Re: [Elecraft] P3 generates noise on 144MHz
Last night I cleaned all mating metal surfaces in the P3 but, alas, this made no difference. The noise was still at S6. When I had the I/O printed circuit board out of the case, I noticed that the earth planes for the power supply section and RS232 section are separated by a zig zag gap. They are commoned together towards the inner edge. This has the effect of lengthening the route of RF currents from the earthy side of C500 to the metal case. My thoughts are that C500 will do a great job of removing RF differential mode voltages across the power cord. It will not do such a good job of removing common mode RF voltages from the power cord. I believe it is the common mode RF voltages that are the problem. This was confirmed when I replaced the power cord with a coaxial one (as suggested by John, G4ZTR). This made no difference to the noise. I found a ferrite ring which was large enough to pass the power plug through it. In fact I passed the plug through it three times to form two tight turns and I moved this choke as close as possible to the P3 power socket. The result was that the noise dropped to S4. Almost there! I think it would be better if the filter was totally inside the P3 case. I will try that next. Roger Crofts, VK4YB __ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html
Re: [Elecraft] P3 generates noise on 144MHz
I fitted a commercial EMI filter (intended for 240V AC operation) inside my P3. It is a fully sealed unit with solder tags. This unit was connected in the + and - DC supply line. This filter completely removed any trace of noise from the power cord to the P3. I used a short coax with a one inch extension of the center wire to act as a "sniffer" to determine where the noise was coming from. This coax was connected to the K144XV input. Did this solve the problem? No! not entirely. I discovered there was now a small amount of noise being radiated from the RS232 lead and the IF input coax. These cables had been free of noise before. It seems to me that it is acting a bit like a boiling kettle - Block the spout and the steam comes out the holes in the lid! Why was there so much noise generation in the first place? I used my sniffer inside the P3 box to try and track down the precise source of the noise. But no luck. The noise was overwhelming everywhere at S9 + 20dB rising to S9 + 40dB near any of the ICs. This was on any frequency in the 2 meter band and beyond. At this point I re-visited the earthing between the panels that make up the P3 box. If you have been following this thread, you will know that I had previously cleaned all the mating surfaces. I looked again at the front panel. It is connected by four screws to the corner blocks. You can see where rectangles of masking tape have been placed to prevent the black paint from reaching the area which will mate with the corner blocks. The masking tape has not quite reached the edge of the board, leaving a one mm wide strip of paint right at both corner edges. It seemed very unlikely that this was preventing a good connection between the panels. Nevertheless I removed all trace of the paint in all four corners of both the front and back panels using a file and sharp knife. Did this have any effect? Well Yes! It reduced the noise by about 6dB. The noise which had started at S9 + 5dB was now down to about S3 which was the target value. The carriers at 19KHz spacing were just, but only just, perceptible with the 2 meter antenna connected. It would have been interesting to know what the noise would now be like without the EMI filter. Sorry guys (and gals), I did not do that experiment. When you have got something that works - leave it alone! I was also able to connect my other accessories to the switched 12V line from the K3 without affecting the noise. Another effect of my paint scraping exercise was that the position of the P3 was no longer critical. Before, if I pulled the P3 forward, the noise went down, but returned with the P3 in its correct position aligned with the K3. Roger Crofts, VK4YB __ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html
Re: [Elecraft] P3 generates noise on 144MHz
Roger saith: "When I had the I/O printed circuit board out of the case, I noticed that the earth planes for the power supply section and RS232 section are separated by a zig zag gap." Ouch! That indicates that whoever laid out the board had no idea that 'gapping' the board is a BAD IDEA(tm). Please, whoever was responsible for laying out the I/O board for the P3, go to this site, and learn what REALLY has to be done to properly design PC boards for EMC: http://www.learnemc.com/index.html I attended a whirlwind 4-hour course on automotive EMC at last year's Freescale Technology Forum that was a very compressed version of the 2-day courses shown on the LearnEMC home page, and it was a real eye-opener. If there is anything critical as a take away from that experience, it was that the PCB layout will make-or-break your EMC issues, and is the best and most cost-effective way to stop EMC issues, rather than trying to use ferrites, etc. after the fact, all of which will never be as effective. If you can't do the two-day course (highly recommended) there are free extensive tutorials to be found at the "EMC Tutorials" link on the main page: http://www.learnemc.com/EMC-Tutorials.html It's all good stuff, and should be mandatory learning for anyone laying out PC boards these days. 73, -- Dave, N8SBE > ---- Original Message ---- > Subject: Re: [Elecraft] P3 generates noise on 144MHz > From: "Roger Crofts" > Date: Tue, February 12, 2013 6:02 pm > To: "elecraft reflector" > > > Last night I cleaned all mating metal surfaces in the P3 but, alas, this made > no difference. The noise was still at S6. > When I had the I/O printed circuit board out of the case, I noticed that the > earth planes for the power supply section and RS232 section are separated by > a zig zag gap. They are commoned together towards the inner edge. This has > the effect of lengthening the route of RF currents from the earthy side of > C500 to the metal case. My thoughts are that C500 will do a great job of > removing RF differential mode voltages across the power cord. It will not do > such a good job of removing common mode RF voltages from the power cord. I > believe it is the common mode RF voltages that are the problem. This was > confirmed when I replaced the power cord with a coaxial one (as suggested by > John, G4ZTR). This made no difference to the noise. > I found a ferrite ring which was large enough to pass the power plug through > it. In fact I passed the plug through it three times to form two tight turns > and I moved this choke as close as possible to the P3 power socket. The > result was that the noise dropped to S4. Almost there! I think it would be > better if the filter was totally inside the P3 case. I will try that next. > > Roger Crofts, VK4YB > __ > Elecraft mailing list > Home: http://mailman.qth.net/mailman/listinfo/elecraft > Help: http://mailman.qth.net/mmfaq.htm > Post: mailto:Elecraft@mailman.qth.net > > This list hosted by: http://www.qsl.net > Please help support this email list: http://www.qsl.net/donate.html __ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html
Re: [Elecraft] P3 generates noise on 144MHz
What does the "zig zag" gap look like and what is the issue with it? Or...how does it contribute to the noise? Thanks, Sent from my iPad Chuck, KE9UW (Jack for BMW motorcycles) On Feb 18, 2013, at 11:42 AM, "Dave New, N8SBE" wrote: > Roger saith: > > "When I had the I/O printed circuit board out of the case, I noticed > that the earth planes for the power supply section and RS232 section are > separated by a zig zag gap." > > Ouch! That indicates that whoever laid out the board had no idea that > 'gapping' the board is a BAD IDEA(tm). > > Please, whoever was responsible for laying out the I/O board for the P3, > go to this site, and learn what REALLY has to be done to properly design > PC boards for EMC: > > http://www.learnemc.com/index.html > > I attended a whirlwind 4-hour course on automotive EMC at last year's > Freescale Technology Forum that was a very compressed version of the > 2-day courses shown on the LearnEMC home page, and it was a real > eye-opener. If there is anything critical as a take away from that > experience, it was that the PCB layout will make-or-break your EMC > issues, and is the best and most cost-effective way to stop EMC issues, > rather than trying to use ferrites, etc. after the fact, all of which > will never be as effective. > > If you can't do the two-day course (highly recommended) there are free > extensive tutorials to be found at the "EMC Tutorials" link on the main > page: > > http://www.learnemc.com/EMC-Tutorials.html > > It's all good stuff, and should be mandatory learning for anyone laying > out PC boards these days. > > 73, > > -- Dave, N8SBE > >> Original Message >> Subject: Re: [Elecraft] P3 generates noise on 144MHz >> From: "Roger Crofts" >> Date: Tue, February 12, 2013 6:02 pm >> To: "elecraft reflector" >> >> >> Last night I cleaned all mating metal surfaces in the P3 but, alas, this >> made no difference. The noise was still at S6. >> When I had the I/O printed circuit board out of the case, I noticed that the >> earth planes for the power supply section and RS232 section are separated by >> a zig zag gap. They are commoned together towards the inner edge. This has >> the effect of lengthening the route of RF currents from the earthy side of >> C500 to the metal case. My thoughts are that C500 will do a great job of >> removing RF differential mode voltages across the power cord. It will not do >> such a good job of removing common mode RF voltages from the power cord. I >> believe it is the common mode RF voltages that are the problem. This was >> confirmed when I replaced the power cord with a coaxial one (as suggested by >> John, G4ZTR). This made no difference to the noise. >> I found a ferrite ring which was large enough to pass the power plug through >> it. In fact I passed the plug through it three times to form two tight turns >> and I moved this choke as close as possible to the P3 power socket. The >> result was that the noise dropped to S4. Almost there! I think it would be >> better if the filter was totally inside the P3 case. I will try that next. >> >> Roger Crofts, VK4YB >> __ >> Elecraft mailing list >> Home: http://mailman.qth.net/mailman/listinfo/elecraft >> Help: http://mailman.qth.net/mmfaq.htm >> Post: mailto:Elecraft@mailman.qth.net >> >> This list hosted by: http://www.qsl.net >> Please help support this email list: http://www.qsl.net/donate.html > __ > Elecraft mailing list > Home: http://mailman.qth.net/mailman/listinfo/elecraft > Help: http://mailman.qth.net/mmfaq.htm > Post: mailto:Elecraft@mailman.qth.net > > This list hosted by: http://www.qsl.net > Please help support this email list: http://www.qsl.net/donate.html __ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html
Re: [Elecraft] P3 generates noise on 144MHz
The point is that splitting the ground planes in an attempt to isolate them, often makes the situation the PCB designer thought they were fixing much worse. It's a case of knowing where the actual return paths are, and minimizing the length of those return paths. The other half of the EMC equation is proper loading of signal lines, to shape the edges, and suppress ringing. It is far more effective to put series 'ballast' resistors in the signal lines, rather than try to suppress EMC with capacitors across the lines to ground. Please refer to the posted EMC tutorial links I posted below for graphics/explanations. Designing for EMC is not a easy chore, but starting with a bad PCB design will make it almost impossible to suppress systems issues that may crop up after the layout is 'frozen' and the unit is headed for production. At that point, trying to fix the issues with shielding and/or ferrite cores is very expensive, and often doesn't produce the desired results. Nothing short of re-laying out the offending PCB will suffice in the majority of cases. 73, -- Dave, N8SBE > Original Message -------- > Subject: Re: [Elecraft] P3 generates noise on 144MHz > From: "hawley, charles j jr" > Date: Mon, February 18, 2013 1:08 pm > To: "Dave New, N8SBE" > Cc: Roger Crofts ,elecraft reflector > > > > What does the "zig zag" gap look like and what is the issue with it? Or...how > does it contribute to the noise? > Thanks, > > Sent from my iPad > Chuck, KE9UW > (Jack for BMW motorcycles) > > On Feb 18, 2013, at 11:42 AM, "Dave New, N8SBE" wrote: > > > Roger saith: > > > > "When I had the I/O printed circuit board out of the case, I noticed > > that the earth planes for the power supply section and RS232 section are > > separated by a zig zag gap." > > > > Ouch! That indicates that whoever laid out the board had no idea that > > 'gapping' the board is a BAD IDEA(tm). > > > > Please, whoever was responsible for laying out the I/O board for the P3, > > go to this site, and learn what REALLY has to be done to properly design > > PC boards for EMC: > > > > http://www.learnemc.com/index.html > > > > I attended a whirlwind 4-hour course on automotive EMC at last year's > > Freescale Technology Forum that was a very compressed version of the > > 2-day courses shown on the LearnEMC home page, and it was a real > > eye-opener. If there is anything critical as a take away from that > > experience, it was that the PCB layout will make-or-break your EMC > > issues, and is the best and most cost-effective way to stop EMC issues, > > rather than trying to use ferrites, etc. after the fact, all of which > > will never be as effective. > > > > If you can't do the two-day course (highly recommended) there are free > > extensive tutorials to be found at the "EMC Tutorials" link on the main > > page: > > > > http://www.learnemc.com/EMC-Tutorials.html > > > > It's all good stuff, and should be mandatory learning for anyone laying > > out PC boards these days. > > > > 73, > > > > -- Dave, N8SBE > > > >> Original Message > >> Subject: Re: [Elecraft] P3 generates noise on 144MHz > >> From: "Roger Crofts" > >> Date: Tue, February 12, 2013 6:02 pm > >> To: "elecraft reflector" > >> > >> > >> Last night I cleaned all mating metal surfaces in the P3 but, alas, this > >> made no difference. The noise was still at S6. > >> When I had the I/O printed circuit board out of the case, I noticed that > >> the earth planes for the power supply section and RS232 section are > >> separated by a zig zag gap. They are commoned together towards the inner > >> edge. This has the effect of lengthening the route of RF currents from the > >> earthy side of C500 to the metal case. My thoughts are that C500 will do a > >> great job of removing RF differential mode voltages across the power cord. > >> It will not do such a good job of removing common mode RF voltages from > >> the power cord. I believe it is the common mode RF voltages that are the > >> problem. This was confirmed when I replaced the power cord with a coaxial > >> one (as suggested by John, G4ZTR). This made no difference to the noise. > >> I found a ferrite ring which was large enough to pass the power plug > >> through it. In fact I passed the plug through it three times to form two > >> tight turns and I moved this choke
Re: [Elecraft] P3 generates noise on 144MHz
I would add further "noise" to the noise concerns... during the process of repairing PC's/Macs/etc. it is very apparent there are a plethora of pathetic power supplies and converting devices in the shack. My P3 detects them very well. The great thing about the P3 is I can disconnect and find the offenders. It's nice my antennas are relatively far away but the idea of shack noise, which does affect all my receiving equipment, it is very helpful to note the noise floor on the P3... The power connection to the P3 appears to need coaxial shielding and filtering, IMHO. I have mine connected via the K3 power, and it is via a simple power splitter. This isn't so hot. 73, Bill K9YEQ -Original Message- From: elecraft-boun...@mailman.qth.net [mailto:elecraft-boun...@mailman.qth.net] On Behalf Of Roger Crofts Sent: Sunday, February 17, 2013 4:55 PM To: elecraft reflector Subject: Re: [Elecraft] P3 generates noise on 144MHz I fitted a commercial EMI filter (intended for 240V AC operation) inside my P3. It is a fully sealed unit with solder tags. This unit was connected in the + and - DC supply line. This filter completely removed any trace of noise from the power cord to the P3. I used a short coax with a one inch extension of the center wire to act as a "sniffer" to determine where the noise was coming from. This coax was connected to the K144XV input. Did this solve the problem? No! not entirely. I discovered there was now a small amount of noise being radiated from the RS232 lead and the IF input coax. These cables had been free of noise before. It seems to me that it is acting a bit like a boiling kettle - Block the spout and the steam comes out the holes in the lid! Why was there so much noise generation in the first place? I used my sniffer inside the P3 box to try and track down the precise source of the noise. But no luck. The noise was overwhelming everywhere at S9 + 20dB rising to S9 + 40dB near any of the ICs. This was on any frequency in the 2 meter band and beyond. At this point I re-visited the earthing between the panels that make up the P3 box. If you have been following this thread, you will know that I had previously cleaned all the mating surfaces. I looked again at the front panel. It is connected by four screws to the corner blocks. You can see where rectangles of masking tape have been placed to prevent the black paint from reaching the area which will mate with the corner blocks. The masking tape has not quite reached the edge of the board, leaving a one mm wide strip of paint right at both corner edges. It seemed very unlikely that this was preventing a good connection between the panels. Nevertheless I removed all trace of the paint in all four corners of both the front and back panels using a file and sharp knife. Did this have any effect? Well Yes! It reduced the noise by about 6dB. The noise which had started at S9 + 5dB was now down to about S3 which was the target value. The carriers at 19KHz spacing were just, but only just, perceptible with the 2 meter antenna connected. It would have been interesting to know what the noise would now be like without the EMI filter. Sorry guys (and gals), I did not do that experiment. When you have got something that works - leave it alone! I was also able to connect my other accessories to the switched 12V line from the K3 without affecting the noise. Another effect of my paint scraping exercise was that the position of the P3 was no longer critical. Before, if I pulled the P3 forward, the noise went down, but returned with the P3 in its correct position aligned with the K3. Roger Crofts, VK4YB __ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html __ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html
