From bobkatz at digido.com Sat Dec 2 09:21:11 2023 From: bobkatz at digido.com (Bob Katz) Date: Sat, 2 Dec 2023 12:21:11 -0500 Subject: [ProAudio] Muncy Revisited: Pin 1 to chassis through a capacitor? Message-ID: Hi, Pro audio! Happy Holidays from Bob. Naturally, I started a firestorm on facebook asking a question about grounding in a headphone amplifier I'm building. I made a mistake by forgetting to show the connection between the 0 volt (transformer center tap) and chassic ground... oops, so the conversation went downhill and even after I revised my diagram some people were responding to my first diagram with the error :-( I insisted (and still insist) that EMI/RFI interference at the audio input is ameliorrated in Muncy's "rule 2": by connecting pin 1 to chassis directly at the audio input. This is the centerpiece of Muncy's seminal article "Noise Susceptibility in Analog and Digital Signal Processing Systems" I know this from experience as well as theory, having successfully installed a microphone preamplifier on the 101st floor of the World Trade Center in 1995 or so. Now on Facebook, a genuine expert in power supply regulation design said "that will make the regulator unstable cause it sees every ground current as an error". I'm not expert enough to either verify or refute the above statement, but perhaps my response to his statement could be to recommend connecting pin 1 to chassis through a capacitor. Would the experts on this group please comment on these thoughts? Thanks. I think that this reflector does not permit attachments, so here is a link to my simple block diagram, which is essentially a simplification of Muncy's Figure 1 in his paper if you'd like to refer to the diagram: https://crush.digido.com/?u=pro_audio&p=muncy Thanks in advance, Bob -- If you want good sound on your album, come to Bob Katz 407-831-0233 DIGITAL DOMAIN MASTERING STUDIO Author: *Mastering Audio* Digital Domain Website No trees were killed in the sending of this message. However a large number of electrons were terribly inconvenienced. -------------- next part -------------- An HTML attachment was scrubbed... URL: From jkc7324 at jkc-lab.com Sat Dec 2 10:08:40 2023 From: jkc7324 at jkc-lab.com (John Chester) Date: Sat, 2 Dec 2023 13:08:40 -0500 Subject: [ProAudio] Muncy Revisited: Pin 1 to chassis through a capacitor? In-Reply-To: References: Message-ID: <476f6c96-d086-4ff4-b96e-b6fe75f3d81b@jkc-lab.com> On 12/2/23 12:21 PM, Bob Katz via ProAudio wrote: > Naturally, I started a firestorm on facebook asking a question about > grounding in a headphone amplifier I'm building. A ground loop running through both the headphone amp and the power supply board doesn't seem like a good idea to me.? Here's my revision of your diagram: https://photos.app.goo.gl/gWfABiqTc82mKZLV9 I wouldn't opine on whether the ground loop will upset the power supply without seeing both the power supply schematic and the PC board layout -- assuming it's not a potted blob.? Best to just get rid of the ground loop.? Zero volts to chassis ground should be on the DC side of the power supply, never on the AC side. Zero volts to chassis should occur before the zero volt line gets to the headphone amp, not after.? IMO the most cautious approach to power supply wiring would be to run 3 twisted wires from the power supply to the chassis ground point, where the zero volt line gets connected to chassis, and then continue with 3 twisted wires to the headphone amp board. -- John Chester From lists at richardhess.com Sat Dec 2 12:02:20 2023 From: lists at richardhess.com (Richard L. Hess) Date: Sat, 2 Dec 2023 15:02:20 -0500 Subject: [ProAudio] Muncy Revisited: Pin 1 to chassis through a capacitor? In-Reply-To: <476f6c96-d086-4ff4-b96e-b6fe75f3d81b@jkc-lab.com> References: <476f6c96-d086-4ff4-b96e-b6fe75f3d81b@jkc-lab.com> Message-ID: John, I agree with you, but I think the Pin 1 of the input connector should be connected directly to chassis, preferably using a purpose designed XLR connector if you're going to be in very high RF. I think Neutrik has an XLR connector with Pin 1 bonded to the connector housing which is then bonded to the chassis. The power supply should be grounded to the chassis via a short lead -- remember ohms-per-square? The chassis is a lower impedance than any wire (as long as it's metal. I agree with the twisted triplet from PS to amplifier board. I'm not so certain the chassis connection needs to come from that triplet, but rather a shorter jumper. The output, while not shown, should be a floating jack with a capacitor directly to ground for RF bypass with the common signal lead coming from the amp module. If there are two amp modules (one per channel), that requires additional wiring. I defer to Bill Whitlock in this. I think I'm accurately portraying his suggestions, but there is a lot of stuff on the Jensen Transformers website (or at least there was before it was bought by Radial Engineering). Cheers, Richard On 2023-12-02 1:08 p.m., John Chester via ProAudio wrote: > On 12/2/23 12:21 PM, Bob Katz via ProAudio wrote: > >> Naturally, I started a firestorm on facebook asking a question about >> grounding in a headphone amplifier I'm building. > > > A ground loop running through both the headphone amp and the power > supply board doesn't seem like a good idea to me.? Here's my revision > of your diagram: > > https://photos.app.goo.gl/gWfABiqTc82mKZLV9 > > I wouldn't opine on whether the ground loop will upset the power > supply without seeing both the power supply schematic and the PC board > layout -- assuming it's not a potted blob.? Best to just get rid of > the ground loop.? Zero volts to chassis ground should be on the DC > side of the power supply, never on the AC side. > > Zero volts to chassis should occur before the zero volt line gets to > the headphone amp, not after.? IMO the most cautious approach to power > supply wiring would be to run 3 twisted wires from the power supply to > the chassis ground point, where the zero volt line gets connected to > chassis, and then continue with 3 twisted wires to the headphone amp > board. > > -- John Chester > > > _______________________________________________ > ProAudio mailing list > ProAudio at bach.pgm.com > http://bach.pgm.com/mailman/listinfo/proaudio -- Richard L. Hess AES Life Fellow Aurora, Ontario, Canadarichard at richardhess.com http://www.richardhess.com/tape 647 479 2800 http://www.richardhess.com/notes Quality tape transfers -------------- next part -------------- An HTML attachment was scrubbed... URL: From jim at audiosystemsgroup.com Sat Dec 2 13:04:21 2023 From: jim at audiosystemsgroup.com (Jim Brown) Date: Sat, 2 Dec 2023 13:04:21 -0800 Subject: [ProAudio] Muncy Revisited: Pin 1 to chassis through a capacitor? In-Reply-To: References: Message-ID: On 12/2/2023 9:21 AM, Bob Katz via ProAudio wrote: > Now on Facebook, a genuine expert in power supply regulation design said > "that will make the regulator unstable cause it sees every ground > current as an error". This is the equivalent of loudspeaker I/O, which is also susceptible to RFI, but not hum/buss. The ancient and accepted solution is a bypass cap to chassis, with an RF choke in series with audio return. That's the equivalent of the Neutrik EMC connector. Jim Brown From dlj at josephson.com Sat Dec 2 14:26:16 2023 From: dlj at josephson.com (David Josephson) Date: Sat, 02 Dec 2023 22:26:16 +0000 Subject: [ProAudio] Muncy Revisited: Pin 1 to chassis through a capacitor? In-Reply-To: References: Message-ID: <0d858a68-71d4-4839-8460-9062d79c858c@mtasv.net> Bob, there are already three good answers from John, Richard and Jim -- I don't have much to add except to puzzle about what you would consider 0V on the AC side of the power supply (and why you would want to tie that to chassis.) Ground and neutral are connected at one point only on your premises, as you know. I think your power supply commenter might have been referring to that -- tying the AC neutral to the chassis would be wrong for a bunch of reasons, including potentially injecting ground current noise into the regulator. I have seen another variant of that problem, with filter capacitor ground paths sharing a common path with something else and injecting hum into the other path through common impedance coupling. As Jim said, T-R-S headphone jacks should have the sleeve (audio return) tied to chassis with the same care that you do with the input, but if there is a reason you don't want to do that, you can use a capacitor so at least all the RF gets to the enclosure ground through a short path. David Josephson -------------- next part -------------- An HTML attachment was scrubbed... URL: From crispin at crookwood.com Sat Dec 2 14:37:03 2023 From: crispin at crookwood.com (crispin at crookwood.com) Date: Sat, 2 Dec 2023 22:37:03 -0000 Subject: [ProAudio] Muncy Revisited: Pin 1 to chassis through a capacitor? In-Reply-To: References: Message-ID: <06b801da2570$12731df0$375959d0$@crookwood.com> Hi Bob, Well, there?s theory and practice? In your case, there are two things to consider: 1. There is no such thing as ground 2. You are making essentially a power amp (big current) Point 1) means that you the second you pass any current through a conductor, it ceases to be a ground reference. Point 2) means that you are going to be dumping an amp say (if it?s a decent studio headphone amp) through your ground return ? it really will stop being a ground. Typically if you?ve got a nice metal box, you can treat this as a chassis ground for RFI. Relatively low level currents will circulate around the chassis, and you can tie pin 1 to this OK. The chassis will float to a potential due to the RFI currents, and all of the pin 1?s will mostly be shorted together. Every input can be connected to a differential amp, balanced, or unbalanced, so the inputs are all sorted. However, your power amp is unbalanced. If you connect the return of your headphones to this chassis by shorting the output ?ground? to the chassis, it will all go Pete Tong, because you?re now dumping 1A through the chassis. The current flow goes, 1A through the transformer secondary, ref the 0V of the secondary, via the regulators to the power amp, via the output transistors, via the headphones, back down the return ground. This return ground needs to be connected to two points. 1 to the power amp ground reference differential pin, because the power amp output is referenced to this, and 2 directly to the 0V tap of the secondary, and nowhere else. You can connect the return ground to pin 1/ chassis via a 10N ceramic if you want some RFI attenuation however. Finally, you need to connect your 0V secondary to the chassis at one point, and one point only. This ground partitioning will give you the best RFI and audio performance. Digital grounding is different however. Note mass production often uses ground planes. These are convenient, but do not provide a good analogue ground, especially if you?re dumping any current through them. Hope this helps. Kind Regards Crispin Herrod-Taylor Managing Director, Crookwood www.crookwood.com Tel: +44 (0)1672 811 649 Mobile:+44(0)7910 637 634 Sign up for our great newsletter here! and keep up to date with the audio world From: ProAudio On Behalf Of Bob Katz via ProAudio Sent: 02 December 2023 17:21 To: proaudio Subject: [ProAudio] Muncy Revisited: Pin 1 to chassis through a capacitor? Hi, Pro audio! Happy Holidays from Bob. Naturally, I started a firestorm on facebook asking a question about grounding in a headphone amplifier I'm building. I made a mistake by forgetting to show the connection between the 0 volt (transformer center tap) and chassic ground... oops, so the conversation went downhill and even after I revised my diagram some people were responding to my first diagram with the error :-( I insisted (and still insist) that EMI/RFI interference at the audio input is ameliorrated in Muncy's "rule 2": by connecting pin 1 to chassis directly at the audio input. This is the centerpiece of Muncy's seminal article "Noise Susceptibility in Analog and Digital Signal Processing Systems" I know this from experience as well as theory, having successfully installed a microphone preamplifier on the 101st floor of the World Trade Center in 1995 or so. Now on Facebook, a genuine expert in power supply regulation design said "that will make the regulator unstable cause it sees every ground current as an error". I'm not expert enough to either verify or refute the above statement, but perhaps my response to his statement could be to recommend connecting pin 1 to chassis through a capacitor. Would the experts on this group please comment on these thoughts? Thanks. I think that this reflector does not permit attachments, so here is a link to my simple block diagram, which is essentially a simplification of Muncy's Figure 1 in his paper if you'd like to refer to the diagram: https://crush.digido.com/?u=pro_audio &p=muncy Thanks in advance, Bob -- If you want good sound on your album, come to Bob Katz 407-831-0233 DIGITAL DOMAIN MASTERING STUDIO Author: Mastering Audio Digital Domain Website No trees were killed in the sending of this message. However a large number of electrons were terribly inconvenienced. -------------- next part -------------- An HTML attachment was scrubbed... URL: From bobkatz at digido.com Sat Dec 2 17:09:41 2023 From: bobkatz at digido.com (Bob Katz) Date: Sat, 2 Dec 2023 20:09:41 -0500 Subject: [ProAudio] Muncy Revisited: Pin 1 to chassis through a capacitor? In-Reply-To: <476f6c96-d086-4ff4-b96e-b6fe75f3d81b@jkc-lab.com> References: <476f6c96-d086-4ff4-b96e-b6fe75f3d81b@jkc-lab.com> Message-ID: <831BD386-D6D8-4714-8A97-C04284FEFE89@digido.com> I like it, John. It? also more of a star ground approach and it takes care of the pin 1 issue at the same time! As for the ?AC side? versus the ?DC side? of the 0 V line inside the power supply, the schematic shows it to be a continuous trace from One side of the Board to the other so I?m not sure of the distinction if any unless we?re looking at micro currents I imagine. The two ?0 v? points at the AC entrance of the ps board are just two adjacent screws of a terminal strip tied together. I should have shown them as tied together Bob My iPhone made me write this. Quotes only below this line. Nothing more to see here. ---------------------------------------------------- > On Dec 2, 2023, at 1:08?PM, John Chester via ProAudio wrote: > > ?On 12/2/23 12:21 PM, Bob Katz via ProAudio wrote: > >> Naturally, I started a firestorm on facebook asking a question about grounding in a headphone amplifier I'm building. > > > A ground loop running through both the headphone amp and the power supply board doesn't seem like a good idea to me. Here's my revision of your diagram: > > https://photos.app.goo.gl/gWfABiqTc82mKZLV9 > > I wouldn't opine on whether the ground loop will upset the power supply without seeing both the power supply schematic and the PC board layout -- assuming it's not a potted blob. Best to just get rid of the ground loop. Zero volts to chassis ground should be on the DC side of the power supply, never on the AC side. > > Zero volts to chassis should occur before the zero volt line gets to the headphone amp, not after. IMO the most cautious approach to power supply wiring would be to run 3 twisted wires from the power supply to the chassis ground point, where the zero volt line gets connected to chassis, and then continue with 3 twisted wires to the headphone amp board. > > -- John Chester > > > _______________________________________________ > ProAudio mailing list > ProAudio at bach.pgm.com > http://bach.pgm.com/mailman/listinfo/proaudio From bobkatz at digido.com Sat Dec 2 17:16:39 2023 From: bobkatz at digido.com (Bob Katz) Date: Sat, 2 Dec 2023 20:16:39 -0500 Subject: [ProAudio] Muncy Revisited: Pin 1 to chassis through a capacitor? In-Reply-To: <06b801da2570$12731df0$375959d0$@crookwood.com> References: <06b801da2570$12731df0$375959d0$@crookwood.com> Message-ID: <2F2E1131-4034-42DF-A175-BF99884B3D3B@digido.com> An HTML attachment was scrubbed... URL: From jkc7324 at jkc-lab.com Sat Dec 2 18:07:46 2023 From: jkc7324 at jkc-lab.com (John Chester) Date: Sat, 2 Dec 2023 21:07:46 -0500 Subject: [ProAudio] Muncy Revisited: Pin 1 to chassis through a capacitor? In-Reply-To: <831BD386-D6D8-4714-8A97-C04284FEFE89@digido.com> References: <476f6c96-d086-4ff4-b96e-b6fe75f3d81b@jkc-lab.com> <831BD386-D6D8-4714-8A97-C04284FEFE89@digido.com> Message-ID: <957263d3-c6c9-475d-be4d-f1e16eb5728b@jkc-lab.com> On 12/2/23 8:09 PM, Bob Katz via ProAudio wrote: > As for the ?AC side? versus the ?DC side? of the 0 V line inside the > power supply, the schematic shows it to be a continuous trace from One > side of the Board to the other so I?m not sure of the distinction if any > unless we?re looking at micro currents I imagine. Current through the 0 V trace on the AC side will be capacitor charging current, with a fairly high peak value. Depending on the PC board layout, this may create several millivolts of voltage drop across the trace between the 0 V terminals and the filter capacitors. Thus, there may be significant noise voltage between the 0 V terminals on the AC and DC sides. If the power supply is well designed, the voltage regulators will be referenced to the 0 V trace close to the DC output terminals. The 0 V terminals on the AC side will be noisy relative to the DC side. So, we want to tie the clean side to ground, not the noisy side. -- John Chester