Re: Short circuit tests in GR-1089
Hello All: Just a quick followup on our discussion about the short circuit tests: I just received my copy of Issue 3 of GR-1089, and when I went to replace Issue 2 I found a 1-page bulletin from Telcordia, dated December 1999, tucked in the front of my Issue 2 binder. The bulletin specifically addresses the "permissible response" of equipment to the short circuit test. Following is the text from that bulletin: "It is the Telcordia Technologies interpretation that when the short-circuit is applied to a circuit pack the operation of a fuse, circuit breaker, semiconductor fuse (e.g., diode, transistor, FET), and/or other current-limiting means (e.g., fold-back) without a fire, electrical safety, or fragmentation hazard is permissible. Further, it is the Telcordia interpretation that a sacrificial element such as a fuse, diode(s), transistor(s), semiconductor, or polymer over-current device(s) may fail as part of this test. The circuit pack is not required to be operational following this test. However, compliance shall be demonstrated by application of cheesecloth as specified in GR-1089-CORE, Section 4.5.2." I think the above statement clears up most, but not all, of the uncertainty regarding permissible failure modes under Issue 2. Of course, now we have Issue 3 with language that is not identical to this earlier statement. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. Visit our web site at: http://www.ewh.ieee.org/soc/emcs/pstc/ To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Ron Pickard: emc-p...@hypercom.com Dave Heald: davehe...@attbi.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org All emc-pstc postings are archived and searchable on the web at: http://ieeepstc.mindcruiser.com/ Click on "browse" and then "emc-pstc mailing list"
Re: Short circuit tests in GR-1089
In a message dated 11/27/2002, Marko writes: > So what did you decide? Is a fuse blowing acceptable? > I'm sure others would be interested as well. > Hi Marko: So you want me to go on the record so I can get flamed? OK, here goes: The revised text in Issue 3 of GR-1089 (kindly posted to the group by Alain Servais) explicitly states that fuses are acceptable. The text is not 100% clear on whether the fuse has to be located on the module where the short was induced or whether it can be anywhere in the system, but it appears that the fuse could be anywhere. The only remaining gray area is whether the failure of a component or circuit trace is acceptable *provided* that it is located on the same circuit card where the short was induced, *and* that no safety hazard resulted. My interpretation is that this would be acceptable. This conclusion is based on the following considerations: 1) Issue 3 makes it very clear that failure of something other than a fuse is NOT permitted if it located on a module other than the one where the short was applied. 2) Issue 3 also says that the module where the short was applied does not have to work after the test is over. 3) It seems unreasonable to require a circuit card or module to continue to function after a short was applied on that module. The short itself simulates a fault condition that would require replacement of the module. If so, what's the point of requiring that nothing else on the module is damaged? It doesn't serve any of the reliability goals that are part of the implied intent of the requirement. For the module itself, it would seem that a sufficient criterion would be that no safety hazard occurs. I suspect that others may disagree with this interpretation, and I am open to discussion about it. The fact is that I would be inclined to use fuses at that module level on any new design unless I was VERY sure that the failure mechanism would not create a "fire, fragmentation, or electrical safety hazard" as stated in GR-1089. In the case that prompted me to post this question in the first place, the failure mode was an open circuit trace, on an inner layer of the circuit card where the short was applied, in an existing product. Under the circumstances, I am not inclined to insist on redesign of that product to retrofit a fuse. However, I would recommend the inclusion of a fuse in any similar product designed in the future. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Short circuit tests in GR-1089
Hello All: Many thanks to all who responded to my question about this requirement. Many of the responses were quite interesting and persuasive, even though some of them were directly opposed. I think that the expanded description in the new third edition of GR-1089 helps resolve most of the uncertainty I had with regard to this requirement. I was not aware that the new edition of GR-1089 had been issued, so this discussion was doubly useful. Telcordia has now received some of my $$ for a copy of the new edition. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. Visit our web site at: http://www.ewh.ieee.org/soc/emcs/pstc/ To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Ron Pickard: emc-p...@hypercom.com Dave Heald: davehe...@attbi.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org All emc-pstc postings are archived and searchable on the web at: http://ieeepstc.mindcruiser.com/ Click on "browse" and then "emc-pstc mailing list"
Short circuit tests in GR-1089
Hello All: I am hoping that some of you can help clarify the intent of requirement R9-20 in Telcordia GR-1089. Taken literally, the requirement says that there shall be no damage to equipment, conductors, or components when the DC power supply is shorted at the load. This could even be interpreted to preclude the use of a fuse that has to be replaced. One test lab has told me that as long as no fire hazard is created from this test, it is considered to have been passed. Needless to say, this differs a bit from the literal interpretation. I guess it would help if I had a better feel for the overall goal of the short circuit testing. Any insight on this would be most appreciated. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. Visit our web site at: http://www.ewh.ieee.org/soc/emcs/pstc/ To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Ron Pickard: emc-p...@hypercom.com Dave Heald: davehe...@attbi.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org All emc-pstc postings are archived and searchable on the web at: http://ieeepstc.mindcruiser.com/ Click on "browse" and then "emc-pstc mailing list"
Re: Clearance requirements for Telecomm PWB design
In a message dated 10/21/2002, Paul Denomme writes: > I am currently working on a design for the telecom industry. I need to > determine the minimum specifications for setting up my design rules for > both internal and external layers and connector selection. According to > UL60950 table 2N the minimum clearance that I need to adhere to is .2mm > (voltages up to and including 125 V r.m.s or DC). This translates to a > .008" clearance. Will this be acceptable? > Hi Paul: If your circuits interface to telephone lines, you first need to determine whether the type of circuit is TNV-1, TNV-2, or TNV-3, per clause 2.3.1 of UL 60950, Third Edition. Once you classify the circuit, other clauses that may apply include 2.3.2 and all of clause 6. In general, creepage and clearance distances for separating telecom circuits from SELV and from earth ground are somewhat larger than what you mention above, but the specific requirements that apply depend on the TNV classification and on the equipment installation. If you can provide more details either to the group or to me offline, I would be happy to try and give you more specific answers to your questions. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Conducted Rf Failures...fixes?
In a message dated 10/8/2002, Alex McNeil writes: > As I am in panic mode, is there some guidance out there for me e.g. is it > usually a matter of changing the interface IC's or hanging special > components onto the lines? > Hi Alex: This test is indeed a difficult one to pass, but in general the transformer-type interfaces that you are dealing with are less susceptible than some other types. I have seen ISDN and Ethernet interfaces pass this test without the need for extreme measures. In the absence of more specific information about your situation, I can offer just a few general guidelines: 1) EN 55024 allows you to use either a CDN or a clamp to induce the RF. Some people find that they get more favorable results with the clamp than with the CDN. 2) The interference is applied common mode. In *theory* your transformer coupled interfaces will have fairly good common mode rejection, but this can be compromised by a bad layout or by some types of surge protection devices. 3) If you have to add some filtering to pass, adding a common mode choke is a good first step toward reducing susceptibility. Adding caps from each conductor to chassis ground also helps, especially if used in conjunction with a common mode choke, with the choke on the line side of the caps. Note, however, that high value caps can interfere with proper operation of the circuit. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Repeat postings
In a message dated 9/19/2002, Ken Javor writes: > I am getting multiple postings, even I think into the next day. > Me too. I have even seen my own postings come back around with the "CC" field populated with multiple copies of the listserver address, even though that is not how the CC field looked when the original posting was sent out. This appears to be happening to other people's postings as well. My guess is that there is a problem with the listserver. I have copied "owner-emc-p...@majordomo.ieee.org" on this email in an effort to reach the list administrators. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com In a message dated 9/19/2002, Ken Javor writes: I am getting multiple postings, even I think into the next day. Me too. I have even seen my own postings come back around with the "CC" field populated with multiple copies of the listserver address, even though that is not how the CC field looked when the original posting was sent out. This appears to be happening to other people's postings as well. My guess is that there is a problem with the listserver. I have copied "owner-emc-p...@majordomo.ieee.org" on this email in an effort to reach the list administrators. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Repeat postings
In a message dated 9/19/2002, Ken Javor writes: > I am getting multiple postings, even I think into the next day. > Me too. I have even seen my own postings come back around with the "CC" field populated with multiple copies of the listserver address, even though that is not how the CC field looked when the original posting was sent out. This appears to be happening to other people's postings as well. My guess is that there is a problem with the listserver. I have copied "owner-emc-p...@majordomo.ieee.org" on this email in an effort to reach the list administrators. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Pulse Dialing
In a message dated 9/19/2002, Matt Aschenberg writes: > Our engineering team is looking for a spec that defines U.S pulse dialing. > Any suggestions? > Hi Matt: As you may know, there are no regulatory requirements for pulse dialing in FCC part 68. This means that any requirements you apply will be self-imposed. Simply stated, US pulse dialing uses a 60/40 break/make ratio and a nominal 10 pulses per second. Thus, each pulse consists of a nominal 60 mS break followed by a 40 mS make. The number of pulses corresponds to the digit dialed, except that a "0" is ten pulses. If you want a detailed specification for pulse dialing, you can use TIA/EIA-470-B, "Telephone Sets With Loop Signalling." Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Measuring AC Line Impedance
In a message dated 9/19/2002, you write: > So far, I've taken a large resistive load and measured the voltage drop on > the AC line. From that I calculated the total impedance of the AC line. > However, as you may suspect, with a resistive load, the power factor is > 1.0. > So I can't vectorly, calculate the resistive and inductive components. > Hi Dave: I am not familiar with the specific measurement you are trying to make, so the following may not be applicable to your situation. However, some of the basic concepts may be useful. In the telecom business it is sometimes necessary to know the complex impedance presented by a 2-wire port. In theory, if you drive the port with a voltage source applied through a known resistance, you can calculate the complex impedance of the port based on measurements of the following two things: 1) The voltage drop across the source resistor 2) The phase of the current through the resistor, relative to the phase of the source I designed a simple fixture to measure the voltage and phase, then derived the necessary equations and created a spreadsheet to calculate the complex impedance. Soon after completing my measurements this way, I got a good price on a used test instrument that makes this measurement directly. I was very pleased to find that the values measured by the test instrument matched my calculated values almost exactly. This gave me good confidence in my earlier test method, even though I no longer need to use it. It might be possible for you to adapt this method for the test you want to make. If so, I would be happy to send you a copy of the equations and a copy of the spreadsheet I used. Since this was developed internally for my own use, the notes are a little sketchy, but I think I could fill in the gaps with a telephone conversation. While I am not familiar with the details of the test you are trying to perform, I can see that one possible complication would occur if the test is supposed to be performed under a specified load condition. In that case, you would have to find a way to ensure that the AC impedance of your load does not affect the measured impedance of the mains. I think there are ways to accomplish this either with a test fixture or by making the spreadsheet calculations take the load impedance into account. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com In a message dated 9/19/2002, you write: So far, I've taken a large resistive load and measured the voltage drop on the AC line. From that I calculated the total impedance of the AC line. However, as you may suspect, with a resistive load, the power factor is 1.0. So I can't vectorly, calculate the resistive and inductive components. Hi Dave: I am not familiar with the specific measurement you are trying to make, so the following may not be applicable to your situation. However, some of the basic concepts may be useful. In the telecom business it is sometimes necessary to know the complex impedance presented by a 2-wire port. In theory, if you drive the port with a voltage source applied through a known resistance, you can calculate the complex impedance of the port based on measurements of the following two things: 1) The voltage drop across the source resistor 2) The phase of the current through the resistor, relative to the phase of the source I designed a simple fixture to measure the voltage and phase, then derived the necessary equations and created a spreadsheet to calculate the complex impedance. Soon after completing my measurements this way, I got a good price on a used test instrument that makes this measurement directly. I was very pleased to find that the values measured by the test instrument matched my calculated values almost exactly. This gave me good confidence in my earlier test method, even though I no longer need to use it. It might be possible for you to adapt this method for the test you want to make. If so, I would be happy to send you a copy of the equations and a copy of the spreadsheet I used. Since this was developed internally for my own use, the notes are a little sketchy, but I think I could fill in the gaps with a telephone conversation. While I am not familiar with the details of the test you are trying to perform, I can see that one possible complication would occur if the test is supposed to be performed under a specified load condition. In that case, you would have to find a way to ensure that the AC impedance of your load does not affect the measured impedance of the mains. I think there are ways to accomplish this either with a test fixture or by making the spreadsheet calculations take the load impedance into account. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Measuring AC Line Impedance
In a message dated 9/19/2002, you write: > So far, I've taken a large resistive load and measured the voltage drop on > the AC line. From that I calculated the total impedance of the AC line. > However, as you may suspect, with a resistive load, the power factor is > 1.0. > So I can't vectorly, calculate the resistive and inductive components. > Hi Dave: I am not familiar with the specific measurement you are trying to make, so the following may not be applicable to your situation. However, some of the basic concepts may be useful. In the telecom business it is sometimes necessary to know the complex impedance presented by a 2-wire port. In theory, if you drive the port with a voltage source applied through a known resistance, you can calculate the complex impedance of the port based on measurements of the following two things: 1) The voltage drop across the source resistor 2) The phase of the current through the resistor, relative to the phase of the source I designed a simple fixture to measure the voltage and phase, then derived the necessary equations and created a spreadsheet to calculate the complex impedance. Soon after completing my measurements this way, I got a good price on a used test instrument that makes this measurement directly. I was very pleased to find that the values measured by the test instrument matched my calculated values almost exactly. This gave me good confidence in my earlier test method, even though I no longer need to use it. It might be possible for you to adapt this method for the test you want to make. If so, I would be happy to send you a copy of the equations and a copy of the spreadsheet I used. Since this was developed internally for my own use, the notes are a little sketchy, but I think I could fill in the gaps with a telephone conversation. While I am not familiar with the details of the test you are trying to perform, I can see that one possible complication would occur if the test is supposed to be performed under a specified load condition. In that case, you would have to find a way to ensure that the AC impedance of your load does not affect the measured impedance of the mains. I think there are ways to accomplish this either with a test fixture or by making the spreadsheet calculations take the load impedance into account. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: DSL Modems
In a message dated 9/17/2002, Rob Keller writes: > So, because of the ring signals, which would exceed the > limits of SELV, connecting to the input the modem, should the > classification be TNV-3. Hi Rob: If the tip/ring pair that the DSL modem interfaces to has power ringing on it (such as for an ADSL modem used on a standard POTS line), the classification would certainly be TNV-3. It really doesn't matter that the DSL modem does not use the ring signal. What matters is that the signal is present. Note that some DSL modems (HDSL, SHDSL) operate on lines that do not have power ringing. For these applications, the typical classification is TNV-1. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Replacements for Gas Tubes
In a message dated 9/12/2002, Scott Douglas writes: > We use gas tubes in our products for protection against lightning and/or > surge. We now have an edict to get them out of all of our products. So what > does one use to provide similar protection without using a gas tube? Hi Scott: Well, the answer depends on what sort of signal lines you are trying to protect and the level to which you are trying to protect them. The main attributes of gas tubes are their low capacitance (good for high speed signal lines) and their ability to carry huge surge currents (good for cases where there is no other primary protection device). There are many disadvantages associated with gas tubes, such as the fail-open mode that you mention and their slow response. There are some other, more esoteric phenomena associated with gas tubes, such as their tendency to have a "glow voltage" in the transition region where a plasma forms and they get red hot. In general, I avoid gas tubes whenever possible, due to the disadvantages listed above. If your lightning surges are limited to less than a few hundred amps, you may be able to use an appropriate Teccor sidactor. However, sidactors have greater capacitance, and this capacitance can be nonlinear in certain configurations, so you have to be careful that you don't compromise signal integrity. In summary, the correct answer to your question would require a more detailed understanding of the requirements of your particular application. If you are willing to provide more information, feel free to post to the group or contact me offline. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Grounding architectures for COs
Marko Radojicic writes: > Battery Return is also problematic. In NA, it *must* be held isolated from > CG/FG. In the EU, it *must* be bonded to the CG/FG! > Hi Marko: Thank you for your very informative response. I would just like to clarify any possible distinction between battery return and signal ground when we are talking about isolation from frame ground. You point out that the "requirement" to isolate signal ground from frame ground is now routinely violated in North America, and seems to be fading away as a requirement. Is this also true for the "requirement" to isolate battery return from frame ground? An example of where isolation becomes problematic is an analog PSTN SLIC (Subscriber Loop Interface Circuit) line card. Here we are feeding DC current to the loop directly from the -48V battery. The circuit is typically implemented using an integrated SLIC chip that has inputs for both the -48V battery and various low level signals such as transmit, receive, and some logic control. Since both battery and signal ground coexist on the same physical silicon chip, it is not possible to separate them in the chip. To meet the isolation "requirement," I have seen some designers go to extraordinary lengths to fully isolate the codec-SLIC subsystem from the remainder of the electronics. This is typically done with an array of high speed optical isolators inserted in the PCM highway. I have never implemented this isolation in my own designs, and so far I have not had a problem with any isolation "requirement" for the battery. However, I see the confusion on this subject and I do see some other designers implementing an isolation barrier so that the -48V battery circuits are fully isolated from frame ground within the equipment. Can you clarify whether your earlier remarks are applicable to battery return as well as signal ground? Thanks, Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Isolated grounds in central office
In a message dated 9/10/2002John Juhasz writes: > Be careful Dorin. For Central Offices, they need to be isolated. > Hi Guys: I changed the subject heading on this response (was "Inrush and EN61000-3-3") to better reflect the direction this thread has taken. The subject of isolating circuit ground from frame ground in a CO seems to be controversial. On one hand, I have heard stories such as John's about various RBOCs insisting on this. On the other hand, I have not seen the actual written requirements that were being imposed or the rationale behind them. For safety, functionality, and EMC considerations, I think it is generally better to tie circuit ground to frame ground in a robust fashion. One person told me that Verizon had insisted that circuit ground be brought out to an isolated terminal post that could then be externally tied to frame ground. This made a mess out of his design, and so far I have avoided this approach in my own designs without getting nailed. It seems that there is some confusion about whether this isolation is required and if so, why it is required. The whole subject of grounding within a CO can become quite complicated, especially when lightning is taken into consideration. I understand that TR-NWT-000295, "Isolated Ground Planes: Definition and Application to Telephone Central Offices" addresses some of these issues, but not in a clear manner (imagine that for a Telcordia document!). If anyone can provide a clear explanation of when isolation of circuit ground and frame ground is required in a C.O. and why, I would be most interested to know. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Curious about Tripp-Lite "isolation transformer"
In a message dated 8/30/2002, mra...@attbi.com writes: > What you got is exactly as stated in their specification. > > http://www.tripplite.com/products/product.cfm?productID=228 > This helps a lot. I had performed a web search earlier for specs on the IS-1000, but I did not turn this up. Thanks, Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Curious about Tripp-Lite "isolation transformer"
In a message dated 8/30/2002, Don Borowski writes: > But there is common mode supression (hot and neutral against safety ground) > offered by the transformer, assuming that the capacitive coupling between > the > primary and secondary is minimal. > > ...There is one effect of a long run of bundled conductors - the > disturbance (at > least at frequencies above several hundred kHz) becomes a floating source > at the > power outlet end. The transformer will break the common mode current loop. Hi Don: Your explanation makes sense if we focus on high frequency common mode disturbances where the inductance of the ground wire is significant enough to make a difference. In this case, having the ground wire tied to one side of the AC at the service entrance does not ensure that these two wires are at the same instantaneous potential at the AC outlet on the other end of a long wire. At low frequencies, it does not appear that the isolation transformer provides any benefit for common mode noise. Regarding my surge scenario that Bob Johnson commented on, it seems that this type of isolation transformer provides only limited protection from lightning surges. Common mode surges coming in to the service entrance will be converted to differential surges when one side is tied to local earth, and these differential surges will sail right through the transformer (subject to its frequency response transfer function). On the other hand, ground strikes that raise the local earth potential at the service entrance will not be passed through on the AC pair, but they will come right through on the ground wire. While this is different from what would happen without the isolation transformer, I'm not sure it provides any added protection. In summary, it sounds as though this type of isolation transformer is primarily intended to attenuate high frequency common mode noise. Is this correct, or have I missed something? Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Curious about Tripp-Lite "isolation transformer"
In a message dated 8/30/2002, Chris Maxwell writes: > Most "isolation" transformers have a non-magnetic, metal shield between the > primary and secondary. The shield does not appreciably degrade the > transformer's magnetic coupling at 60Hz; however, it does attenuate the > high frequency differential noise passed from primary to secondary. > Hi Chris: Thanks for your response. This transformer does indeed have a grounded shield between the primary and secondary. I guess I don't know enough about transformer design to understand why the shield would attenuate differential noise. I can certainly see how it would reduce common mode coupling at high frequencies, since the grounded shield will reduce the inter-winding capacitance. Do you know of some typical applications for this type of "isolation" transformer? There must be enough demand for them, since Tripp-Lite and a few other manufacturers offer them as a catalog item. In the mean time, I plan to modify the wiring as I described previously to obtain a floating, ungrounded AC supply. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Curious about Tripp-Lite "isolation transformer"
Hello All: Some of the laboratory tests that I perform require a source of AC mains power that is fully isolated from earth. To provide this function, I made a catalog purchase of a Tripp-Lite model IS-1000 "isolation transformer." This unit is equipped with a grounded (3-prong) 120 VRMS plug on the AC mains cord used on the input side, and has a 3-prong AC mains socket on the isolated side. When the unit arrived, I was surprised to find that the green-wire ground on the input side was carried across to the ground pin on the AC socket of the "isolated" side. Furthermore, one side of the 120 VRMS socket on the output side was also tied to the green wire ground. In other words, the output of the unit is not isolated from ground, even if I use a "cheater adapter" to break the connection of the ground pin on the output socket. I also noted that the unit has capacitors and what appears to be a large MOV placed across the "isolated" output. My theory is that this type of isolation transformer is intended to provide protection from transients or noise on the AC mains. However, I can't see what protection the transformer provides, except maybe from common mode transients on the AC mains pair. Even so, I am having difficulty visualizing any scenario where a common mode surge would appear on the AC mains pair but not on the ground wire (given that in North America, one side of the AC mains pair is tied to ground at the service entrance). Is it possible that the transformer is designed with frequency-dependent coupling so that frequencies above 60 Hz do not couple well? Such a scheme would provide some degree of protection from differential noise. I am wondering if any of you are familiar with this type of "isolation transformer" and the applications where someone might need such a device. On the surface, I can't see what "isolation" benefit the transformer is providing when it is wired this way. It is relatively easy for me to modify the wiring to provide the fully floating, balanced AC output that I was seeking in the first place. The only thing I haven't decided yet is whether to leave the capacitors and MOV connected across the isolated output. If any of you are familiar with this type of "isolation transformer" and can explain the intended function, I would be interested to learn about it. Also, if you see any flaws in my rewiring plan, please let me know. I'm a little bit uneasy modifying a design that I don't fully understand. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. Visit our web site at: http://www.ewh.ieee.org/soc/emcs/pstc/ To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Ron Pickard: emc-p...@hypercom.com Dave Heald: davehe...@attbi.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org All emc-pstc postings are archived and searchable on the web at: http://ieeepstc.mindcruiser.com/ Click on "browse" and then "emc-pstc mailing list"
Re: Insulation Distance Between Circuitboard Layers (Safety)
In a message dated 8/28/2002, Chris Maxwell writes: > Can anyone see any "pinholes" in my reasoning? Can anyone recall the thread > regarding multiple layers of thin insulation? > Hi Chris: Wow, you are really pushing the limits with your board design. I work with EN 60950, and the clause in the third edition that is relevant to your question is 2.10.5.3, "Printed Boards." The default requirement is 0.4 mm insulation thickness, but there are alternative methods listed in Table 2M. For instance, you can use three layers of sheet insulating material including pre-preg. You can even use two layers if you run a routine production test for electric strength. I don't know how much of this is relevant to EN 61010-1, but hopefully the info on 60950 will be useful in pursuing this. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Creepage on PCB Internal Layers
In a message dated 8/15/2002, Gary McInturff writes: > Okay, now don't everyone write to my layout guys, but I never bothered to > tell them they could change the constraint files for closer routing on > inner layers, so they do apply the spacings. I have try to keep an eye for > it causing us extra layers or any real grief and it hasn't. > Hi Gary: I agree that for designs that involve only a single telecom port or single power supply, there is often no incentive to take advantage of the tighter spacings allowed on inner layers. For such jobs it is easier for the layout folks if they have one set of simple rules. However, on high density cards with multiple ports, the tighter spacings allowed on inner layers can be a godsend. It is important to keep this in mind when you plan your layout. Also keep in mind that another degree of freedom that was not discussed in this thread is using layer-to-layer separation for the isolation barrier. That invokes another set of complicated considerations, but on some high density boards it can make the difference between "feasible" and "impossible." Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Creepage on PCB Internal Layers
In a message dated 8/13/2002, you write: > To me, it's sort of funny in that it just says that the Creepage and > Clearance distances do not apply on inner layers of void free PCBs. > That's nice; but I can't find where a distance is specified. I mean, I > would think that there should be some minimum distance between an AC line > and a 5V SELV line on an inner layer of the board > > Our layout guy has a military specification that specifies insulation for > 300-500V circuits. It specifies .100" on external layers and .010" on > internal layers...which would work out to a factor of 10 reduction between > surface layers and inner layers Hi Chris: I work with EN 60950, not EN 61010-1, so my comments here are only general. In EN 60950 the creepage and clearance requirements can be eliminated for inner layers, but all other requirements for the relevant insulation still apply. For example, both supplementary insulation and reinforced insulation still require a minimum of 0.4 mm through solid insulation. Also, any relevant hipot tests for the insulation in question still apply. In general, a well constructed circuit board will have extraordinary insulation between traces on the same layer, due to the high dielectric strength of the prepreg and the bonding agents that are used. Even a separation .002" will, in theory, have a dielectric strength of more than 1000 volts. However, I prefer to be cautious here, in the event that a defect creates a small air gap in the isolation area. I would suggest using as large a gap as you can reasonably achieve, even if the regulations call out no minimum or a minimum of only 0.4 mm. The 0.8 mm target that you have nominated sounds OK to me, but if you can bump it up to 1.0 mm or more, that would be nice. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: NEBS compliance for 100baseT / 1000base T
Hi Muhammad: The discussion on this thread has prompted me to think further about why I have not used integrated magnetics for NEBS applications in the past. My earlier response was based on what I could recall at the moment, and was only partially accurate. If you use a device that has a 1500 VRMS rating, it will most likely have no trouble with the common mode surge, so that is probably not an issue. As Scott Lemon noted earlier, very little energy will get through the transformers on a common mode surge, so any protection on the secondary side will not have to deal with large currents that could upset your ground plane. I think the key concerns for using integrated magnetics are the differential lightning surge and the 120 VRMS power cross test. To pass the tests in GR-1089, the primary winding will have to survive a 100 amp, 2x10 uS differential surge. The small wire used in the integrated transformers may have trouble surviving this. Due to the fact that the primary side is not accessible outside the jack, you will not be able to put a sidactor across the primary winding to help protect it. The other issue you need to watch out for is the failure mode during the 120 VRMS power cross test. This test is applied through a "line simulator fuse" that must not blow open during the test. In other words, your primary side must fail open before the line simulator fuse opens. Compliance with this requirement is typically guaranteed by putting a fuse in series with the primary side, but with the integrated magnetics this is not possible. You will be depending on the fusing characteristics of the primary winding itself, which may be unpredictable. I think the above considerations are the ones you should evaluate carefully before committing to the use of an integrated magnetics jack. Failing either of these tests after the board design is complete would represent a major setback for your development program. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: NEBS compliance for 100baseT / 1000base T
In a message dated 7/22/2002, Muhammad writes: > For boards with copper ports (100baseT and 100base T), is it possible to use > rj45 connectors with integrated magnetics and still get pass these tests. Hi Muhammad: Previous respondents have clarified that typically you will be subject to the intrabuilding lightning surges and the 120 VRMS power cross test. They have also noted that you can avoid the lightning tests if you use a shielded cable that is grounded on both ends (although such grounding creates its own EMC issues that this group has discussed previously). In the past, I have avoided using the RJ-45 jacks with integrated magnetics because I was skeptical about their ability to survive the lightning surges. The very thin wire used on the windings is susceptible to damage from the 100 amp differential surge, and the isolation barrier is susceptible to damage from the 1500 volt common mode surge. You can probably provide adequate differential protection to protect the winding wire during the differential surge, but the isolation barrier has to take the full 1500 volt common mode surge (providing protection to ground would not be cost effective or space-efficient). If you locate a supplier of integrated magnetics that claims they survive the intrabuilding surges, I would be interested to know. In the absence of a GR-1089 compliant, integrated-magnetics device, the two most viable options are: 1) Use an RJ-11 with integrated magnetics and then use the shielding exemption to avoid the lightning tests. 2) Use external magnetics with more robust isolation that can handle the 1500 volt common mode surge. I have used the second approach successfully on several designs where NEBS compliance was required. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Australian Telecom test
In a message dated 7/8/02, Dieter Baldamus writes: > I have a telecom unit that already passed all tests according to CS03, FCC > part 68 and NRT3. What are the chances to pass the Australian TS001 and > TS002 without modifying the unit? > > Hi Dieter: The following items are the key ones you should check to see if changes are likely to be needed: 1) Return loss requirements for Australia differ from the requirements you listed above, and are not easy to meet. 2) Pulse dialing in Australia requires special limiting of inductive transients, but the easiest way out of this is to not offer pulse dialing. Almost all central offices in Australia accept DTMF dialing. 3) For the first 300 mS after an off-hook transition, the DC V-I must be less than 6 volts at 20 mA. The easiest way to accomplish this is to have the steady-state DC V-I meet this requirement as well, and then have a good transient response that meets the 300 mS requirement. If your existing design already has a pulse dialing shunt relay, you can use this instead. 4) Safety isolation requirements for creepage and clearance are identical to EN 60950, but the hipot and surge test levels are higher. These four parameters are the ones that I usually see existing FCC designs fail. Also, if your product is a voice system, you may be subject to TS 003 and/or TS 004. These requirements are not likely to be met unless the design was done with them in mind. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. http://www.randolph-telecom.com
Re: Telecomm Safety clearance
In a message dated 7/8/02 10:36:19 AM Eastern Daylight Time, Alex McNeil writes: > Q1. Is 2.5mm OK (Norway and Sweden are to be included) and what is the > minimum? > Q2. Does this "creepage and clearance" need to apply to all 6 layers (and 4 > layers)? > Hi Alex: I am out of my office for the next week, so I do not have access to my copy of EN 60950 to give you specific clause references. However, the following points should be considered in your design: 1) In the Third Edition of EN 60950, Annex ZB no longer requires supplementary insulation for Norway and Sweden. Only the 0.4 mm distance through insulation requirement has been retained, along with the stipulation that capacitors bridging the isolation must be Y2 caps. The creepage/clearance requirement now reverts to the requirement called out in the main body of the standard (typically 1.6 mm and 1.0 mm). That being said, I continue to use the supplementary values of 2.5 mm and 2.0 mm in designs where the room is available in the layout. 2) In the section of EN 60950 that addresses creepage/clearance, there is an explicit section for printed circuit boards. I won't try to summarize all the requirements here, but some key points are that separation between traces on the same layer is treated as solid insulation, and separation between traces on different layers does not necessarily have to meet the 0.4 mm requirement if you are willing to subject the board construction and/or testing to additional constraints. I don't necessarily recommend using these exemptions, but you should be aware they are available if your layout is extremely dense. 3) Remember that the required separation is between your TNV-3 circuits and SELV/GROUND. There is no regulatory requirement to separate different parts of the TNV circuits from each other. The only considerations you need to think about for separating different parts of the TNV circuits are various potential differences that can develop during on-hook conditions, ringing, lightning, and power cross. 4) If your product is a simple modem with only one phone line interface, the simplest layout approach is to create a TNV "island" around the phone jack. If the TNV connector is a conventional modular jack that is edge mounted, make a 2.5 mm wide, U-shaped, keepout "moat" around the TNV circuits that captures the jack in the open end of the U. To simplify verification of the isolation barrier, it is easiest to make a 2.5 mm moat appear in the same location on all layers, but this is not required by EN 60950. There should be no power or ground plane in the moat or in the TNV island. The only devices that should bridge the moat are isolation devices such as transformers, optos, and Y2 caps. If you have room available in your layout and can follow this simplified approach, it will be very easy to verify compliance with the isolation requirements. The only other thing you will have to watch out for is separation from grounded mounting holes and from parts of any metal housing material. If you still have questions about this, feel free to call me after I get back to my office next week. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Transient Voltage Suppressor - Europe approvals?
In a message dated 6/19/2002, Dave Hutchins writes: > Energy has nothing to do with the capability of the TVS device. It is also > proven that the energy in the transient threat is not the energy dissipated > in the TVS component, just ask any old timer even those at Harris. Hi Dave: I'm not sure I agree with you on this one, but perhaps the issue is simply one of semantics. The "energy" I referred to was the energy in the surge, not the protection device. For example, a gas tube can handle, without damage, surges of far higher energy than a Transzorb. If I recall correctly, MOVs can also typically handle surges of higher energy than a Transzorb. In the past, when I have used a Transzorb or comparable device, it was for tertiary protection inside a circuit that was already protected by other devices with the ability to handle much higher energy surges. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Transient Voltage Suppressor - Europe approvals?
In a message dated 6/19/2002, Bob Wilson writes: > Transzorbs (or "TVS"s) are basically zener diodes that are rated for > their energy absorption capability. Hello All: One thing to keep in mind is that the energy handling capability of a TransZorb is generally not very high compared to other devices such as gas tubes, MOV's, and sidactors. Michael, your original posting did not specify the intended application, but the reference to being "rated for mains voltage" makes me wonder if you are trying to protect from surges on an AC mains input. If so, I do not think that a TransZorb would have adequate energy handling capability. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: FCC part 68 vs CTR21
In a message dated 6/19/2002, Alex McNeil writes: > For the PSTN analogue modems I am not so sure: > FCC Part 68 vs CTR21 (TBR21)? > Can anyone offer some advice here? > I assume if the product meets CTR21 then the modem should function OK in > Italy? > Does Italy have other modem details that I should know about? > Hi Alex: If you modem was not designed to meet CTR 21, it will probably fail the CTR 21 tests for current limiting and for transient response. Fortunately, CTR 21 is no longer a mandatory requirement under the RTTE directive. The main regulatory requirements that apply are those for safety and EMC. Regarding general compatibility with the Italian network, keep in mind that dial tone in Italy is cadenced, so your North American dial tone detector may not properly detect it. As far as the other parameters are concerned, it is hard to say with certainty without reviewing the actual design. Since FCC Part 68 is a harms-based standard, the tests are only intended to show that the equipment will not harm the public network. It is not necessary for the equipment to actually function at all, even on the North American network. That being said, I do think that a well designed modem intended for North America is likely to operate reasonably well on the Italian network, provided that the cadenced dial tone does not present a problem. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Lightning and power cross in Japan
Hello All: I am trying to determine whether there are any published requirements for lightning immunity or power cross for wireline PSTN terminal equipment in Japan. The JATE requirements for PSTN terminal equipment do not contain any such tests, and I am not aware of any similar requirements imposed by various industry specifications. Do any of you know of any specific requirements for Japan? On a related note, how severe is the actual lightning environment in Japan, compared to say, North America or Europe? Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. Visit our web site at: http://www.ewh.ieee.org/soc/emcs/pstc/ To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Ron Pickard: emc-p...@hypercom.com Dave Heald: davehe...@attbi.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org All emc-pstc postings are archived and searchable on the web at: http://ieeepstc.mindcruiser.com/ Click on "browse" and then "emc-pstc mailing list"
Re: R&TTE Directive
In a message dated 5/8/2002, John Juhasz writes: > it appears that it is not required to test the interface itself to > 'telecom specs' such as those in the CTR21. > Hi John: Your interpretation is correct. For wireline POTS products/interfaces, the only requirements that apply under the RTTE directive are safety and EMC. There are no regulatory requirements whatsoever for the type of specifications called out in CTR 21. CTR 21 ceased to be a regulatory requirement when the RTTE directive came into force on April 8, 2000. That being said, many manufacturers are uncomfortable with having no regulatory requirements for telecom. One option is to continue voluntary testing to CTR 21. Another option is to obtain the individual "reference" standards that the RTTE directive requires each operator of a public network to publish. Most of the major operators have posted these on their web sites. A third option is to review CTR 21 and the relevant "reference" specifications, then use engineering judgment to design and test your interface. This is the preferred approach if you are seeking minimum cost and/or maximum compatibility with the various national networks. The bottom line is that the telecom aspects of your product performance are now a matter between you and your customer, rather than between you and the regulators. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: How to Minimize Global Approvals?
In a message dated 4/19/2002, you write: > My "summarized" Analogue Modem Requirements > TBR21 + EG 201 121 (Europe) > FCC Pt 68 (Mexico and Canada acceptance?) > TS002 (Australasia, no PTC 200?) > Hi Alex: As Ron Pickard has already pointed out, Mexico's telecom requirements are not fully harmonized with FCC Part 68, so there is a separate test and approval process for Mexico. I just wanted to add a couple comments about other aspects of your proposed plan: 1) As you know, TBR 21 and EG 201 121 are strictly voluntary, so if you are seeking to cut testing costs this is something you should look at. If you have good confidence in your modem design and you do not expect your customers to insist on this testing, it is an avoidable expense. I typically perform internal testing for those parts of TBR 21 and EG 201 121 that are relevant, but the formal testing at an outside lab is now of questionable value. 2) You use the term "Australasia" in reference to TS 002. Some people use this term to refer to Australia and the entire Asian region. Please keep in mind that a TS 002 report will only get you approval in Australia. Other countries, such as New Zealand, China, Taiwan, and Singapore each have their own requirements that are not identical to TS 002. Fortunately, there is considerable overlap in the requirements of the different countries, so a carefully designed modem can meet the requirements of all these countries. Other than a few software-controlled characteristics related to dialing and such, the only hardware related parameter where there is a conflict is return loss (Australia and Singapore can not both be addressed with a single compromise impedance). 3) It's not clear whether your plan for worldwide approvals implies that you have a single, worldwide modem design, but you may want to give some thought to how you will address the various worldwide requirements. With some advance planning it is possible to have a single design that can be used worldwide, but without such planning you will likely be forced to have multiple versions of your modem. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Analog PSTN approvals in Czech Republic
In a message dated 3/7/2002, John Woodgate writes: > If such a modem meets the 60 mA requirement, is there really a problem? > Hi John: A modem that does not provide 60 mA limiting will be suitable for use in a wider range of countries, and will be less expensive to implement. The need for this limiter is being phased out in France, the only country that ever actually needed it for network compatibility. While the limiter was never needed on the Czech republic network, it is included in TBR 21 and is therefore technically a requirement if the manufacturer chooses the optional TBR 21 compliance path in the Czech Republic. In the past, if the manufacturer chose to use the Czech Republic national requirements instead, there were other onerous requirements on the ringer impedance. Based on input from another member of the group, there is evidently a new Czech national requirement that eliminates the unusual ringer requirement and also requires no current limiting. This represents some welcome progress toward harmonization of analog PSTN requirements. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Analog PSTN approvals in Czech Republic
In a message dated 3/7/2002, Roger Magnuson writes: > According to TZP 018, ring impedance should be >= 4 kohm i.e. there is no > upper limit anymore. > Hi Roger: Thanks for the update. Is TZP 018 available on the web? I would like to obtain a copy to review. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Analog PSTN approvals in Czech Republic
Hello All: Recent discussions on the emc-pstc listserver indicate that the Czech Republic has made an effort to harmonize their regulatory scheme with the European Union's RTTE directive. I have reviewed the text of the Czech legislation, Act 151/2000, and find that it has some similarities to the RTTE directive, but also has significant differences. It is also unclear how far they have moved in actually implementing the new law. Do any members of the group have up-to-date information on the Czech requirements for approval of analog PSTN terminal equipment, such as a modem? The last I heard, applicants could either comply with an old Czech national standard that had unusual requirements for ringing impedance, or they could comply with TBR 21 and its unusual requirement for 60 mA current limiting. I would like to know if it is now possible to get approval in the Czech Republic for an analog PSTN modem that uses a more "worldwide" type of interface that has high ringing impedance and no 60 mA current limiter. Any insight that members may have on this would be greatly appreciated. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. Visit our web site at: http://www.ewh.ieee.org/soc/emcs/pstc/ To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Ron Pickard: emc-p...@hypercom.com Dave Heald: davehe...@mediaone.net For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org All emc-pstc postings are archived and searchable on the web at: http://ieeepstc.mindcruiser.com/ Click on "browse" and then "emc-pstc mailing list"
Re: Bellcore LATA
In a message dated 2/28/2002, Alex McNeil writes: > I am not too familiar with the North American requrements: > Our customer has requested Surge Protection testing to the Bellcore LATA > Switching General Requirements. 600 Vpk metallic, and 2500 Vpk > longitudinal. > > 1. Are you familiar with this spec? > 2. Are these requirements covered in any of the Regulatory Telecomms or > Safety specs or is our Customer wanting something beyond the regulatory > requirements? > 3. How can I get a copy of this standard? > Hi Alex: These requirements are described in Telcordia (formerly Bellcore) document GR-1089 CORE, "Electromagnetic Compatibility and Electrical Safety - Generic Criteria for Network Telecommunications Equipment." It can be purchased from the Telcordia web site at http://www.telcordia.com/. GR-1089 is not a regulatory requirement. It is simply an industry specification that can be called out on a customer specification. In the USA, compliance with GR-1089 is typically required by the major network operators for equipment that will be used in their central offices. Anyone else can also call out GR-1089 on a purchase specification if they choose to do so. In general, the lightning requirements in GR-1089 are more severe than those ITU K.20 or EN 300 386. Even more significant is the difference in power cross tests, which are much more severe in GR-1089 than the tests called out in K.20 or EN 300 386. While compliance with GR-1089 can be challenging and may add slightly to the cost of your interface circuits, it does provide a reasonable guarantee that the equipment will have very few field failures due to lightning and power induction. This makes it an attractive specification for customers to call out on a purchase specification. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Pollution Degree vs. Creapage Distance
In a message dated 2/4/2002, Chris Maxwell writes: > What would happen (assuming 60950) if you had AC on the component side > of a board and SELV on the solder side. (assuming that any through holes > were properly cleared out). > Hi Chris: That usually works out fine, since the board is treated as solid insulation. The minimum thickness of solid insulation for supplementary or reinforced insulation is 0.4 mm (about .017"). Most boards are much thicker than this. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Pollution Degree vs. Creapage Distance
In a message dated 2/1/2002, John Juhasz writes: > In short, they noted that while possible, it is difficult to maintain the > pollution degree 1 in a printed circuit. Typically the pollution degree > is applied to 'potted' items. > Hi John: In the first edition of EN 60950 there was little guidance on how to treat internal layers of a multilayer board. This led to the kind of discussions that you describe, with each NRTL developing their own interpretation of what requirements to impose. In the current (third) edition, clause 2.10.5.3 explicitly addresses internal layers and eliminates most of the ambiguity. Depending on the construction and compliance method chosen, it is not always necessary to perform routine tests or some of the other onerous tests that used to be imposed. If a manufacturer wants to qualify a solder mask per clause 2.10.6, however, there are still many additional tests that will be required. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Pollution Degree vs. Creapage Distance
In a message dated 1/31/2002, Rick Busche writes: > Does the application of a solder mask allow for a change from pollution > degree 2 to pollution degree 1? I understand that conformal coating > requires significant testing when used to reduce spacings per table 7, but > in this case I am only asking if solder mask can be used to improve the > pollution concern. > Hi Rick: You do not mention which standard you are looking at, but if it is one of the IEC 950 derivatives there are some clauses that specifically address the questions you have. For example, in EN 60950, Third Edition, clause 2.10.5.3 addresses printed circuit boards, and clause 2.10.6 addresses solder mask. To the extent you can use inner layers, clause 2.10.5.3 should provide you some of the relief you seek. I have not ever tried to qualify a solder mask under clause 2.10.6 due to the burden of the additional tests, but you may want to consider it. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Use of resistor on the front end of E1design
Hi Bob: The 120 ohm resistor you mention is typically placed across the line on the transceiver chip side of the isolation transformer. While you might be able to get away with putting it on the line side, I do not see how this would provide any protection. There is a big difference between putting the resistor across tip/ring (as you have suggested) and placing it in series with tip/ring. This particular resistor, which establishes the AC impedance presented to the incoming E1 signals, can not be placed in series with the line. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. Visit our web site at: http://www.ewh.ieee.org/soc/emcs/pstc/ To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Michael Garretson:pstc_ad...@garretson.org Dave Healddavehe...@mediaone.net For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org All emc-pstc postings are archived and searchable on the web at: No longer online until our new server is brought online and the old messages are imported into the new server.
Re: T1/E1 compliance requirements
In a message dated 1/15/2002, mailto:jfinlay...@telica.com";>jfinlay...@telica.com writes: > You'll need to comply with the Intrabuilding Lightning requirements > of GR-1089. I'm not sure about the 12 mil question. > Hi Bob: Joe is correct about the intrabuilding surges in GR-1089. I would like to elaborate with a few additional comments. Assuming your T1 interface is not intended for connection to outside lines, you will only be subject to the requirements for inside wiring (intrabuilding wiring). If you connect to outside lines, the requirements are much more stringent. GR-1089 is not a regulatory test, but if your equipment is intended for central office use in the USA, you will likely be asked to comply with GR-1089. The intrabuilding tests in GR-1089 include both lightning and power cross tests. I would suggest that you stick with the 12 mil traces (1 ounce copper) to make sure they don't fuse open during the intrabuilding lightning tests. You may also need to add some protection devices to make sure that your interface is not damaged by the lightning surges. There is an option to waive the lightning tests if you use shielded cables that are grounded on both ends. To pass the GR-1089 power cross test you will likely need a suitable fuse. Your original posting does not mention whether the present design contains a fuse. If you plan to sell the same product in Europe configured for E1, there is a new regulatory requirement EN 300 386-2 called out under the EMC directive. This standard also contains some lightning tests for intrabuilding wiring, with no exemption for shielded cables. These lightning tests only apply to cables longer than 10 meters that connect to other manufacturers' equipment. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com FEES.doc Description: Binary data
Re: Production Line Hipot Test for Network Powered Equipment
In a message dated 11/27/2001, Jerry Martin writes: > I'm working on a product that is telecom network powered (-190 Vdc) and will > most likely require a production line hipot test. Being network powered, it > is exposed to outside plant surges and has transient protection designed in > (fuse and sidactor). > > Hi Jerry: If the protected circuitry in your product has a permanent connection to earth ground, it may be possible to avoid the production hipot test. For instance, if your telco port was a conventional TNV-1 or TNV-3 interface, there are exemptions available for equipment that has a permanent ground. Unfortunately, the -190 volt power that you mention falls outside of the TNV limits and would normally be classified as a hazardous voltage circuit. USA vendors of T1 equipment with span powering (notably Adtran) have worked with UL to get a special evaluation procedure developed that avoids some of the onerous requirements that would result from classifying the telco port as a hazardous voltage circuit. I do not know if these procedures could be used to provide any relief in your situation. The most straightforward approach for your problem is to simply perform the requested hipot test (with removable links added), but if you really want to avoid this approach there may be a way to use the permanent ground exemptions to eliminate the requirement for having an isolation barrier and performing a hipot test. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: DC voltage ratings
In a message dated 10/25/01, Dave Heald writes: > If you want to ship to Europe, the nominal voltage can be -60Vdc. Here > it gets tricky as charging voltages are typically 72Vdc. Most NRTL's > testing to the new UL60950 will treat DC input voltages at these levels > as TNV-2, which is a lot better for design reasons than treating your > input circuit as a hazardous voltage circuit. > Hello All: Just a followup on Dave's comment above. In Europe (EN 60950), the definition of TNV voltages in clause 2.3.1 allows levels up to 120 VDC. In the USA there is a D2 deviation that severely limits the use of voltages over 60 VDC, but clause 3.6 of UL 60950 explicitly allows centralized DC supplies up to 80 VDC to be classified as TNV-2. Thus, depending on the specifics of the application, the 60 VDC limit in the UL definition of TNV is not a problem in Europe and may not be a problem in the USA either. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. http://www.randolph-telecom.com
Re: TNV Circuits
In a message dated 10/23/01, Gregg Kervill writes: > Have lines ever been struck by Lightning? > > One of my clients had a cupboard full of telephone products destroyed by > all sorts of bazaar events - like the temporary three phase power line to a > factory that "fell' across domestic phone lines > > Anything outside must be considered hazardous and if that hazard is limited > to +_200 then we are VERY lucky. > Hi Gregg: UL 60950 makes a distinction between working voltages that are present all the time (such as span power) and transient overvoltages caused by lightning and power cross. Both types of situations are addressed in the standard, but they are treated differently, as they should be. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. http://www.randolph-telecom.com
Re: TNV Circuits
In a message dated 10/23/01, Perry Hopkins writes: > My question is; by definition, shouldn't the DSL circuits fall under the > TNV-3 and not Hazardous Voltage secondary circuit. The major difference is > when you are determining the creepage and clearance distances. At TNV-3 > you are allowed basic insulation between TNV-3 circuits and SELV circuits > but at the Haz Voltage Reinforced insulation is required. > Hi Perry: If you have the reported 190 VDC span powering on your DSL lines, you may well be forced into the hazardous voltage category. In clause 2.3.1(b), UL 60950 limits TNV voltages to 60 VDC, with voltages up to 120 VDC allowed if current limiting is included. Clause 3.5.3 allows central supply voltages up to 80 VDC to be declared TNV-2 under certain circumstances. However, voltages of 190 VDC exceed any definitions of TNV that I am aware of. The issue of span powering and the requisite insulation requirements created quite a commotion a year or two ago when the Second Edition of UL 1950 was about to become mandatory. As written, the Second Edition would have rendered noncompliant a wide range of T1 equipment that was widely deployed and still in production. As I recall, the effort to bring this to UL's attention was led by Adtran. I seem to recall seeing a document that was worked out between the industry and UL to address the legacy products that used high voltage span powering. I do not recall the details, but if it would be helpful I can look into it for you. I do not know whether the agreement that was reached for these legacy products could be applied to your situation. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. http://www.randolph-telecom.com
Re: Ethernet query
In a message dated 8/29/01, Andy White writes: > When does an Ethernet interface require JATE approval? Is the approval only > reqd. when connected to a Leased Line? or does it apply when connected to a > local building ethernet network? > Hi Andy: Ethernet is not a service offering on the public network, so JATE will not be involved. I have not ever heard of Ethernet being used on leased lines, but I suppose that if you wanted to connect to a conventional leased line you would encounter some regulatory requirements, at least for signal power. However, the limited line length of Ethernet (100 meters?) seems to make it unsuitable for use on leased lines. If your Ethernet port is a standard port for an intra-building LAN, all you should have to concern yourself with is the standard safety and EMC requirements for ITE. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Immunity testing of Headset
In a message dated 8/24/01, Ken Hall writes: > What equipment is generally used to test the acoustic noise levels specified > in CISPR 24 Annex A while performing Radiated and conducted immunity? > Hi Ken: There are special microphones and measuring equipment that can be used for acoustic measurements, but often the microphone and preamp end up being exposed to the RF impairment during the radiated immunity test. This introduces the possibility that any demodulated 1 KHz that is measured may come from demodulation within the microphone and preamp, rather than the equipment under test. In general, I think it is better to use Method 2 in Annex A, where the signal is measured as an electrical signal, rather than an acoustic signal. A lowpass filter can be used to keep conducted RF out of the measuring equipment. The use of the reference tone calibration procedure in Method 2 compensates for differences in the nominal gain between the phone line and the point where the signal is being measured. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. http://www.randolph-telecom.com --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. Visit our web site at: http://www.ewh.ieee.org/soc/emcs/pstc/ To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Michael Garretson:pstc_ad...@garretson.org Dave Healddavehe...@mediaone.net For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org All emc-pstc postings are archived and searchable on the web at: No longer online until our new server is brought online and the old messages are imported into the new server.
Re: Am1 to CISPR24 in OJ?
In a message dated 8/20/01, David Gelfand writes: > Having had much difficulty meeting these limits, could you briefly summarize > the changes in Am1? Do they apply also to PSTN lines also? Hi David: I'm out of the office this week, so I don't have the amendment handy. However, as I recall all of the changes were to the limits applied to conducted susceptibility on PSTN lines. The frequency bands were revised so the tables now look a bit different, but the overall effect was to relax the limits in several frequency bands. I still have no word from any other list members about when this amendment is likely to show up in the OJ. Joe Randolph --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. Visit our web site at: http://www.ewh.ieee.org/soc/emcs/pstc/ To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Michael Garretson:pstc_ad...@garretson.org Dave Healddavehe...@mediaone.net For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org All emc-pstc postings are archived and searchable on the web at: http://www.rcic.com/ click on "Virtual Conference Hall,"
Am1 to CISPR24 in OJ?
In a message dated 7/23/01, Ghery Pettitwrites: > You can purchase a copy of Amendment 1 to CISPR 24 from the IEC web page > (http://www.iec.ch) for all the details. The amendment primarily deals with > analog telephone sets. > Hi Ghery: Thanks for the info on this. I downloaded the amendment and was pleased to see that it mostly represents a relaxation of the limits. Do you or anyone else on the listserver have an idea of when this amendment is likely to be published in the Official Journal? My understanding is that this amendment has no regulatory significance until that occurs. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Immunity requirements outside europe?
In a message dated 6/29/01, David Gelfand writes: > What other countries besides CE have immunity requirements similar to EN > 55024 > or EN 50082? > Korea is one example. I think Singapore is phasing in some immunity requirements. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Conducted Emissions Test in Telephones
In a message dated 6/28/01, Geoff Lister writes: > EN55022:1998 section 9.5 indicates that measurements must > be made on telecommunication ports, and covers, in great detail > Hi Muriel: Geoff has provided a succinct answer to the question you posed. I just want to add that you will also have to test for immunity per EN 55024 if you plan to CE mark your product. This standard includes about seven different immunity tests (static discharge, lightning, etc.). Of the tests included in EN 55024, perhaps the most difficult is the conducted immunity test. For this test, common mode RF signals of 3 VRMS, 150 KHz to 80 MHz, are applied to the phone line. The RF carrier is 80% AM modulated at 1 KHz to simulate an AM radio station. Limits are placed on the amount of demodulated 1 KHz that appears on the phone line and in the handset. My experience with this new test suggests that most conventional telephone designs will have difficulty passing. You may need to add some special filtering in strategic locations order to pass. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Small form factor serial cables
In a message dated 6/22/01, Chris Maxwell writes: > I have a test sample that has one of these serial ports on it, and I can't > find a cable assembly to mate with it. I have called my usual cable > assembly vendors with no luck. I now throw myself at the collective wisdom > and mercy of the group. Any hints? > Hi Chris: A couple of years ago I saw a very tiny D-sub connector on a cell phone fax adapter. The designer told me the connector had been specifically developed for cell phones. If it turns out that that this is the connector you are asking about, there may be some cable manufacturers that serve the cell phone industry who could help you. Sorry, but this is all I can contribute on this topic without a lot more digging. The company that made that product has gone out of business, but I know where a couple of the engineers ended up. So, if this tidbit sounds promising and you get no other leads, let me know and I will track down the engineers to see if they can provide any further info. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Loop start fail over relays
In a message dated 5/29/01, Dan Roman writes: > I am looking for a relay that can be used without restriction for a fail > over function between CO trunks and station sets in a PBX type application. > The fail over relays switch select CO trunks over to individual phones > should the PBX fail or lose power. The coil to contact insulation must > protect SELV from TNV-3 and the contact to contact insulation must protect > TNV-2 (which is not isolated from SELV) from TNV-3. Hi Dan: Sounds like what you want is a relay that provides "supplementary insulation for a working voltage of 250 volts" between open contacts. I've started down that path before, but have usually found another solution before I had to actually source such a relay. Back in 1993 I looked at the Omron G5H2 relay, which provided supplementary insulation between open contacts. The internal design of this relay was an engineering marvel. The relay footprint was 12 mm by 27 mm, so it was larger than most telecom relays, and it was more expensive. I have heard that this relay may have been discontinued, but it was certainly a nice part if you had to have supplementary insulation between open contacts. In practice, I have found that the requirement for supplementary insulation between open contacts can usually be waived by exercising certain clauses in EN 60950. In particular, the exemptions involving a permanent connection to earth ground are very useful for some types of equipment, including PBXs. You might look into this approach some more before you go crazy trying to source a special relay. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: EN 55024 Annex A.1
In a message dated 5/24/01, David Gelfand writes: > When we inject as per IEC 61000-4-6, the 1 kHz tone is demodulated and is > VERY loud in the telephone earpiece! This happens no matter which i/o > cable we inject, even the power cord. Hi David: While it is not surprising for an unprotected SLIC port to have trouble with the 61000-4-6 test, I think it is abnormal for the problem to be equally pronounced when cables other than tip/ring are being tested. There may be a problem with your ground connections and/or test setup. Are you using CDNs or the clamp method? Do you have a good ground plane as specified in the standard? I recently purchased most of the equipment required to perform this test, and I have been experimenting in the lab to characterize various DAAs and SLICs. I have found that the results I get are somewhat dependent on the specifics of the test setup. The first thing I would suggest is that you review your test setup to be sure it fully complies with 61000-4-6. I have not tested the AM79R70, but I did run some quick tests on a circuit that used the Intersil HC55183, which is also a ringing SLIC. In that test, the SLIC was pretty quiet at carrier frequencies above 1 MHz, but there was some 1 KHz demodulation at lower frequencies. I have not had time to experiment with possible fixes. I do know of someone who is using the AM79R70 and claims that they have passed this test. Unfortunately, I do not know the details. By the way, don't feel too bad if you are having trouble with this test. It is a brutal test, and most people I know who have tested their existing designs have failed. Since passing this test becomes mandatory on July 1 for CE marking, I expect that there will be a lot of activity as the July 1 date approaches. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: TS002 (Australian Telecom)
In a message dated 4/24/01, Kevin Harris writes: > We just ran across a nifty piece of standards writing. In Australia in of > their telecom standards (TS002) there is the following safety statement in > it. > > "5.1.1.2 CE should not cause harm or damage to a Telecommunications Network > or Facility when CE is operated outside the range of operating voltage and > environmental conditions specified by the manufacturer" > Hi Kevin: I agree with you that this is a poorly worded requirement. However, it may help that this is not a mandatory requirement. In TS 002, the word "shall" is used for mandatory requirements, and the word "should" is used to describe performance that is desirable, but not mandatory. See clause 1.1 for the definitions of "shall" and "should." The word "shall" is used for the more specific fail-safe criteria in clause 5.1.1.1. These are somewhat easier to test for. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: GR 1089 Intrabuilding Surges (2)
In addition to the various mechanisms for intra-building lightning that have been postulated on this thread, there is one other that I have heard reference to. If lightning strikes the top of a building, the path to ground is often through the steel frame of the building. When this happens, cables routed with long runs inside the building can pick up significant energy via induction. The coupling factor is small, but the magnitude of the surge current can be quite large, so the net energy coupled into the cable can be a problem. Whether it is for this reason or the others cited previously, there are some valid mechanisms that can (theoretically) induce lightning surges on intrabuilding wiring. Even though the folks at Bellcore/Telcordia tend to be a bit too conservative sometimes, my guess is that their requirements for immunity to intrabuilding lightning surges are based at least in part on past experience and/or field studies. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Korean immunity requirements
Dear All: It is my understanding that South Korea has adopted a set of EMC immunity requirements that are similar, but not identical, to the ones that the EU imposes. I believe that the scope of the requirements includes ESD, radiated RF immunity, EFT, surge, and voltage drops. So far, I have only heard verbal references to these requirements, but I would like to get more definitive information. Can anyone in the group help me with the following questions? 1) What are the official standards (publishing entity and standard numbers) for EMC immunity? 2) What types of equipment fall within the scope of these requirements? (My particulat interest is telecom equipment, both TTE and central office). 3) Are the immunity requirements mandatory at this time or will they be phased in at a later date? 4) Is there a way to get English translations of the requirements and test procedures? 5) What test labs are certified to perform the tests? Thanks for any assistance you can provide. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Need a CDN-T2 for 61000-4-6 conducted immunity test
Dear Listmembers: I would like to purchase a CDN (coupling-decoupling network) for performing the conducted RF immunity test called out in EN 55024. The specific CDN that I am interested in is the "CDN-T2" for 2-wire unshielded twisted pair, as shown in Figure D.4 of 61000-4-6. Can any of you recommend some suppliers who sell this device? So far I have identified only one small company in California. When I called to get more information I had a most unpleasant experience, so I would prefer to spend my money elsewhere if possible. I also have a couple of questions that I'm hoping you can help me with: 1) What specific performance parameters should I should be looking for in this CDN? 2) I note that the ISN (impedance stabilization network) in Figure D.1 of EN 55022 has a similar schematic to the CDN-T2. Is it possible to use the CDN-T2 for the EN 55022 conducted emissions test as well? My primary interest is in being able to perform the conducted immunity test per EN 55024, but it might be nice to be able to perform the conducted emissions test per EN 55022 as well. Any suggestions for recommended vendors would be welcome. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: FCC Part 68 Continuing Compliance
In a message dated 1/17/01, Courtland Thomas write: > I have a question concerning coninuing compliance for Part 68. I believe > that we are required to submit units for testing every six months. If the > FCC audits and finds that the test reports aren't up to date, then there can > be problems. This testing costs around $1500.00 for our ITE products. This > becomes a very expensive process. Is there anything that can be done to > eliminate this cost. I would like to test the units in house and log the > results and use that info as my continuing compliance data. Any thoughts on > this? > Hi Courtland: To the best of my knowledge, there is no reason why you can't do the continuing compliance testing yourself. In fact, you can even do the initial testing yourself if you are equipped to do so. For the initial testing you would have to submit a written report to the FCC, but there is no filing requirement for continuing compliance. You simply have to keep the test data on file. If you have a lot of different products in production, the economics favor in-house testing. A few years ago one of my clients who had about 30 different products decided to do their continuing compliance testing in-house. They figured the savings of at least $60,000 per year would easily justify the cost of the test equipment and a part time technician. They purchased the FCC Part 68 test equipment from Compliance Design, but never actually used it. I guess they just got too busy with other priorities to set up the in-house capability. Recently they told me they would like to unload the equipment, so if you are interested I will put you in touch with them. I wouldn't necessarily recommend the Compliance Design equipment for a new purchase, but if the price is right on the used equipment, it might be an attractive option. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Transient Surge Suppressor
In a message dated 1/15/01, John Juhasz writes: > I am trying to find a second-source for a TVS (Transient Voltage Suppressor) > that I now use to meet the 61000-4-5 1.2/50us 1kV transient. > Hi John: To finalize the requirements on the surge suppressor you need, you will have to determine the short circuit current of the surge, the required standoff voltage of the TVS device, the required clamping voltage, and the allowable capacitance. With these parameters in mind, some vendors to consider are Teccor (http://www.teccor.com), ST Microelectronics (http://www.st.com), and Semtech (http://www.semtech.com). All of these vendors have some very nice surface mount "crowbar" type devices, some of which are in packages common to two or more vendors. For instance, the Teccor sidactor in the DO-214 package is similar to some ST Microelectronics parts. If the surge is not very high energy, you could also consider using surge suppressor zener diodes such as the devices in the SMA, SMB, and SMC packages that many manufacturers of discrete components make. Some of these can handle up to 1500 watts for one millisecond. Motorola (now On Semiconductor, http://onsemi.com), General Semiconductor (http://www.gensemi.com), and Central Semiconductor (http://www.centralsemi.com) all make these devices. Examples from the Motorola family of "600 watt" TVS diodes are the 1SMB series and the P6SMB series. For high energy surges, I prefer the crowbar type devices such as teh Teccor sidactor. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: TNV-1 in Norway
In a message dated 1/2/01, Gaétan Hogue writes: > I have an TNV-1 interface (ISDN) on a card going in a computer. > Some of my designers told me that we need reinforced isolation > transformer(3kv) in Norway. > Hi Gaétan: The main requirement for reinforced insulation is between the TNV-1 circuit and the AC power mains. This is usually taken care of by the insulation in your power supply. Your designers may be referring to the requirement for supplementary insulation between TNV-1 circuits and earth ground. In the main text of EN 60950, only a hipot test is required (clause 6.4.1), but Annex ZB in EN 60950 calls out additional requirements for Norway and Sweden. For Norway, you will need supplementary insulation unless you qualify for the exemptions related to permanent grounding or a Type B mains plug. If you decide to implement supplementary insulation, keep in mind that in addition to a 1500 VRMS hipot test, there are specific requirements for creepage distance, clearance distance, and distance through insulation. This will affect your board layout, your choice of transformer, and (sometimes) your choice of connector. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Surge Protection
In a message dated 12/27/00, Zohar Zosmanovich writes: > I'm concern about designing a proper protection from Surges, applies on > ports for indoor signal lines (E1, Ethernet) at telecommunication centers. > According to ETSI EN 300 386 a surge of 1.2/50 us, 0.5 kV shell be applied > between Tip & Ring. Because of lack in space on the PWB we can't put the > overvoltage protector (sidactor, TVS) isolation transformer and ... close to > the connector were we applying the surge. My questions are: > > 1. Can I locate only the Sidactor close as possible to the connector, while > the isolation transformer is about 45 cm ahead ? > 2. Can the trace width after the sidactor, till the transformer, be thin as > 5 mills? > Hi Zohar: It's not clear to me that you need a sidactor to pass this test. The EN 300 386 surge is applied common mode (tip/ring to ground), so there should not be much in the way of a differential voltage between tip and ring. Assuming your interfaces are transformer coupled, and if the transformers have at least 500 volt isolation, the common mode surge should not damage anything since there is no path to ground. This assumes that your board layout can handle 500 volts from tip/ring to ground. It should not be hard to accomplish this, especially if you use internal layers. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: TNV to Earth Insulation Requirement
In a message dated 12/21/00, you write: > He also said that meeting the 500 V ac or 707 V dc requirement was not > Hi David: Thanks for the feedback. Can you elaborate on the reference above to the "500 VAC or 707 CDC requirement"? What clause does this requirement derive from? Was your CSA rep saying that this requirement applies even with a permanent ground, or was he saying that this requirement applies if you do not have a permanent ground? Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: TNV to Earth Insulation Requirement
In a message dated 12/12/00, Peter Merguerian writes: > Assume you have equipment with external power earthed (Class 1) power > supply with an earthed appliance inlet for conenction to the mains. The > power supply output is connected to a dc unit with a TNV-2 interface with > no separation to SELV. The dc unit has a earth stud for connection of an > external earthing conductor. All exemptions for basic insulation in Table > 19 is applied. > > My questions is: can the stud on the dc unit serve as my permanent > Hi Peter: Based on your description, my tentative answer would be yes. I'm not sure what the "DC unit" that you refer to is, but the crux of your question seems to be whether it is OK to have a separate grounding lug on the chassis of the equipment that mixes TNV-2 and SELV/ground, as opposed to requiring a hard wired ground associated with the AC power input. The use of the separate grounding lug is quite common on pluggable Type A equipment that needs to qualify for the exemptions under Table 19. I believe that the requirements for the grounding lug are contained in clause 2.5. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: TNV to Earth Insulation Requirement
In a message dated 12/11/00, Dave Lorusso writes: > If a telecom product is permanently connected (-48 Vdc Central Office > Equipment), is it exempt from paragraph 6.3.3 Separation of the > telecommunication network from earth, of UL 1950? I'm reading this as there > is no physical spacing required between TNV and Earth for permanently > connected equipment. How about impulse/hipot requirements? > > Is there anything else in UL 1950, IEC 60950, EN 60950, that contradicts > this? > Hi Dave: This topic seems to generate a lot of confusion and controversy, since most telecom equipment that the test labs see is subject to the isolation requirements between TNV circuits and SELV/ground. Many test labs seem to be unaware that the isolation requirements can be waived if certain conditions are met. However, UL 1950 does have provisions for certain types of equipment to not have this isolation. For example, central office equipment and large PBXs typically do not have isolation between TNV and SELV/ground. The requirements for isolation appear in several clauses in UL 1950 (6.2.1.2, 6.2.1.4, 6.2.1.5, 6.2.2.1, 6.3.3.1, 6.3.3.2, and 6.4.1). To qualify for the exemptions, it is necessary to step through each clause and show compliance with the conditions for which the exemptions are allowed. In general, though, equipment which has a permanently connected earth ground can usually be exempted from the isolation requirements. For clause 6.3.3 that you mentioned, the permanent ground exemption can be found in clause 6.3.3.2. If your equipment qualifies for the exemptions from having isolation, there should be no requirement for a hipot test between your TNV circuits and SELV/ground circuits. If you plan to exercise the exemptions, I suggest that you review your rationale ahead of time with your test lab, before they try to hipot your equipment. The requirements in EN 60950 are pretty much the same as UL 1950, except that Norway and Sweden have additional requirements for the exemption that can be found in Annex ZB. These requirements call for permanent connection or pluggable Type B connection for the AC mains. This may or may not affect your product if the input is strictly -48 VDC. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com
Re: Telecom/Safety Question
In a message dated 11/30/00, Jim Lyons writes: > The gas tube would look pretty much like a short circuit once it starts to > conduct, and I would expect the Buss MDL2 fuse they used to blow. Their > failure criteria would seem to exclude the use of gas tubes directly > across the tip and ring lines. Is this a proper assumption? Hi Jim: In general, you will need to have fuses or PTC devices in front of your gas tube. To pass the M1 test, your equipment must become an open circuit before the fuse in the test set opens. Since you have protection from tip to ground and ring to ground, you will need a fuse in tip and a fuse in ring. Otherwise you will fail the L-1 test. I generally use a 1.25 amp time lag fuse such as the Teccor F1250T (surface mount) or the Bussmann C515 (through hole). I also prefer Teccor sidactors over gas tubes, since they are faster and more precise. Some people use the Raychem TR600 Polyswitch instead of a fuse, but a well chosen conventional fuse will be a little more robust for lightning immunity. Whatever protection scheme you use, you need to consider how your circuit will respond to the M3 and M4 tests. These tests can force quite a bit of current through the circuit and overheat it, since they are designed to "sneak in" below the trigger thresholds of your fuses and overvoltage protection. Lastly, I should mention that there are some ways to avoid putting in fuses, but they are not often used. You will have to specify the use of a 26 gage or heavier line cord, and you may need to have a fire enclosure. Take a look at Figure 18b in UL 1950 to see a flow chart of possible compliance paths. This flow chart is not easy for the uninitiated to follow, so don't feel bad if it seems confusing. Your UL test lab should be able to walk you through it. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 http://www.randolph-telecom.com --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
Re:
In a message dated 9/26/00, John Juhasz writes: > Personally I see this as a Medical Device with 'Telecommunications Feature'. > Therefore, I envision the application of the applicable Medical Device > standard with a separate test (report addendum) on the telecom port utilizing an > appropriate telecom standard. Peter, John: I agree with John's interpretation above. A couple of years ago I worked on a portable defibrillator/EKG device that had a modem in it for sending data back to the hospital. We used the relevant requirements for a medical product, but did supplementary testing for the telecom interface using EN 60950 and UL 1950. For Europe, you might even consider EN 41003 (if it is still around), since that was intended to cover the generic case of connection to the telecom network. For most products the ITE safety standard EN 60950 has superseded EN 41003, but for a medical product EN 41003 might be a good match. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. http://www.randolph-telecom.com --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
Re: Ringing wave surge testing for Modem
In a message dated 8/31/00, Matt Aschenberg writes: > The fast transient has a ringing (underdamped) wave form. Our primary > concern is to protect the modem against lightening surges. We currently test > to 6kV underdamped and overdamped. Hi Matt: If your primary concern is lightning immunity, you should probably refer to some industry standards that address this specific topic. In most cases the peak voltage of the surge is limited to less than 1500 volts by the primary protector located where the cable enters the building. However, the peak currents can reach 100 amps, and the total durations can reach 1000 uS. In other words, there is still a lot of energy to deal with. You do not mention the equipment you are using to generate the surges, but if it was not designed specifically for simulating lightning surges, it probably does not generate enough energy to accurately simulate lightning. Two references that describe recommended immunity tests for lightning are ITU-T K.21 and Bellcore/Telcordia GR 1089-CORE. When you develop your lightning protection scheme for the modem, keep in mind that the same circuit must also meet the power cross requirements in UL 1950 (if the product will be used in the North America). You will find that the UL 1950 compliance and lightning immunity are somewhat at odds with each other, so it takes a well thought out design to perform well for both tests. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. http://www.randolph-telecom.com --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
Re: EU Requirements for PSTN Equipment
In a message dated 7/28/00, Allan Carr writes: > COMMISSION DECISION of 26 May 2000 > concerning the request by France to maintain pursuant to Article 18(3)of > Directive 1999/5/EC of the European Parliament and of the Council (the > "R&TTE Directive") a requirement for telecommunications terminal > equipment intended for connection to the analogue public switched > telephone network of France Telecom (2000/373/EC) > > This requires limiting of the line current for TTE used in France. > > There are no other Harmonized PSTN Standards so you can self-declare > that you comply with all the applicable specifications under the R&TTE > Directive (which is NONE.!) unless you are selling into France. Hi Allan: I have a slightly different interpretation of the Commission Decision that you reference. My interpretation is that the Commission specifically declined to require current limiting. Instead, they said that manufacturers must provide information to users about the intended use of the TE. TE without current limiting must contain a warning that it should not be used on lines that require current limiting. To me, this sounds like a political solution, where the Commission tried to appear sympathetic to France Telecom's request, but basically refused to impose current limiting for France. Thus, it appears to me that it is up to the manufacturer to decide whether to implement current limiting for France. I would not necessarily recommend that a manufacturer leave out current limiting, but it does appear to me that they have this option. There are some situations where exercising this option might make sense. For instance, if the manufacturer has an existing design that is otherwise compliant with EU requirements, and service personnel would be installing the equipment in only a few known locations, it might make sense for the manufacturer to confirm that the local phone lines do not require current limiting instead of undertaking a costly redesign to add current limiting. In France Telecom's filing with the Commission, they stated that 22% of their existing lines require current limiting, but that steps are underway to upgrade these lines. They also stated that this process would be completed within the maximum 30 month delay that the RTTE directive allows member states to request. My guess is that with the response that they got from the Commission, France Telecom may feel compelled to accelerate their retrofit program. Do you agree that manufacturers have the option to omit current limiting (with appropriate warnings), or have I misunderstood the Commission decision? Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. http://www.randolph-telecom.com --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
Dialers
Hi Peter: Back on July 11, you posted a question asking about the safety and EMC requirements for dialers. I meant to respond, but got busy and forgot. Did you get the info you needed? I am not sure what you mean by "dialer," but in my mind this usually refers to a device that is placed in the phone line in series with a TE such as a fax machine or phone. The dialer intercepts the DTMF digits dialed by the TE and, for certain dialed numbers, redirects the call to a different number. Typically, the dialer is co-located with the TE, although sometimes it is placed in the telephone closet with the PBX. Is this the type of device you are asking about? I have designed a few such devices in the past, and have treated them as ITE devices that are subject to the EMC Directive and the Low Voltage Directive. This would suggest EN 55022, EN 55024, EN 60950, etc. A key issue that is somewhat unique to dialers relates to the safety isolation. Typically there is a mix of TNV-3, TNV-2, and SELV circuits inside the dialer, so that providing the required safety isolation in all operational modes can become a problem. There are a couple of ways to address this issue, depending on what other types of interfaces appear on the dialer. Aside from the safety isolation issues, the remaining compliance problems are similar to other types of ITE. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. http://www.randolph-telecom.com --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
Re: Breakdown voltage between pcb layers
In a message dated 5/23/00, David Gelfand writes: > We have an emissions problem on a board and I would like to suggest a ground > plane in the area of an RJ-45 jack (TNV-1). But we have always asked our PCB > designers to leave TNV traces free of ground and power planes to avoid > arcing during surge and dialectric strength tests. > > Does anyone know where to find specs on breakdown voltages between PCB layers? > Has anyone successfully used ground planes above or below TNV traces? We > are testing to UL1950 and Part 68. David: I do not think it is a good idea to put ground plane under your TNV-1 circuits. It is possible to do this, but I do not recommend it. You do not mention the type of EMC problem you are trying to resolve, but if it is an emissions problem, the first choice would be to add a common mode choke in series with tip/ring. If that alone is not sufficient, you can sometimes add small value caps (100 pF or so) from tip to chassis and from ring to chassis. Since these capacitors bridge the isolation barrier, they must be suitable high voltage caps. Capacitors in these locations can be very helpful if you have a "quiet" metal chassis to connect them to, but they are useless and even potentially harmful for EMC if the product has only a plastic housing with no real chassis. When adding capacitors, you must also watch out for detrimental effects on the intended signal. If you are still intent on adding a ground plane under your TNV-1 circuit, the main requirement for both FCC Part 68 and UL 1950 will be to pass a 1000 VRMS hipot test. You can do this by carefully specifying the required insulation between the relevant layers, but now your board stack-up will be subject to special requirements that both you and your board vendor must keep track of. Another potential problem is that TNV vias which pass through the ground plane must have a large enough clearance hole in the ground plane to avoid hipot failures at that point. There is one special case where the requirements of the above paragraph are easy to meet. This would be when the TNV circuits are on the top layer and the ground plane is on the bottom layer, with no copper on any of the internal layers. In this case, the entire thickness of the board is insulation, which is more than adequate. However, as soon as you start using the inner layers, you must get involved with specifying the layer-to-layer dielectric strength. While the overall thickness of the board is always specified, individual board fabricators typically exercise considerable freedom in selecting the thickness of the insulation between inner layers. In summary, I do not think it is a good idea to extend your ground plane under the TNV circuits, but it is technically possible to do so by carefully specifying the circuit board stack-up. Before you resort to this solution, I think you should carefully examine the reasons why this seems to be necessary, and consider alternative solutions. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. http://www.randolph-telecom.com --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
Re: EMC and product safety split?
In a message dated 3/10/00, Steve Phillips writes: > EMC gets a lot more traffic (if appears > to me anyway) - and I am more concerned with > safety, which leaves me filtering and deleting a > lot more messages than I would like to have to - > it wouldn't be so bad if EMC and Safety messages > were separated - perhaps many people feel the > same. My perspective is the same as Steve's. I would prefer to have the two lists separated, because my interest in this list is more area of safety than EMC. I currently subscribe to the treg listserver for telecom issues, and to the emc-pstc listserver for safety issues that relate to telecom. For me, EMC is a distant 3rd in terms of my particular interest level. It would be nice if people had a listserver for each topic, so they could be more selective about the amount of email traffic they sign up for. Of course, current subscribers who are interested in both safety and EMC could simply subscribe to both lists. I think the main issue is that splitting the emc-pstc list may require having two list administrators, which could present a resource problem. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
Re: Source for leadless high-voltage capacitors?
In a message dated 3/9/00, Scott Lacey writes: > I am looking for a source for leadless disc capacitors. I require 10,000 pF > of at least 3kV. Basically, what I'm looking for is to buy capacitors that > have not yet had the lead attachment and coating operations performed. I am > trying to eliminate the lead inductance. Scott: Could you just use a high voltage surface-mount cap instead? Surface mount capacitors meeting your description are available from several vendors such as Novacap, Murata-Erie, Johanson Dielectrics, and AVX. Joe Randolph --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
power line crosses / induction issue
In a message dated 1/21/00, John Boucher writes: > What I'm really wrestling with is whether we need to be concerned about power > line cross / induction issues outside of North America. > John: I too have heard the argument that AC mains wiring outside of North America is "different," and that as a result, power cross is less of a risk outside North America. I'm not sure that I believe this. Even in North America, power cross and induction at levels that present an actual safety risk are extremely rare events. My impression is that the people developing safety standards for North America were perhaps the first to identify power cross as a possible risk, and the first to include power cross requirements in their standards. Even so, there was considerable internal debate about whether power cross and induction presented enough of a safety risk to justify inclusion in the safety standards. I have heard that at the IEC level, many member countries have been skeptical about the actual severity of the power cross risk, and have been reluctant to include power cross requirements in IEC standards such as IEC 950. Some observers think this may change in the future. I would note that the 1996 version of the ITU K.20 requirements now contains several power cross tests. In addition, the new EN 300 386 (for telecom network equipment) includes several power cross tests. This indicates that there is now at least some acceptance at the international level of the risk of power cross and induction. My suggestion would be to use your North American line interface design (including UL 1950 power cross compliance) for your international applications as well. This will eliminate any doubt about whether you have chosen the proper path for international applications. A properly executed line interface design will add very little cost to the interface, so the cost penalty for including power cross protection in your international version should be small (I have seen some designs that were quite expensive, but they don't have to be. There are several ways to comply with the requirements, and some designs are much better than others). An added benefit of using the same line interface in all countries is that it increases your unit volumes for components, and it reduces the number of different line cards that must be manufactured and supported. These savings can help to offset the small added cost for power cross protection in the international applications. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 - This message is coming from the emc-pstc discussion list. To cancel your subscription, send mail to majord...@ieee.org with the single line: "unsubscribe emc-pstc" (without the quotes). For help, send mail to ed.pr...@cubic.com, jim_bac...@monarch.com, ri...@sdd.hp.com, or roger.volgst...@compaq.com (the list administrators).
Using RTTE directive before April 2000?
Listmembers: I have a question that perhaps some of you can help me with. I'm developing a regulatory compliance plan for a new telecom product that is scheduled to begin shipping in the first quarter of 2000. The exact date is not certain, but it is likely to be before the April 8, 2000 date that appears in the RTTE directive. If possible, I would like to avoid the whole notified body route called out by the current directive 98/13/EC, especially since it would only be required for the brief period until April 2000. I seem to recall that a new directive can be used as soon as *any* member state has transposed it into national law. If so, this suggests that the RTTE directive could be used prior to April 2000 if at least one member state has transposed it into national law. In the case of the UK, however, recent postings on the emc-pstc listserver indicate that the draft legislation for the UK calls out an effective date of April 8, 2000. In other words, even if the UK transposes the directive prior to April 2000, the national law itself will call out an effective date of April 8. I do not know what the other member states are planning to do. So, am I stuck with using directive 98/13/EC and the notified body route if the product ships prior to April 8, 2000? Joe Randolph Telecom Design Consultant Randolph Telecom, Inc.
Re: UL1459
In a message dated 12/3/99, Bruce Benzie writes: > UL was talking about integrating 1459 (telephone equip.) into UL1950 (ITE) > can anyone update me as to status? Bruce: UL 1459 is being phased out, so you should use UL 1950 for new designs. Most of the telecom-specific aspects of UL 1459 (such as power cross testing) have been incorporated into the Third Edition of UL 1950. Joe Randolph Telecom Design Consultant Randolph telecom, Inc. - This message is coming from the emc-pstc discussion list. To cancel your subscription, send mail to majord...@ieee.org with the single line: "unsubscribe emc-pstc" (without the quotes). For help, send mail to ed.pr...@cubic.com, jim_bac...@monarch.com, ri...@sdd.hp.com, or roger.volgst...@compaq.com (the list administrators).
Intrinsic safety for factory communication?
Hello All: I am hoping that some of you can help me gain an understanding of the requirements for "intrinsic safety" for communication wiring that is used in a manufacturing environment. My background is in telecom and ITE, not factory automation. However, I have been asked to assist a client with developing some communication technology that will be used for sensors and controls on a factory floor. My client tells me that for communication wiring that is used in a factory, there are requirements for "intrinsic safety" of the wiring. Among other things, there is reportedly a test where connections are made and broken in an atmosphere that contains explosive gases. Since any sparks generated from the make/break of the connections could ignite the gases, it is reportedly necessary to design the communication scheme in such a way that no sparks are generated. When I asked where these requirements are documented, I was told that there is no written standard for intrinsic safety. Rather, the system must be submitted to an independent agency for review, and this agency will evaluate the system according to their own (internal) criteria. Now, coming from a background in ITE, I find this hard to believe. In the ITE world, we have documented standards for safety such as UL 1950 in the USA and EN 60950 in Europe. I find it hard to imagine that there are no comparable standards for factory automation. Can any of you clarify this issue for me? Where does the concept of "intrinsic safety" come from, and how is compliance determined? I am interested in addressing this issue for both North America and Europe. Any assistance you could provide would be greatly appreciated. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2840 - This message is coming from the emc-pstc discussion list. To cancel your subscription, send mail to majord...@ieee.org with the single line: "unsubscribe emc-pstc" (without the quotes). For help, send mail to ed.pr...@cubic.com, jim_bac...@monarch.com, ri...@sdd.hp.com, or roger.volgst...@compaq.com (the list administrators).
Re: connectors that meet hipot requirements
In a message dated 7/7/99, Susan Beard writes: << Anyone know of any connectors with 0.1" spacings that survive a 2000 volt rms hipot requirement? >> Susan: Don't forget that for most connectors, the minimum spacing actually occurs on the printed circuit board, at the land areas around the pins. For a through-hole connector that has .100" pin spacings and .060" round pads, the spacing on the board will be .040". Several times I have seen people fail to take this into account. For through-hole pins, you can increase the separation by using an oval pad. Surface mount connectors often allow greater spacing for the same pin pitch. Another way to provide the required separation is to use a connector that has a smaller pitch, but allocate one or more "dead" pins (not connected to anything) between the pins you are trying to separate. In this case, the effective separation will be the total of all the smaller separations between pins. Sometimes this approach allows you to use a connector that is otherwise more desirable than one selected specifically for its pin separation. In practice, a .040" spacing might pass your 2000 VRMS test, but I would recommend a minimum of .050". Anything above this amount would increase your margin of safety. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 (voice) 781-721-0582 (fax) - This message is coming from the emc-pstc discussion list. To cancel your subscription, send mail to majord...@ieee.org with the single line: "unsubscribe emc-pstc" (without the quotes). For help, send mail to ed.pr...@cubic.com, jim_bac...@monarch.com, ri...@sdd.hp.com, or roger.volgst...@compaq.com (the list administrators).
Re: Hipot requirements for TNV circuits
In a message dated 7/1/99, Peter Merguerian writes: > By the way, a transformer is not always needed on a modem. The > new design of modems have capacitors (not required to be > Approved) between the TNV-3 and SELV. Hi Peter: It is certainly true that some of the new modem DAA devices use capacitors instead of a transformer (Silicon Labs, Krypton, Conexant, Siemens, Analog Devices, etc). However, I do not agree that the capacitors somehow do not have to be approved. On what basis have you reached this conclusion? There certainly does seem to a debate about what requirements might apply to capacitors used in this application. For basic insulation, some people have argued that as long as the physical size of the cap meets the creepage and clearance, and the cap passes the hipot test, you are all set. My impression is that some safety labs accept this interpretation, while others do not. Sometimes the design goal is to provide supplementary insulation, to comply with requirements in Norway and Sweden. In this case, the isolation requirement includes a minimum distance throgh solid insulation of 0.4 mm. I believe that most multilayer ceramic caps violate this requirement. It is still possible to use such a capacitor by using an approved Y2 cap (per IEC 384-14), but these caps tend to be physically large and somewhat expensive in surface mount. I am aware of a couple of vendors who are trying to get smaller caps approved for bridging supplementary insulation, but I have not heard yet that these efforts have been successful. Do you have more information on this subject? Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 (voice) 781-721-0582 (fax) - This message is coming from the emc-pstc discussion list. To cancel your subscription, send mail to majord...@ieee.org with the single line: "unsubscribe emc-pstc" (without the quotes). For help, send mail to ed.pr...@cubic.com, jim_bac...@monarch.com, ri...@sdd.hp.com, or roger.volgst...@compaq.com (the list administrators).
Re: Hipot requirements 4 TNV circuits
In a message dated 7/1/99, George Sparacino writes: > I have a modem product that failed the hipot requirement defined in > cls 6 of 950. My understanding is that an isolation xfmr between Telco & > SELV is needed and clearance from Telco & gnd is to be 3mm minimum. Does > anyone have any other general design guidelines that they can share on this > topic ? George: The isolation requirements that apply to your modem will differ based upon several variables, including the types of cables attached, how the SELV circuits are grounded, and where the modem is installed. However, in the vast majority of cases (such as a typical box modem) IEC 950 requires "basic" insulation between the TNV3 circuits and the SELV circuits, per clause 6.2.1.2. This translates to 1.0 mm clearance, 1.6 mm creepage, and 1000 VRMS hipot. In my experience, a circuit that provides the required creepage and clearance distances will easily pass the hipot test. You referenced the "950" spec, but did not mention whether you are using UL 1950 or EN 60950. If you are using the second edition of UL 1950, there are some provisions that allow you waive the creepage and clearance requirements and use a simple 1000 VRMS hipot test. The second edition will be withdrawn in April 2000 and replaced with the third edition, which does not allow this option. If your target market includes Europe, you should be aware that Norway and Sweden have taken exception to the requirement for basic insulation that appears in EN 60950, and have inserted national deviations that require supplementary insulation. This requires 2.0 mm clearance, 2.5 mm creepage, and 1500 VRMS hipot. For supplementary insulation, there is an additional requirement for at least 0.4 mm distance through any solid insulation that is used in the barrier. Basic insulation does not have this requirement. If you are looking for a manufacturer's app note on this subject, you might check Midcom at http://www.midcom-inc.com. I seem to recall that they have such an app note. Also, there was a lengthy discussion on this topic about a year ago on the treg listserver. You can search the archives of postings at http://www.rcic.com. Lastly, if you are still confused about this subject, give me a call. If you can describe the specifics of your application, I can quickly identify the applicable clauses in IEC 950 and walk you through the calculation of the requirements for your specific application. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 (voice) 781-721-0582 (fax) - This message is coming from the emc-pstc discussion list. To cancel your subscription, send mail to majord...@ieee.org with the single line: "unsubscribe emc-pstc" (without the quotes). For help, send mail to ed.pr...@cubic.com, jim_bac...@monarch.com, ri...@sdd.hp.com, or roger.volgst...@compaq.com (the list administrators).
Re: Network Equipment and UL 1459/1950
In a message dated 3/24/99, jim.wi...@adtran.com writes: > The only design criteria with regard to insulation in Bellcore standards or UL 1459 is a hi-pots test. > Creepage and clearance do not exist in traditional C.O. equipment (just look at wire wrapped > backplanes). Dear Jim and others: Jim Wiese has raised several concerns regarding the transition from UL 1459 to UL 1950 (3rd Edition). One of these concerns relates to the new creepage and clearance requirements for separation of TNV circuits and ground/SELV. I would like to pick up on this one issue (I know that Jim has raised several other issues as well). In particular, I would like some feedback from others in the group regarding my interpretation of how the separation requirements in UL 1950 apply to certain types of equipment. Just this week I met with a PBX manufacturer who is in the process of re- designing the backplane and all of the line cards in their PBX to "comply with UL 1950." The PBX is presently approved to UL 1459, but the manufacturer has its eye on the March 2000 date for new or modified products to comply with UL 1950. Needless to say, the redesign effort is an expensive one. The biggest headache in the redesign is complying with the creepage and clearance distances for separation of TNV and SELV circuits. The PBX manufacturer seemed incredulous when I stated that I did not think the creepage and clearance requirements applied to their product, since the PBX has a permanent (hardwired) connection to ground. My interpretation is based primarily on the following statement in clause 6.2.1.2 in UL 1950, paraphrased below: " Basic insulation is not required provided that all of the following conditions are met: - the SELV circuit is connected to protective earth...in accordance with 2.5; and - the installation instructions specifya permanent connection to earth; and - the test of 6.2.1.3 is carried out... (where applicable)" There are other clauses that call out isolation, such as 6.3.3.1 and 6.4.1, but the "permanent ground" exemption appears to apply here as well. In my view, these exemptions are specifically targeted at equipment such as PBXs and network equipment that are typically installed by service personnel and include hardwired grounding. Without these exemptions, it is almost impossible to separate certain types of TNV circuits from SELV and ground. For example, a feed circuit that provides 48V battery (SLIC, FXS, DID, etc.) is inherently referenced to ground. The situation with a ground-start FXO interface is not much better. Do others in the group agree with this interpretation? If not, how are feed circuits supposed to be isolated? Does anyone have direct experience with getting a product through UL with these exemptions? I recognize that the original thread here related to network equipment, but the "permanent ground" exemption should apply to network equipment as well. I also recognize that these exemptions only apply to circuits that qualify as TNV, and do not address some of the other issues that Jim Weise raised concerning things like 200 volts DC for repeaters. However, for simple TNV isolation, it seems that network equipment could use the "permanent ground" exemptions from having to provide creepage and clearance (or in fact, any isolation at all). Any input from others in the group would be welcome. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 781-721-2848 (USA) - This message is coming from the emc-pstc discussion list. To cancel your subscription, send mail to majord...@ieee.org with the single line: "unsubscribe emc-pstc" (without the quotes). For help, send mail to ed.pr...@cubic.com, j...@gwmail.monarch.com, ri...@sdd.hp.com, or roger.volgst...@compaq.com (the list administrators).
Re: Crepage/Clearance On Telecom Modem design
In a message dated 98-10-12, Mel Pederson writes: << If one considers the worst case (DC) voltages that could possibly exist on TNV-3 lines and also SELV circuits, and considers the likelihood that these are at opposite polarity, one would obtain a worst case 180VDC = -120VDC (TNV-3) - 60VDC (SELV). >> Mel: I have a vague recollection that somewhere in EN 60950, it states that for purposes of determining working voltage, SELV circuits can be assumed to be zero volts. Perhaps this explains the slight difference in your calculations of working voltage from the calculations of the labs you reference. Joe Randolph - This message is coming from the emc-pstc discussion list. To cancel your subscription, send mail to majord...@ieee.org with the single line: "unsubscribe emc-pstc" (without the quotes). For help, send mail to ed.pr...@cubic.com, j...@gwmail.monarch.com, ri...@sdd.hp.com, or roger.volgst...@compaq.com (the list administrators).
CE marking vs. Annex ZB in EN60950
I am trying to sort out an an apparent contradiction between the main body of the "harmonized" EN 60950 and the national deviations that appear in Annex ZB. I would appreciate some assistance from the safety professionals and regulatory experts in the emc-pstc and treg forums. My particular problem has to do with the specified insulation between between TNV-3 circuits and SELV circuits, although the question about Annex ZB applies to other national deviations as well. I will use the TNV/SELV issue as an example. Clause 6.2.1.2 of EN 60950 specifies basic insulation between TNV-3 and SELV circuits. However, in Annex ZB, it states that Norway and Sweden require supplementary insulation. Annex ZB states that "for the countries in which the special national conditions apply, these provisions are normative. For other countries they are informative." I am trying to understand how this distinction is applied for CE marking. I thought that the intent of CE marking was that products could be evaluated against "harmonized" standards and then have the CE marking applied. I thought that products with the CE marking were allowed to move freely among the member countries in the EU. Annex ZB seems to complicate this simple interpretation. My tentative conclusions about the above situation are: 1) A manufacturer who intends to market his product in a list of EU countries that excludes Norway and Sweden can apply the CE marking to a product that has only basic insulation between TNV-3 and SELV circuits. 2) A manufacturer who intends to market his product in a list of EU countries that includes Norway or Sweden can only apply the CE mark if the product provides supplementary insulation between TNV-3 and SELV circuits. The above two cases result in products that each bear identical CE marking, but one of them (case 1) is technically illegal for use in Norway and Sweden. >From looking at the product marking, there is no way to determine that the product in case 1 should not be used in Norway or Sweden. I suppose that this distinction could be made in the user manual. I also presume that the manufacturer in case 1 would be obligated to ensure that the product is not offered for sale in Norway or Sweden. Maybe this is enough to comply with the letter of the law, but I suspect that it would be easy for the product in case 1 to find its way into Norway and Sweden. Prior to the adoption of CTR 21 for analog modems, the above issue was largely academic, since individual national approvals and the corresponding national labels were still required for connection to the PSTN. However, now that CTR 21 has been adopted, I expect there will be a rush toward "full" CE marking (including the crossed hockey sticks) of analog modems. National approval labels will no longer be required. I think this will bring additional attention to the situation I have described above. Can anyone offer some some insight into how this situation should be handled? Is my interpretation correct? Any comments would be welcome. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc. 325 Highland Avenue Winchester, MA, USA 01890 781-721-2848 (voice) 781-721-0582 (fax)
Re: Basic insulation
Judd Stewart writes: << We are currently designing an ac filter assembly. The board is extremely > small and we need to put text in etch (copper) on board. About the only > place is an area that provides basic insulation from ground. >> Judd: Would it be possible to provide your text in a (non conductive) silkscreen, instead of in copper? If your board uses silkscreen to show component outlines and reference designations, you can add the text to the silkscreen without compromising the insulation barrier. Joe Randolph
emc-pstc@listserv.ieee.org
In a posting dated 98-02-18, Rich Nute writes: << I am assuming that the capacitors are connected between mains and secondary circuits. (If they are connected between a high-voltage secondary and a low-voltage secondary, then this discussion does not apply.) Such capacitors would be Y capacitors in the IEC scheme, or UL-Recognizied capacitors in the UL scheme. >> Rich: I am interested in your parenthetical remark about the case where the capacitors are bridging the barrier between two secondary circuits. I have been having trouble getting a definitive answer regarding a special case of this situation. I would be interested to hear your thoughts on this. For telecom applications, EN 60950 requires basic insulation between a TNV-3 circuit (regular phone line) and a SELV circuit, per clause 6.2.1.2. A TNV-3 circuit is, by definition, a secondary circuit (clause 1.2.8.8). My question is, what requirements apply to capacitors which bridge the basic insulation barrier between TNV-3 and SELV? Two interpretations have been presented to me: 1) The capacitors simply have to meet the required dielectric strength test and provide the required creepage/clearance between their terminals. 2) The capacitors must be safety rated Y2 caps. Initially, my interpretation was #1 above. I believe that UL has a similar view. However, I am told that BABT has adopted interpretation #2. Reportedly, they justify this by pointing to EN 60950 clause 1.5.1, which says that "where safety is involved, components shall comply with the requirements for this standard or the safety aspects of the relevant IEC component standard." This leads them to require that the capacitors comply with IEC 384-14. However, my view is that IEC 384-14 is not relevant to this case, because IEC 384-14 is directed at capacitors which bridge the barrier from a *primary* circuit. Note 1 in clause 1.5.1 of EN 60950 states that "an IEC component standard is considered relevant only if the component in question clearly falls within its scope." I do not think that capacitors which bridge two secondary circuits fall within the scope of IEC 384-14. The difference between interpretations 1 and 2 has great significance for designers of phone line interfaces. In many of today's space-constrained products (such as PCMCIA card modems), it is often difficult or impossible to use Y2 caps to bridge the barrier. The use of interpretation 2 effectively eliminates certain design approaches that might otherwise be considered. Can you comment on this issue, and describe your interpretation of what requirements apply? I would also be interested to hear from any other emc- pstc members who may have comments on this subject. Thanks for any insight you can provide. Joe Randolph Randolph Telecom, Inc.
Re: Varistors Allowed - Line to Ground? for European App's
In a message dated 97-08-14 02:21:07 EDT, Art Michael writes: >Have heard from some sources that surge-suppressors (Varistors, MOVs, >Tranzorbs, etc) are not allowed between Line to Ground, in EN 60950 >products Art: A few years ago, I saw a letter on VDE or TUV letterhead that stated that MOVs would not be allowed connected line-to-ground. The concern was the degradation over time that MOVs can exhibit, leading to high leakage currents and possible overheating. The letter made no mention of other types of protectors, such as sidactors or transzorbs. Let me know if you want me to dig out this letter. I'm away from my office right now, but I think I have the letter on file. Joe Randolph
Re: Blue Angel (is Class B required?)
In a message dated 97-03-19, Jerry Roberton writes the following regarding the "Blue Angel" program: << Its the German "environmental excellence" marque. >> I have heard about the Blue Angel program from a client. My client's product is a large machine that is intended for use in commercial office environments. In my judgement, this product clearly qualifies as Class A under EN 55022. However, my client claims that to obtain the "Blue Angel" mark, the equipment must comply with the Class B limits. Can anyone clarify the following questions for me? 1) How widely recognized is the Blue Angel program? 2) Is it directed at consumer products, commercial products, or both? 3) Does the Blue Angel program impose stricter requirements than EN 55022 would otherwise impose? In particular, does it require that all products be Class B? I would certainly appreciate any clarification that others on emc-pstc might be able to offer. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc.
Re: Spread spectrum clock oscillator
In a message dated 97-02-18, Darrell Upson writes: << Does anyone know of a source for a spread spectrum clock oscillator. We have a sample of a 20MHz spread spectrum oscillator that we find reduces radiated emissions readings from our products by several dB. The company name was IMI in Milpitas, California, but no longer exists. Any leads to a similar product would be greatly appreciated. >> Darrell: I recently saw some sales literature on such a device. Apparently, the oscillator frequency is modulated slightly to spread the energy out a little bit, rather than having it all concentrated at exactly one frequency. The product that I have literature on is made in a standard 14-pin dip oscillator package. Frequencies of 14 to 120 MHz are available. The claimed reduction in emissions (at the fundamental and its harmonics) is 10 to 20 dB. Following is the contact info: Product name: EMC530 clock oscillator Manufacturer: EMC Component Group, Inc., Mountain View, CA Phone: 415-969-2286 Fax: 415-969-2287 Contact: Deborah Chen This is an interesting idea, and I would like to hear what experiences anyone on emc-pstc may have had with such devices. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc.
3750 Volt surge test on telecom relays/products
In a message dated 96-11-22, Harold Leipold writes: >We have been hearing from sales reps that customers are telling them someone >is getting ready to require a 3750 volt surge test on some telecom relays >and other telecom products. Harold: This might be a misunderstanding of the sort that often occurs with info via the sales channel. Back in the days when IEC 380 applied, the dielectric strength test for reinforced insulation was 3750 volts rms. IEC 950, which superseded IEC 380, reduced the test voltage for reinforced insulation to 3000 volts rms. However, the 3750 volt number still persists in the minds of many people. To this day, I still run into people who are convinced that the dielectric strength test for reinforced insulation is 3750 volts. Furthermore, as you probably know, reinforced insulation is not usually required for telecom relays. Until recently, clause 6.3.3 of IEC 950 called out supplementary insulation from earth ground. This level of insulation calls for a dielectric strength of 1500 volts rms (plus additional requirements on creepage and clearance). Under the recently adopted Amendment 4 to IEC 950, the requirement for supplementary insulation has been dropped. What remains is the requirement for basic insulation in clause 6.2.1.2. By my interpretation, this requires a dielectric strength of 1000 volts rms for applications involving a standard (TNV-3) analog phone line. Of course, there is also clause 6.4, which requires a dielectric strength of 1000 volts rms for most applications. Either way, you end up with a 1000 volt requirement. So, it would appear that the overall trend in dielectric strength requirements is down, rather than up. For surge testing (as opposed to dielectric strength testing), most countries that have such requirements use either 1000 volts or 1500 volts. South Africa is a notable exception, with a 4000 volt surge. Bellcore, as you mentioned, has a 2500 volt surge that the equipment must survive. In summary, I am not aware of any new regulatory requirement to withstand a 3750 volt "surge". The 3750 volt number sounds suspiciously like the old dielectric strength test for reinforced insulation under IEC 380. Aside from the 4000 volt surge in South Africa, the Bellcore 2500 volt surge represents the worst case that I am aware of. However, for high reliability applications of residential CPE, I recommend to my clients that they use relays which can withstand 4000 to 5000 volt surges. This is because the primary protector is either missing or malfunctioning at a small percentage of residential installations. Commercial installations tend to be better maintained. So, unless you have encountered an internal purchasing specification from one of my clients, I would be surprised if the 3750 volt "surge" requirement is real. But, then again, I have been surprised before! Please let me know if you find out that this "new requirement" is real. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc.
Re: Relay Needed
In a message dated 96-09-25, Ron Pickard writes: >I have been asked to post a request to identify a telecommunications relay >which has a published Dielectric Strength (Hi Pot) of 1500 vac, ( or 2121 >vdc) between open contacts which is sometimes called "Contact-to-Contact >Dielectric Rating". > >This relay has been reportedly difficult to find, and if found could remove >costly additions for isolation of a serial port, to meet EN 41003/EN60950. > Ron: Are you SURE that a 1500 volt dielectric strength test will be sufficient for EN 41003 and EN 60950? If you are trying to meet the requirements of supplementary insulation (clause 4.4.2 in EN 41003, and 6.3.3 in EN 60950), I think that creepage and clearance distance requirements will also apply. If you have information to the contrary, I would like to hear about it. A few years ago, the only relay I could find that met the requirements for supplementary insulation between open contacts was the Omron G5H-2. This part actually has reinforced insulation between open contacts! Nice part, but very expensive. I would be interested to hear about any other relays that might be suitable. Joe Randolph Telecom Design Consultant Randolph telecom, Inc.
Re: Y Capacitors - reply
In a message dated 96-07-24, Brent Taira writes: >Is anyone aware of an agency approved Y1 capacitor (IEC 384-14, 1993 edition) >that is stable at high temperatures (X7R)? > >One more item - a VDE Engineer had informed me that I needed to use either 3 - >VDE 565-1 or 2 - VDE 560-2 approved capacitors in series between primary and >secondary (Reinforced Insulation). Could anyone provide insight into this? >Perhaps, I did not argue strongly enough? Brent: The precise safety requirements for capacitors that bridge reinforced insulation have always confused me, and I do not claim to be an expert on this subject. However, I am under the impression that the Panasonic ECK-DNS series is suitable for bridging reinforced insulation. Values up to 1000 pF have a fairly stable temperature characteristic. I think that the notion of placing multiple caps in series stems from the reference to "insulation in thin sheet material" as described in subclause 2.9.4. Here, you can circumvent the requirement for 0.4 mm distance through insulation by using either two or three independent layers of insulation. Each capacitor can be viewed as a "thin sheet" of insulation. In the case of the Panasonic ECK-DNS series, I believe that the thickness of the dielectric exceeds 0.4 mm, so a single capacitor is sufficient. You may be interested to know that draft Amendment 4 to IEC 950 adds a subclause (2.2.8.1) that explicitly addresses bridging capacitors. It says that reinforced insulation can be bridged by any of the following: 1) A single capacitor that complies with IEC 384-14: 1993, subclass Y1 2) Two series caps that each comply with IEC 384-14: 1981, class U or Y 3) Two series caps that each comply with IEC 384-14: 1993, subclass Y2 or Y4 This amendment is not yet official, but it probably serves to show the direction that things are headed. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc.
Re: 'Trisil' source of supply
In a message dated 96-06-20, Victoria Pham writes: > I need to find the sources of supply for this TRISIL: Part number > T2503AA. Please advice which company I should contact to. Victoria: As far as I know, Trisils are made by SGS-Thomson. They are a "crowbar" type overvoltage protector. My SGS-Thomson data book is old, and it does not list this part. However, you might try calling them at 214-466-6000 in Texas. Another company that makes similar devices (called sidactors) is Teccor Electronics. They are also in Texas, at 214-580-1515. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc.
Re: Italian Telephone requirements
In a message dated 96-06-20, Jeff Welsh writes: >I am invovled with basic (Plain Old Telephones) that are used mostly in the USA. >We have an inquiry to use them both in Italy and the Czech Repulblic. > >Aside from the agency approvals, does anyone know what typical issues must >be adressed? I'm having difficulty in getting the specs. Does it invoolve >total redesign or just the front end? > > Jeff: To my knowledge, the current spec for POTS phones in Italy is CEI 103-5. The requirements are quite extensive. A translation of this spec can be obtained from BSI Language Services for about 620 pounds. Their fax number is 011-44-181-996-7047. I have the Czech spec for fax equipment, but not for POTS phones. The most notable aspect of the Czech requirement that I have is that the ringer capacitance must exceed 0.8 uF, and the total ringer impedance must be less than 10 K ohms. I suspect that the same requirements apply to POTS phones. If you are working with a simple, loop powered POTS phone, I think you will have difficulty getting a USA version approved in most European countries. In Europe, most countries place strict requirements on ringer response, return loss, speech levels, and frequency response. Most USA phones will fail one or more of these tests. You would probably have much better luck obtaining an approved POTS phone from an in-country supplier. This is what most companies do when all they need is a basic POTS phone to attach to their equipment. On the other hand, if your POTS function is integrated into the equipment and is not loop powered, you have more flexibility. I have had good success with designs that use transformer-coupled interfaces, with a digital signal processor to implement the POTS functions. This allows me to modify most of the ringing, dialing, and speech parameters in software, rather than hardware. In a few countries, POTS functions must be loop powered, so this approach is not universally applicable. However, the requirement for loop powered POTS can sometimes be waived, depending on the application. I hope that the above info gets you pointed in the right direction. Joe Randolph Telecom Design Consultant Randolph Telecom, Inc.