Re: The transformer screen/shield conundrum
I read in !emc-pstc that Rich Nute wrote (in <200307011612.jaa01...@epgc264.sdd.hp.com>) about 'The transformer screen/shield conundrum' on Tue, 1 Jul 2003: >So, you can't really consider North American 240 V >mains circuits in this discussion. We don't. > >* I think that's debatable, but probably doesn't need any further debate at this point. > >I do agree that electricity consumption per head in >North America is probably greater than in Europe. > >Having lived in Spain, I believe that North American >appliances are bigger and therefore consume more >electricity. It depends indirectly on where you lived in Spain. Some rural supplies are rated at only 6 kVA per household. Naturally, they don't have too many 3 kW appliances in those households. [snip] > >Neither of these topics has anything to do with the >current available into a fault before the circuit >breaker operates. They are about matters which you raised in your responses, and indirectly refer to the crucial matter of supply impedance. -- Regards, John Woodgate, OOO - Own Opinions Only. http://www.jmwa.demon.co.uk Interested in professional sound reinforcement and distribution? Then go to http://www.isce.org.uk PLEASE do NOT copy news posts to me by E-MAIL! 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: emc_p...@symbol.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org Archive is being moved, we will announce when it is back on-line. All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
Re: The transformer screen/shield conundrum
Hi John: > >If the breakers are 230 V/16 A and 120 V/15 A, then > >the power distributed at 230 V is roughly twice that > >at 120 V. > > But your '120 V' is also '240 V' for some equipment. Besides, I think I > was being 'generous'; electrical energy consumption per head in USA is > probably greater than in Europe. In North America, the 240 V is supplied to high-power- consumption equipment such as water heaters, electric ovens, electric cook-tops, electric clothes dryers, electric furnaces, and air conditioners. The 240-V circuits are rated 20, 30, and 50 amps. In North America, typical equipment is not rated for use on 240 V, but 120 V. While 240 V is supplied to each home, use of 240 V for applications other than those I mentioned, would require a special electrical installation. This is extremely rare. So, you can't really consider North American 240 V mains circuits in this discussion. We don't. * I do agree that electricity consumption per head in North America is probably greater than in Europe. Having lived in Spain, I believe that North American appliances are bigger and therefore consume more electricity. Also, NA electrical appliances tend to be less efficient than Euro appliances. NA clothes and dish washers use much more water than Euro versions. * Neither of these topics has anything to do with the current available into a fault before the circuit breaker operates. Best regards, Rich 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: emc_p...@symbol.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org Archive is being moved, we will announce when it is back on-line. All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
Re: The transformer screen/shield conundrum
I read in !emc-pstc that Rich Nute wrote (in <200306302129.oaa24...@epgc264.sdd.hp.com>) about 'The transformer screen/shield conundrum' on Mon, 30 Jun 2003: >I disagree that the same power is distributed at >230 V and 120 V. We have a certain amount of misunderstanding. > >If the breakers are 230 V/16 A and 120 V/15 A, then >the power distributed at 230 V is roughly twice that >at 120 V. But your '120 V' is also '240 V' for some equipment. Besides, I think I was being 'generous'; electrical energy consumption per head in USA is probably greater than in Europe. > >My subject was *not* supply voltage tolerance, but >system *design* goal for percent voltage drop at max >rated load (the circuit-breaker rating). OK, that's clear now. > >I really believe that EEs in Europe design supply >systems to 3% voltage drop at rated current, 16 A, >just as they do in North America. The reason I so >believe is that the wire sizes for 15 A (NA) or 16 A >(Euro) circuits are nearly the same. The same size >wire at (nearly) the same current would give (nearly) >the same percent voltage drop. Yes, well, it's actually 3.3% for the worst case, see the amendment to IEC/EN 61000-3-3. But 3% normally. > >(If the system source impedance is 0.47 ohms, then, at >230 V, the system voltage drop would be slightly more >than 6% at maximum rated load.) Yes, but that 0.47 ohms is the source impedance at the point of common coupling (PCC), which you weren't addressing, AIUI. Your 3% is the drop >from the PCC to the wall socket? > >We really weren't talking about voltage, but about >the maximum current into a fault in cord-connected >equipment. The voltage tolerance can be ignored for >this purpose. Yes, but the voltage tolerance provides information on the impedance, which is itself not specified, and the impedance is what determines the prospective fault current. > >> That still gives 490 A, which would be embarrassingly big, but it apples >> at the 'point of common connection', not at the wall socket and still >> less at the equipment. > >My comments were for the current available at the >wall socket. In North America, building wiring is >designed for maximum 3% voltage drop at maximum >rated current. Yes, OK, that works in Europe as well. > >In NA, at the point of common connection between the >building and the electicity supplier, the electricity >supplier maintains a nearly constant voltage. He >does so not only by his system design, but also by >dynamically adjusting the voltage of the source. Two points there: 1. Yes, in Europe the voltage may be 'nearly constant', but it is only the 'declared voltage' for some customers. Those nearer the substation transformer get more, and those further away get less, but within legal limits. Typically, one transformer feeds 200 to 500 residential customers. 2. It is possible to adjust the LV voltage but only by adjusting the MV supply to the transformers. On-load tap-changing at LV is extremely rare. In the Americas, this adjustment is not so easy, because of the very large number of transformers, some of which are likely at a given time to be lightly loaded, and some heavily loaded. Most of the source impedance of the LV network is in the transformer, so those that are heavily loaded will deliver fewer volts, and if the MV were increased to compensate, the lightly-loaded transformers would deliver too many volts. Of course, seasonal load variations tend to affect each transformer equally, so MV adjustment does work in this case. -- Regards, John Woodgate, OOO - Own Opinions Only. http://www.jmwa.demon.co.uk Interested in professional sound reinforcement and distribution? Then go to http://www.isce.org.uk PLEASE do NOT copy news posts to me by E-MAIL! 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: emc_p...@symbol.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org Archive is being moved, we will announce when it is back on-line. All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
Re: The transformer screen/shield conundrum
Hi John: > >For 3% voltage drop at maximum rated load, the > >source resistance is about 0.2 ohm. So the > >maximum rms current would be system nominal > >voltage divided by 0.2 ohm, or 600 amps for 120 > >V and 1150 amps for 230 V. > > For 230V, there are two modifications. One is that roughly the same > *power* is distributed as at 120 V, so the system impedance would be > doubled. But, in addition, the supply voltage tolerances are greater > than 3%: in Europe the tolerance is about double, so, overall, the > system impedance could be four times that 0.2 ohm. But in fact, 90% of > supplies are below 0.47 ohms (see IEC 60725, under amendment). I disagree that the same power is distributed at 230 V and 120 V. If the breakers are 230 V/16 A and 120 V/15 A, then the power distributed at 230 V is roughly twice that at 120 V. My subject was *not* supply voltage tolerance, but system *design* goal for percent voltage drop at max rated load (the circuit-breaker rating). I really believe that EEs in Europe design supply systems to 3% voltage drop at rated current, 16 A, just as they do in North America. The reason I so believe is that the wire sizes for 15 A (NA) or 16 A (Euro) circuits are nearly the same. The same size wire at (nearly) the same current would give (nearly) the same percent voltage drop. (If the system source impedance is 0.47 ohms, then, at 230 V, the system voltage drop would be slightly more than 6% at maximum rated load.) We really weren't talking about voltage, but about the maximum current into a fault in cord-connected equipment. The voltage tolerance can be ignored for this purpose. > That still gives 490 A, which would be embarrassingly big, but it apples > at the 'point of common connection', not at the wall socket and still > less at the equipment. My comments were for the current available at the wall socket. In North America, building wiring is designed for maximum 3% voltage drop at maximum rated current. In NA, at the point of common connection between the building and the electicity supplier, the electricity supplier maintains a nearly constant voltage. He does so not only by his system design, but also by dynamically adjusting the voltage of the source. Best regards, Rich 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: emc_p...@symbol.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org Archive is being moved, we will announce when it is back on-line. All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
Re: The transformer screen/shield conundrum
I read in !emc-pstc that Rich Nute wrote (in <200306301811.laa23...@epgc264.sdd.hp.com>) about 'The transformer screen/shield conundrum' on Mon, 30 Jun 2003: >For 3% voltage drop at maximum rated load, the >source resistance is about 0.2 ohm. So the >maximum rms current would be system nominal >voltage divided by 0.2 ohm, or 600 amps for 120 >V and 1150 amps for 230 V. For 230V, there are two modifications. One is that roughly the same *power* is distributed as at 120 V, so the system impedance would be doubled. But, in addition, the supply voltage tolerances are greater than 3%: in Europe the tolerance is about double, so, overall, the system impedance could be four times that 0.2 ohm. But in fact, 90% of supplies are below 0.47 ohms (see IEC 60725, under amendment). That still gives 490 A, which would be embarrassingly big, but it apples at the 'point of common connection', not at the wall socket and still less at the equipment. > >At these currents, the overcurrent device will >operate in about 0.01 second maximum, which is >less than 1 cycle at either 50 Hz or 60 Hz. So, >the current will never rise to the calculated >currents. But the overcurrent device must be rated to break the prospective current. HBC fuses rated at 1500 A are OK, but many EBC fuses are rated at 150 A, and it is doubtful that they can be used with 230 V mains supplies. > >So, the maximum current is set by the overcurrent >device characteristics, not by the resistance of >the supply system. > >If we assume a time of 1 cycle (0.02 s), then >a typical overcurrent device will trip at a >maximum of 20 times the rated current. So the >maximum current for a 15-amp overcurrent device >for a duration of 1 cycle will be about 300 amps. > >If we add in the resistance of a 2-meter power >cord (0.05 ohm per wire) and the maximum >resistance of the equipment (0.10 ohm), then the >maximum fault current is less than 300 amps. >Indeed, it drops to about 150-200 amps. Yes, I agree, but still uncomfortably in excess of 150 A. I think we need supplies of fuses rated at 300 A, not 150 A or 1500 A. -- Regards, John Woodgate, OOO - Own Opinions Only. http://www.jmwa.demon.co.uk Interested in professional sound reinforcement and distribution? Then go to http://www.isce.org.uk PLEASE do NOT copy news posts to me by E-MAIL! 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: emc_p...@symbol.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org Archive is being moved, we will announce when it is back on-line. All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
Re: The transformer screen/shield conundrum
Hi Richard and John: > >[R_Hughes >] I know it as prospective short-circuit current, but maybe I go > >boldly whereas you boldly go? > > Do we know what the current values are in various countries, for > equipment connected to wall-sockets? Assuming that the mains lead has > the resistance that can be deduced from the maximum permissible > resistance of its PEC. The short-circuit current value is limited by the sum of the resistance of the phase conductor and the resistance of the neutral/PE conductor (assuming all three conductors are the same size). The resistance is a function of the physical distance (length of wire) from the distribution transformer and the contact resistances of the various connections and overcurrent devices in the current path. Since the distance from the distribution transformer to the load can vary from relatively small to quite long, the deduction of the maximum permissable resistance of the PE and other conductors seems indeterminable. However, according to my USA EE colleagues, the design of distribution systems is based on no more than 3% voltage drop (wrt nominal voltage) at maximum rated load, regardless of distance >from the distribution transformer to the load. If we can assume this to be true throughout the world, then we can deduce both the mains lead resistance and the PE conductor resistance. See my paper, "Equipotentiality and Grounding, Derivation of Grounding Resistance for Equipment," http://www.ewh.ieee.org/soc/emcs/pstc/tech-spk.htm Click on "equipotentiality and grounding." Page 20 gives the source resistance for mains systems as a function of system voltage drop and maximum rated current. For 3% voltage drop at maximum rated load, the source resistance is about 0.2 ohm. So the maximum rms current would be system nominal voltage divided by 0.2 ohm, or 600 amps for 120 V and 1150 amps for 230 V. At these currents, the overcurrent device will operate in about 0.01 second maximum, which is less than 1 cycle at either 50 Hz or 60 Hz. So, the current will never rise to the calculated currents. So, the maximum current is set by the overcurrent device characteristics, not by the resistance of the supply system. If we assume a time of 1 cycle (0.02 s), then a typical overcurrent device will trip at a maximum of 20 times the rated current. So the maximum current for a 15-amp overcurrent device for a duration of 1 cycle will be about 300 amps. If we add in the resistance of a 2-meter power cord (0.05 ohm per wire) and the maximum resistance of the equipment (0.10 ohm), then the maximum fault current is less than 300 amps. Indeed, it drops to about 150-200 amps. Best regards, Rich 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: emc_p...@symbol.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org Archive is being moved, we will announce when it is back on-line. All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
Re: The transformer screen/shield conundrum
I read in !emc-pstc that richhug...@aol.com wrote (in <6C87B93E.0C9D7087 .0ba45...@aol.com>) about 'The transformer screen/shield conundrum' on Sat, 28 Jun 2003: >[R_Hughes >] I know it as prospective short-circuit current, but maybe I go >boldly whereas you boldly go? Do we know what the current values are in various countries, for equipment connected to wall-sockets? Assuming that the mains lead has the resistance that can be deduced from the maximum permissible resistance of its PEC. -- Regards, John Woodgate, OOO - Own Opinions Only. http://www.jmwa.demon.co.uk Interested in professional sound reinforcement and distribution? Then go to http://www.isce.org.uk PLEASE do NOT copy news posts to me by E-MAIL! 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: emc_p...@symbol.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org Archive is being moved, we will announce when it is back on-line. All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
re: The transformer screen/shield conundrum
Rich, RN> I'm surprised that, today, a SMPS would used screen/ shield construction. [R_Hughes >] Not all power supplies are switch mode, even today. I bet that the HiFi addicts who say that they can hear the difference between a valve amp and a transistor amp would absolutely hate the idea of anything other than a pig-iron transformer (not intended to be a technical description, John). RN> I presume you are referring to the circuit prospective current. (I am not familiar with the proposal to WG8.) [R_Hughes >] I know it as prospective short-circuit current, but maybe I go boldly whereas you boldly go? RN> However, I believe that the circuit prospective current is of little or no consequence when testing a robust earthing circuit complying with the constructional requirements. (Of course we should test this hypothesis.) [R_Hughes >] Exactly the point I made in WG8. Best regards, Richard 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: emc_p...@symbol.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org Archive is being moved, we will announce when it is back on-line. All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
Re: The transformer screen/shield conundrum
I read in !emc-pstc that Rich Nute wrote (in <200306271743.kaa29...@epgc264.sdd.hp.com>) about 'The transformer screen/shield conundrum' on Fri, 27 Jun 2003: >The screen/shield must be capable of carrying >the full fault current and is often tested >using the 25-amp test. C.2 of Annex C of IEC/EN 60950-1:2001 says, in the Notes: ' an overload device [should be 'protective device', I think] will open the circuit before the screen is destroyed.' If the product is fitted with an internal fuse, say 5 A, I don't see any need for the screen to withstand 25 A. I certainly would not want to design a transformer with a screen that would withstand 25 A unless the cost and difficulty of manufacture were amply justified. At 25 A, the current divergence from the fault point and convergence to the lead out could well have significant effects. There are ways of dealing with that, but it's bothersome. -- Regards, John Woodgate, OOO - Own Opinions Only. http://www.jmwa.demon.co.uk Interested in professional sound reinforcement and distribution? Then go to http://www.isce.org.uk PLEASE do NOT copy news posts to me by E-MAIL! 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: emc_p...@symbol.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org Archive is being moved, we will announce when it is back on-line. All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
Re: The transformer screen/shield conundrum
I read in !emc-pstc that Rich Nute wrote (in <200306271956.maa00...@epgc264.sdd.hp.com>) about 'The transformer screen/shield conundrum' on Fri, 27 Jun 2003: >I would agree that this current, at a very small >contact area, is the driver for the hole in the >screen/shield. OK, the trade secret is to use two layers of foil. Only the first one gets the hole. -- Regards, John Woodgate, OOO - Own Opinions Only. http://www.jmwa.demon.co.uk Interested in professional sound reinforcement and distribution? Then go to http://www.isce.org.uk PLEASE do NOT copy news posts to me by E-MAIL! 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: emc_p...@symbol.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org Archive is being moved, we will announce when it is back on-line. All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
Re: The transformer screen/shield conundrum
Hi Richard: > You mention a few times the 25 Amp test. The designs I have seen that try > and use this approach (and I don't ever recall seeing a design I was happy > with) were all switch mode power supplies where there was an input fuse of > about 2A, meaning that the test would be performed at 3A in the past (and > will be performed at 4A in the future, if the CDs go through un-changed). I'm surprised that, today, a SMPS would used screen/ shield construction. The effectiveness of a SMPS transformer is a function of the physical proximity of primary and secondary windings. All of the SMPS transformers that I have been involved with over the past 10 years or so use reinforced/double-insulation between primary and secondary. While I mentioned the 25-amp test, I fully concur that the test current is a function of the relevant overcurrent device. > problem I am not convinced that it is the whole story. For me, another > chapter in this story is the part played by the very high surge current that > flows for a very short period of time. It is here where we could usefully I presume you are referring to the circuit prospective current. (I am not familiar with the proposal to WG8.) I would agree that this current, at a very small contact area, is the driver for the hole in the screen/shield. However, I believe that the circuit prospective current is of little or no consequence when testing a robust earthing circuit complying with the constructional requirements. (Of course we should test this hypothesis.) Best regards, Rich 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: emc_p...@symbol.com For policy questions, send mail to: Richard Nute: ri...@ieee.org Jim Bacher: j.bac...@ieee.org Archive is being moved, we will announce when it is back on-line. All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
re: The transformer screen/shield conundrum
Rich, Thanks for the Email in which you mention many of the points already discussed in IEC TC74. So what are we going to do with the new hazard based standard? Are we going to permit constructions that have been in IEC 60065 and IEC 60950 for years or are we going to introduce a test that is rather more realistic of the actual failure condition than simply running an earth bond test? You mention a few times the 25 Amp test. The designs I have seen that try and use this approach (and I don't ever recall seeing a design I was happy with) were all switch mode power supplies where there was an input fuse of about 2A, meaning that the test would be performed at 3A in the past (and will be performed at 4A in the future, if the CDs go through un-changed). You mention that current density caused by a small point of contact is the reason why a short from the winding to the earthed foil will cause the formation of a hole in the foil. Although I agree that it is part of the problem I am not convinced that it is the whole story. For me, another chapter in this story is the part played by the very high surge current that flows for a very short period of time. It is here where we could usefully apply the surge test proposed in WG8 by Taylor and Bahra, but rejected as being not sufficiently well thought through. So there you have it, some of us Brits just like having a good discussion (what you call an argument). Unfortunately there were no Americans around to argue with in our morning so we just had to make do with arguing with one another! Glad to see you've joined the fray! Richard From: ri...@sdd.hp.com [mailto:ri...@sdd.hp.com] Sent: Saturday, June 28, 2003 5:43 AM To: j...@jmwa.demon.co.uk; richhug...@aol.com Cc: emc-p...@majordomo.ieee.org Subject: The transformer screen/shield conundrum Hi John and Richard: The transformer screen/shield between primary and secondary windings is intended to carry the fault current in the event of a failure of basic insulation between the primary and the screen/shield. The screen/shield must be capable of carrying the full fault current and is often tested using the 25-amp test. The object of the test is to connect to the screen/shield, and then pass the 25 amps through the shield to its transformer terminal and then to the PE terminal. The difficult part is how to connect the tester to the screen/shield. When the screen/ shield is a copper foil, a special transformer must be wound with a wire connected (soldered) to the foil screen/shield and brought out for connection to the 25-amp source. If this cannot be done, then a wire is soldered to the edge of the foil screen/shield by forcefully separating the sheet insulation that extends beyond the edge of the shield, inserting the soldering iron tip, and soldering the wire. (A constructional problem is that the screen/ shield must extend BEYOND the primary winding so that no part of the primary winding is "exposed" to the secondary winding.) (Another constructional problem is that the ends of the foil must be overlapped, but with insulation inserted between the overlaps; otherwise, the screen/shield would constitute a shorted turn and cause all kinds of electrical, magnetic, and thermal problems.) The connection of the tester to the foil screen/shield typically has a very large contact area. Any thickness of foil screen/ shield passes the test. The conundrum is that the failure of basic insulation could result in a point contact between the primary winding and the shield. The point-contact, because of its small area, has very high current density, and will blow a hole in thin foil (that otherwise passes the 25-amp test). With the hole in the screen/sheild, the fault to the foil opens, and the fault no longer exists -- even though there is a failure of basic insulation. The connection of the foil to the PE remains intact, and there is no shock hazard. (Of course, the heat of the point-contact fault may very well cause failure of the foil screen/shield-to-secondary functional insulation, which might very well result in a shock hazard from the secondary SELV circuit.) The point is that the test does not necessarily test what can occur within the transformer, yet will pass the transformer screen/shield construction. Nevertheless, when the basic insulation fault occurs, the hole in the screen/shield may very well end the event safely. Best regards, Rich 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: emc_p...@symbol.com For policy questions, send mail to: Richard Nute: ri.
