Why terminate circuit ground(s) to chassis, rather than directly to mains ground point?

[Peegoo]

The goal here is not more power/volume, but mostly for quietest operation. In the simplest terms, noise gets into the signal anywhere you allow it to, which means every little thing you do to help prevent that really matters. Noise in a signal is cumulative; it's like having leaky windows on a house: with little air leaks in every window, it's the same as having one window all the way open all year long.

Another thing I do is test all of the non-electrolytic capacitors I'm going to use to ensure proper orientation in the circuit. Polarity does matter, even with "non-polarized" caps. Used to be, manufacturers would mark their caps with OUTSIDE FOIL on one end, and/or include a solid bar marking to indicate which conductor was connected to the foil wrap closest to the surface of the cap. Some makers still mark their caps with a bar, but these markings cannot be trusted because I've found a few that were incorrectly marked.

This is important to know because when the cap is installed with the outside foil oriented toward either ground, source or lower-impedance side, it is shielded in the same manner as a coaxial cable. This concept applies to all non-electrolytics: orange drops, green and brown 'chicklets', axials...you name it.

With all these caps installed correctly, your builds will be a whole lot quieter. It seems like a lot of trouble to go through, but believe me it is worth the extra effort.

More info here:

 

[gabasa]

You know, after crashing so many build threads (builds that would have turned out perfectly fine regardless) about how returning current on the chassis is sooo baaadd... we're finally talking about it!!!

And now you are mocking folks for not picking up the concepts right away?

This seems antithetical to your mission.

This thread was started Monday. It's now Wednesday. How long have you been studying electronics? Have some patience with folks why don't ya?

Respectfully, I'd be easy on 2L man because he's made almost 150 posts regarding return current over the last two years, and it's understandable if he's showing frustration for the very first time. In almost 100% of cases, the advice in his posts has not been taken and I think that, with the best of intentions, his only point is that we're collectively making something very simple, very confusing by adding a lot of clutter.

 

[gabasa]

I must confess that I do not fully understand what @gabasa is talking about ^^^here^^^. It seems to have something to do with this thread, which I missed until just now:
https://www.tdpri.com/threads/another-amp-re-grounded.1113580/
but I see in his photo that he's got grounds collected at the PT center tap, using that for ground reference rather than the mains ground connection to chassis.
**This thread seems to be an approach to answering my question above "Why not pretty much eliminate that entirely?" -- but it's not clear to me how this "solves" the issue. As an solution, it certainly seems to be an "outlier", not one I've seen before, and the discussion there regarding being shocked in a club venue is somewhat alarming.

With all due respect, I'd like to clarify one thing that was stated in the OP, as well as highlight how confusion comes into these threads.

In 26 years of modifying and building amps as a hobby, working on it almost daily for fun, one amp on one occasion caused a shock in a nightclub. This happened a little while ago and it sent me down a rabbit-hole of learning how to setup my power supply wiring and grounding better. I read Aiken's website, Merlin's chapter on grounding and 2L man's posts for at least a week. Once I got to the point in which I could apply what I read to the best of my abilities, and could visualize the current paths better in my amps, I felt like it was a great learning moment for me. I implemented this all into that amp, took it back to the nightclub and the shocks vanished. Problem solved. I created that thread to share my new grounding scheme to anyone interested in seeing how I implemented it.

The original post would have you believe that these methods caused a shock issue, while in my thread, I actually wrote the exact opposite.

 

[2L man]

You know, after crashing so many build threads (builds that would have turned out perfectly fine regardless) about how returning current on the chassis is sooo baaadd... we're finally talking about it!!!

And now you are mocking folks for not picking up the concepts right away?

This seems antithetical to your mission.

This thread was started Monday. It's now Wednesday. How long have you been studying electronics? Have some patience with folks why don't ya?

Yes you are right. I should have ask first to define the "ground". I have wrote my definition many times. Often I have wrote instructions when it has been asked. Also sometimes when build was planned or on stage when it would be easy to fix. I have received positive FB and no negative which does not say there isn't cases it did not help.

There was a smiley too but I hope it give another way to think and see the dilemma because return current circuit is much much simpler than the feed circuit.

I have now test "split ground" vs single bus or Star in three amps and all have hum bit more when pre amp and PI current flow thru chassis. Test can be done installing SPDT switch so comparing is fast just hitting the switch.

The use of chassis has been forbid here about 50 years on imports and new builds because it can spread noise to mains.

I began studying electronics 1978 but I had hobbied it few years before. Then electronics was my work over 41 years until it stop two months ago.

 

[2L man]

I always respect your insights, @2L man . Of course you're right. Merlin says the same thing, with a practical coda:

“Although the terms ‘earth’ and ‘ground’ are often used interchangeably, the audio circuit ground does not necessarily have to be connected to planet Earth. The entire amplifier circuit could be built ‘floating’ inside the metal chassis, with no connection to the chassis at all. However, in reality the circuit will be connected to chassis at some point since this ensures the amplifier’s working voltages are properly defined with respect to zero volts, and that the chassis acts as a shield against electric fields.”

Isn't Merlin european, Brit I recall? I have read most content in his pages and it looks we have studied same electronic material but he have more I believe european electronics student whom has not been teached this is close to 70, possibly older?

 

[chas.wahl]

May I reframe my original question somewhat, to help (I hope) illustrate what prompted me to start the thread? Consider two cases:

a) that where the chassis is made of some insulative material, like plastic or fiberglass reinforced something-or-other; and

b) that where the chassis is, in fact, entirely copper.

How would circuit ground reference be effected (meaning made, not affected) in case (a)?

How would case (b), with a single ground bus terminated at/near input(s) be any different than what I originally suggested: direct connection of circuit ground to mains ground point? I will grant in advance @Peegoo's remonstration against a long wire creating an antenna, versus an entire Faraday-like enclosure in copper, as being a difference. But not the only one?

 

[King Fan]

How would case (b), with a single ground bus terminated at/near input(s) be any different than what I originally suggested: direct connection of circuit ground to mains ground point?

Re plastic, my drill's power cord has only two prongs, nonpolarized (ok, there’s a bit more to that setup than plastic, but….)

Solid copper? Now if your chassis was pure silver, which is way more conductive than copper… (checks Magic 8 Ball)… nope, still not a thing.

Heh, I’m not looking up more NEC code. Still, I’m pretty sure it's not just Merlin and Aiken who condemn anything else sharing the mains ground. It should be clear by now I'm no grounding guru, but the idea the safety ground be *mechanically* and electrically separate seems to imply something more about safety engineering than simple ground loop theory.

By the way, can you tell us why you’d want to route the signal ground to the mains anchor? Against code, long wire, etc…. downsides. Any upsides?

Still, maybe there's more to the separation than safety. I don’t know. Line noise going back to the household ground? Something else? I’ll stop being facetious and see what we learn.

 

[Lowerleftcoast]

Consider two cases:

a) that where the chassis is made of some insulative material, like plastic or fiberglass reinforced something-or-other; and

b) that where the chassis is, in fact, entirely copper.

I am not sure I am understanding these questions because I think they have been adequately answered in this thread.

In case a) the non-conductive chassis ideally will be shielded, just like a guitar body, with conductive paint or tape. All we are talking about here is shielding. This has nothing to do with pulsed DC current or B+ DC current.

In case b), the guidelines for the "mains ground point" fastening are clear. Due to hopefully maintaining a good connection... vibration proof... bomb proof. This connection will be fastened all by itself. Now you can place the ground reference right next to this fastener if you please, but connecting directly with this fastener has been deemed not good practice because it might make the connection less secure.

Positioning the reference near the input has been suggested by Blencowe due to the possibility of creating larger than necessary ground loops when other AC powered equipment is connected to this amplifier. Again the possible ground loops have nothing to do with pulsed DC current or B+ DC current.

 

[mcentee2]

May I reframe my original question somewhat, to help (I hope) illustrate what prompted me to start the thread? Consider two cases:

a) that where the chassis is made of some insulative material, like plastic or fiberglass reinforced something-or-other; and

b) that where the chassis is, in fact, entirely copper.

How would circuit ground reference be effected (meaning made, not affected) in case (a)?

How would case (b), with a single ground bus terminated at/near input(s) be any different than what I originally suggested: direct connection of circuit ground to mains ground point? I will grant in advance @Peegoo's remonstration against a long wire creating an antenna, versus an entire Faraday-like enclosure in copper, as being a difference. But not the only one?

Great thread,I think your question is good too!

With a non conductive chassis, where does your 0v reference come from ? The thing that would have been connected to the chassis near the input jack we are all talking about.

Without a zero reference, that amp circuit could float around and maybe get out of control or rise up to some not very safe level Vs external (I think that's what Merlin says somewhere?)

I agree re separating the safety ground from this 0v reference, but in a non-conductive chassis the only places are either the safety ground (via some form of conductive connection, or back directly to one of the input jack ground points.

Does that make sense?

 

[Peegoo]

May I reframe my original question somewhat, to help (I hope) illustrate what prompted me to start the thread? Consider two cases:

a) that where the chassis is made of some insulative material, like plastic or fiberglass reinforced something-or-other; and

b) that where the chassis is, in fact, entirely copper.

How would circuit ground reference be effected (meaning made, not affected) in case (a)?

How would case (b), with a single ground bus terminated at/near input(s) be any different than what I originally suggested: direct connection of circuit ground to mains ground point?

Ref a: Ground reference would not be affected or effected--there is a difference--good observation. Electrons have no dog in the fight when it comes to conductors and insulators; they are simply going to 'go home' via the easiest (least resistive) path. I've built an amp using an ancient wooden cigar box as the chassis, and I didn't shield it with foil or screen; it was quiet, until placed close to other electronic equipment.

Ref b: By doing it this way I think it would work the same way as direct grounding to the chassis, but with the addition of more wire. There may be differences in DC ground potential between various portions of the circuit, and that can create ground loops (current backflow via ground). I think it gets even more complicated in a circuit that contains AC and DC currents because they are in close proximity to each other and noise can be easily induced in a signal path. Add gain to the equation and noise mitigation becomes even more critical; you have to find a way to make it as quiet as possible. Something as simple as changing the orientation of the OT in relation to the PT can make a huge difference in noise in the signal. Then you plug in a guitar or bass and all bets are off.

Every time I do a build I try different things to reduce noise. I am not an expert on any of this stuff, but I am a safety freak and I like amps to be a quiet as possible without strangling 'em with power filtering that's too stiff. It's an ongoing learning experience. This thread is a good one and has given me other ideas to try.

 

[sds1]

I will grant in advance @Peegoo's remonstration against a long wire creating an antenna, versus an entire Faraday-like enclosure in copper, as being a difference.

At the same potential as the chassis, the wire is part of the shield.

 

[Peegoo]

Here's a good article on grounding types.

 

[chas.wahl]

That article is interesting, but mainly tells me how shallow my understanding of grounding is, still. My favorite part is (describing separation/interaction of different grounds): "The girl's emotional fluctuations will not affect the boy's mood, just as they will not affect the mood of a pair of strangers before they become lovers, because they do not have any intersection."

Is "Rose" an AI bot (so much in the news these days)? The syntax strikes me as slightly stilted.

I'm going to guess that Merlin Blencowe's single ground end-to-end is a Type 4 approach, that a Doug Hoffman split ground (as also espoused by @robrob in his pages, at least erstwhile) is a combination of Types 4 and 5, and that leaves the Fender / @Peegoo "ground everything as you go, right to the chassis immediately adjacent" style of managing the "emotional fluctuations" and their intersection (the problems outlined in Part III), in the context of each ground having a "certain grounding resistance" -- which I suppose is both a good thing, and something to be concerned about getting right.

 

[King Fan]

Is "Rose" an AI bot (so much in the news these days)? The syntax strikes me as slightly stilted.

Good point, sir. ‘Stilted’ is being kind. But I see that website belongs to a firm in Hong Kong, so I think this may have been written before ChatGPT by someone for whom English was a second language.

Wih this kind of web content, how much is original and how much is a retread, in less clear terms, of some original source? The embedded YouTube on that web page, which sounds more articulate, is still edited in weirdly repetitive chunks. But it happily leaves out all the digital/analog stuff and does at least emphasize key ideas relevant to amps:

1. The terms and symbols for ground and earth are ambiguous
2. The safety earth ground is critically different from circuit ground
3. Schematics often fail to show circuit ground location
4. Circuit connections are critical regarding ground loops and noise

 

[2L man]

Well, for me the ground reference {<< edit} aspect is less clear, because it's at the end of the circuit, and seemingly passive, rather than being what I can understand as being the "active" part of the circuit, with the PT providing the motive force.

If you like, I could start another thread about how to calculate/predict voltage drop when both a choke and a dropping resistor are used (or resistors in both B+ and ground buses, as in Merlin Blencowe's suggested "balanced" reservoir filter, q.v. Grounding fig. 15.7) ;-)

Also, if irritated, you could simply stop reading the thread.

Actually I think when return current circuit is compared to feed current circuit it is easy to understand how simple it is and that the simpler it is the better it function.

On Star grounding instructions it is mounted to chassis and placed close the power supply and often it is a screvable lug which can have even eight "beams" but perhaps six beam star lug is more common. When the Star is connect to chassis all input jacks must be insulated and no parallel return current paths come.

I should not matter where the secondary zero volt reference "ground" is connect to chassis/safety earth. For example amplifier operate fine even it is not but for safety reason it is often connect. Or "ground loop hum eliminator circuit" is used which does this SE connection "softly". It is needed to burn HV fuse if HV leak to chassis.

I have seen tube amp schematics which use HV positive as the reference which is connect to chassis. They must have something special and cathodes as a reference is more convenient when signal is often fed to control grid which potential is much closer to cathode than anode.

151st time the idea is not to use chassis for operative current

Gabasa said I had wrote it alresdy 150 times I think there must be more? And I know You knew it.

In European regulations mains Safety Earth must use own lug which might be crimped and mounted usind threaded method. I think it is the same in USA as well?

I have used filtering resistors on return current now perhaps at least in five builds, perhaps more? Values selection follow easy ohms law when supply voltage and target B+1 are known. Yet I have changed few resistor values on few builds when PT output did not drop as I had predict. Resistors also increase sag-effect.

 

[Tele_Tom64]

Geez. I'm trying to read this thread but you guys are killing me with all of the abbreviations. HT? OT? PT? ETC...

We all haven't been playing guitars for 30 years so some of us don't know this stuff. You guys might think about writing out the word(s) the first time you use them in a post and then put the abbreviation(s) next to them in parentheses for the benefit of others who use the forum, also. That way we can all learn from tech threads such as this one. New people like me would really appreciate it

 

[gabasa]

Geez. I'm trying to read this thread but you guys are killing me with all of the abbreviations. HT? OT? PT? ETC...

We all haven't been playing guitars for 30 years so some of us don't know this stuff. You guys might think about writing out the word(s) the first time you use them in a post and then put the abbreviation(s) next to them in parentheses for the benefit of others who use the forum, also. That way we can all learn from tech threads such as this one. New people like me would really appreciate it

Here are some of the acronyms in this thread:

PT: power transformer
OT: output transformer
HV: high voltage
HT: high tension
CT: center tap
AC: alternating current
DC: direct current
RF: radio frequency
EMI: electromagnetic interference
NVM: nevermind

 

[wangdaning]

@Tele_Tom64
To add a few more.
PI: phase inverter
MV: master volume
TMB: treble, middle, bass

Always feel free to ask about any of them if you do not know. Nobody here is the type to laugh at people, we can laugh together though if you later realize something was obvious

 

[Scottcurry13]

HT is also referred to as B+ at times.
PPIMV is post phase inverter master volume.

 

[King Fan]

Geez. I'm trying to read this thread but you guys are killing me with all of the abbreviations. HT? OT? PT? ETC...

We all haven't been playing guitars for 30 years so some of us don't know this stuff. You guys might think about writing out the word(s) the first time you use them in a post and then put the abbreviation(s) next to them in parentheses for the benefit of others who use the forum, also. That way we can all learn from tech threads such as this one. New people like me would really appreciate it

A reasonable reaction; we've all been there. And it only gets worse. HT (high tension) is also sometimes called HV (voltage), which may often be distinguished from the DC (direct current) B+ (battery positive, which isn't from a battery), and its subsidiary B+1, 2, 3, etc... all of which may be lumped as the "hot rail" in casual speech.

But take a step in the moccasins of the average TDPRI poster. You start to find you're writing multiple posts per day about PTs, OTs, and CTs (center taps). Spelling 'em out gets old really fast. Now consider the next morning, when you try to add a helpful note about the HT. Did you spell that out *in this particular thread* yesterday? There's a reason technical fields are nostril-deep in acronyms.

Here's a page that may help you get started. Not exhaustive, but a boost onto the ladder. Nostalgia; I bookmarked this my first week doing amps.

http://www.aikenamps.com/index.php/a-glossary-of-common-amplifier-terms

And I can't leave without noting that though this thread interests us nerds no end, 95% of the discussion is about stuff that doesn't make a difference in 95% of the builds you'll see here. Idealized ground theory is nice to learn about eventually, fun to argue about in the bar between missions, but neither easy nor vital to *fully* understand. If you take away the idea that US and EU codes require the household safety ground/earth to be mechanically and electrically separate from any other connection, you'll live long and prosper. When you add the fact the same codes suggest a bombproof mechanical connection to the chassis, then, like a beginning pilot, you can move on from ground school...