PCB ground planes

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Some examples of 'local star ground':
Fig 19 in Crowhurst 1951: http://ken-gilbert.com/images/pdf/ch2345.pdf
Fig 15/2, p.172 in Briggs & Garner, 1952 : http://www.tubebooks.org/Books/Briggs_amplifiers.pdf
Fig 15.14, p.274 by Merlin Blencowe : http://www.valvewizard.co.uk/Grounding.pdf

For noise currents from power transformer secondaries via parasitic capacitance, I note the 1944 WW p.15 had an article on screening with measurement results:
https://worldradiohistory.com/UK/Wireless-World/40s/Wireless-World-1944-01.pdf
 
I thought I understood grounds, but now I am learning vocabulary from Abbey... I suspect "moronic" may be harsher than Abbey thinks it is, Abbey is the kinder, gentler, mod.

JR
 
Some examples of 'local star ground':
Fig 19 in Crowhurst 1951: http://ken-gilbert.com/images/pdf/ch2345.pdf
It shows a perfect example of hierarchical ground, which faithfully follows the signal path, not star.
I agree with the sentence:" It is imperative that all "earthy" points shoud be bonded to chassis at one point only", which does not mean that these "earthy points should be bonded via individual wires, which would then create loops through which magnetic induction would create voltages that would interact with the desired signal..
It just means that the "ground" connection must not make loops.
Fig 15.14, p.274 by Merlin Blencowe : http://www.valvewizard.co.uk/Grounding.pdf
Most of the figures are perfect illustrations of hierarchical ground. Notable exception is the description of a two-channel amp (fig 15-17). However, in practice, for optimal performance, the layout must obey the rules of hierarchical ground, i.e. locate the low-level sensitive stages close to each other. In other words "signal follows ground".
I'll leave the responsibility of questionable terminology (multiple star) to Mr Blencowe.
For noise currents from power transformer secondaries via parasitic capacitance, I note the 1944 WW p.15 had an article on screening with measurement results:
https://worldradiohistory.com/UK/Wireless-World/40s/Wireless-World-1944-01.pdf
Interesting document, but it relates to hertzian transmission of RFI, which is unrelated to the subject at hand.
 
Last edited:
Etching copper off PCBs consumes reactive agents in the etchant. I've shared this before but PCB layout artists at Peavey were advised to leave as much copper on the boards as practical.
==
For some high current power amp traces, the solder mask was left off so extra solder would accumulate on the traces to lower the resistance.

JR
I remember when it was solder mask over reflowed copper. Then they went to SOBC and took my current away... I have a PCB house in china that doesn't charge extra for 2OZ copper...
 
It is not that esoteric, certainly very common terminology among tube amp builders.
One thing to keep in mind about the mystery of the "star ground". Plan out your grounding before just starring it to one place.

Many times, The PSU ground isn't the best place to start.

When doing a mic pre, bring the ground into the module to the point where it meets the transformer and phantom circuit, then ground it there.

When summing, bring the + input to the bus ground and ground it there.

And, for studios, don't bring everything to the ground bar. It may make a cool rock story, but you should source the grounds to the console or the DAW interfaces, then one wire to earth.
 
Some examples of 'local star ground':
Fig 19 in Crowhurst 1951: http://ken-gilbert.com/images/pdf/ch2345.pdf
Fig 15/2, p.172 in Briggs & Garner, 1952 : http://www.tubebooks.org/Books/Briggs_amplifiers.pdf
Fig 15.14, p.274 by Merlin Blencowe : http://www.valvewizard.co.uk/Grounding.pdf

For noise currents from power transformer secondaries via parasitic capacitance, I note the 1944 WW p.15 had an article on screening with measurement results:
https://worldradiohistory.com/UK/Wireless-World/40s/Wireless-World-1944-01.pdf

Many thanks for the effort to explain something that some do not want to see. I learned the hard way that hierarchical grounding is problematic in designs that require exceptional power decoupling of each stage. It was the design of an extremely sensitive amplifier of seismic vibrations (generated in the audio band). The solution was to use special return buses for the decoupling capacitors of each stage, so the grounding ended up looking like tree branches. Very similar to the form of star grounding. After that experience, I no longer used the usual hierarchical grounding for any project.
BTW, have you had any experience with power transformers that have an EMI protective winding between the primary and secondary? A friend who pays attention to details asked a transformer shop to make him a power transformer on which the first winding is for heating the vacuum tubes, then the EMI winding, then the primary, then again the EMI winding, then the winding for the anode voltage +B. The preamp that was powered by that transformer sounded very good.
 
The solution was to use special return buses for the decoupling capacitors of each stage, so the grounding ended up looking like tree branches. Very similar to the form of star grounding.
Really? Decoupling capacitors referencing to a point which inductance is significant? And distance/impedance between this point and the current return node?
BTW, have you had any experience with power transformers that have an EMI protective winding between the primary and secondary? A friend who pays attention to details asked a transformer shop to make him a power transformer on which the first winding is for heating the vacuum tubes, then the EMI winding, then the primary, then again the EMI winding, then the winding for the anode voltage +B. The preamp that was powered by that transformer sounded very good.
I don't doubt it sounded very good, but can one attribute that only to the power transformer? Wouldn't traditional electrostatic shields have similar performance?
Is there any objective comparison with a more conventional transformer?
 
I recall seeing some papers on isolation transformers, and especially for medical equipment, where managing parasitic capacitance between primary and core and secondary was a key issue, and have seen some lab instrument schematics with what would appear to be extra effort in the use of screening between winding sections (but unless one was in the manufacturers design group you wouldn't come across testing or design details). Often vintage power transformers for audio would come with one screen winding (but I wasn't able to confirm that this was more related to enhanced protective earth implementation that was later included in country standards for power transformers, and now typically called an Earth Screen or ES). The quandary for a PT with just one screen is where to link it, as there are two obvious forms of noise to manage - from the mains and from a ss rectifier.

Smps design to comply with CM egress on the mains often goes in to detail about minimising or managing or screening windings to suppress parastic currents caused by capacitive coupling to chassis. The voltage on the secondary winding of a psu PT being rectified by capacitor input filter and using ss diodes is anything but a nice sinewave, with high frequency content transferred through from the mains supply, and diode turn-off transient glitches that can easily permeate to nearby circuitry through other means than direct pcb copper traces.

Imho, it's worthwhile being aware of all forms of noise and hum propagation and not just pedantically focusing on one mitigation technique alone.

This link from 1934 raises an awareness of grounding loops that obviously wasn't well appreciated by those using chokes in power supply filtering: https://dalmura.com.au/static/Note on a cause of residual hum in rectifier filter circuits 1934.pdf.

If I was making layout suggestions to the pcb layout in post #37 I'd suggest taking cathode resistors like R12 separately to their respective stage decoupling cap neg pad like C30, and in particular for routing R26,27,32,33 0V (separatel to other 0V) as they locally loop the highest level of signal current.
 
I recall seeing some papers on isolation transformers, and especially for medical equipment, where managing parasitic capacitance between primary and core and secondary was a key issue,
For medical equipment, there are two aspects, safety and hum rejection, which are dealt with reducing the inter-capacitance between primary and secondary, typically less than 120pF..
Often vintage power transformers for audio would come with one screen winding (but I wasn't able to confirm that this was more related to enhanced protective earth implementation that was later included in country standards for power transformers
Shields between windings existed long before safety standards. Designers were aware of the need to keep mains-induced noise from entering the unit, not so much about preventing conducted emissions.
Most power amp transformers had no shield between windings, because it reduces the efficiency, which was not a problem with low-power electronics.
Smps design to comply with CM egress on the mains often goes in to detail about minimising or managing or screening windings to suppress parastic currents caused by capacitive coupling to chassis.
That's why most smps xformers have a shielding winding. The reason it does not use copper foil is because it would generate too much eddy currents, which would decrease efficiency and increase heat.
Imho, it's worthwhile being aware of all forms of noise and hum propagation
Of course.
and not just pedantically focusing on one mitigation technique alone.
It's important to know what technique addresses which problem. Electrostatic shielding does not solve circulating-current problems.
This link from 1934 raises an awareness of grounding loops that obviously wasn't well appreciated by those using chokes in power supply filtering: https://dalmura.com.au/static/Note on a cause of residual hum in rectifier filter circuits 1934.pdf.
The flaws of the "choke in the negative" are so obvious one wonders who this article was destined to. Certainly, most proper designers of the era had perfect mastery of this subject.
If I was making layout suggestions to the pcb layout in post #37 I'd suggest taking cathode resistors like R12 separately to their respective stage decoupling cap neg pad like C30, and in particular for routing R26,27,32,33 0V (separatel to other 0V) as they locally loop the highest level of signal current.
There's no question about it. Returning currents to the local reference point.
As a perfect counter-example, a vintage Fender amp has all its reservoir and decoupling caps in a single box, with wires distributing.
Fender got away with this by using high-gauge wires (relative to current) and accepting mitigated noise performance.
 
As a perfect counter-example, a vintage Fender amp has all its reservoir and decoupling caps in a single box, with wires distributing.
Fender got away with this by using high-gauge wires (relative to current) and accepting mitigated noise performance.
Not to mention adding that huge angled brass corner stock bus bar that was run along the pots and the front of the chassis. :)
 
Tell me how circuit A can be better than circuit B.
The current drawn by each stage create voltages between stages. Signal reference has to go through these small resistances and colect these voltages. With hierarchical ground these voltages are minimized by creating a very short and direct path for signal. It's been long known as "signal follows ground".
I believe you mean the reservoir capacitors currents? Where do they go? Read Kirchoff. They don't need a ground for closing the loop.

What mistake?

Profound misunderstanding. DC (as well as AC) has no home. Current exists in a loop. In taht respect, all points of teh loop are of equal importance. Calling one ground, or reference, or whatever, is a useful convention for circuit analysis.

Hierarchical ground, of course.

Doesn't make sense to me. Because of resistances introduced by wiring, and circulating currents, nothing is a "zero point". It's an abstraction, but when dealing with noise you can't rely on t, you have to consider all parasitics.

That's why chassis should never be used as an "audio ground". And for teh same reason, a dedicated ground should not make loops.

Correct. Same for many guitar amps, because guitar players have always been used to hum and buzz. The balance between the bonus of extra volume and the nusance of noise is still positive.

I don't need to research it. As I explained in another post, I used it to carry an unbalanced signal with an impedance of 200 ohms across several kilometers of cable.

Of course, currents should be separated as much as possible.

There's no doubt DC heaters are the best choice for low noise. It depends on the expected level of noise performance.
However I've seen poor implementations of DC heaters where incorrect filtering resulted in increased buzz compared to AC heaters. 50/60 cycles doesn't coupled too well capacitively, but their harmonics do.
As I said, I've had very good results with star grounding in audio amplifiers, all analogue, no digital signals.

Perhaps my explanation isn't clear.

It comes down to this for me. Treat the unbalanced input ground as the quietest ground point in the circuit. By referencing my HV regulators to that ground the dc from the regulator is as quiet as the input, assuming the regulators work. Mine do.

The star point is connected to the PS via a fat trace for lo z return.

Each successive stage (2 more, a gain then a follower) has their signal ground attached to the star which keeps the amplified signal referenced to signal ground. Returned dc from the stage may go back to the PS ground or the star ground depending on the current.

The output stage, power supply and it's filter bank of my hybrid amp 200 wpc mosfet amp live on a separate board, thus charging and output ground return currents never see the star point. There is a connection between the output stage ground, with no charging currents, and the star ground.

There is little to no signal at the output ground terminal referenced to the input ground. The point being there is no modulation of DC return paths with AC signal current.

The safety AC line ground is isolated by a 10k resistor in parallel with 2 6 amp diodes in parallel in opposite directions which bleeds off tiny ac leakage currents under .6 volts. If there should be a short to the chassis the diodes will conduct and blow the fuse or trip a house breaker. So the chassis is at signal ground and will connect with earth ground in case of a fault. Meanwhile there is 10K isolation between signal/chassis ground and AC line ground reducing or eliminating ground loops.

The overarching theme is to be conscious of the nature and magnitude of the ground currents and where you want them to go and how that will affect circuit performance.

Critique all you want. Stand up on a soap box ll you want. Doesn't bother me.
 
Each successive stage (2 more, a gain then a follower) has their signal ground attached to the star which keeps the amplified signal referenced to signal ground.
How do you mitigate the fact that the current drawn by each of these stages develops along the wire that goes to the star point, which result in differential voltage between stages?
Replace your wires going to the star point with resistors, you will see what I'm talking about.
If mixers used star ground, there would be a lot of (bad) x-talk.
Modular systems use star ground, because of the large number of "grounds". They resort to differential transmission of signal for eliminating these parasitic voltages.
Indeed, many systems operate with currents low enough and conductors sturdy enough to make this negligible.
 
How do you mitigate the fact that the current drawn by each of these stages develops along the wire that goes to the star point, which result in differential voltage between stages?
Replace your wires going to the star point with resistors, you will see what I'm talking about.
I installed a cooler in the amplifier to keep it at absolute zero. No Temp, no R. ;0)!!!

Do you think I don't know that? Your tone is starting to irritate me.

A few ma on a few milli ohms of gold plated 2oz Cu doesn't introduce much of an error voltage when we're talking voltage swings of 80 volts rms.

The whole point is to respect the unbalanced input ground as the lowest least noisy part of the circuit and keep error voltages isolated to their stage. In small gear it's easy to do and you know what is riding on what where because zero is the input ground star. I think I explained that pretty well.


If mixers used star ground, there would be a lot of (bad) x-talk.
Modular systems use star ground, because of the large number of "grounds". They resort to differential transmission of signal for eliminating these parasitic voltages.
I didn't build a mixer. I'm talking about my amplifier. 3 stages in 2 tubes that are 2 inches apart. Not a desk that's 4 feet long. that's a whole different problem. Obviously needs a different solution. The OP was talking circuit boards. As you say, pro mixers are usually balanced in to out keeping audio off the ground.
Indeed, many systems operate with currents low enough and conductors sturdy enough to make this negligible.
Exactly.
3 stages that are 2 inches apart, 2 oz gold plated copper conductors.
A few ma on a few milli ohms doesn't introduce much of an error.

Still have a problem with that?
 
@abbey road d enfer

Just because it may not work well in every situation does not justify excoriating it on a general basis, as you appear to do, IMO.

BTW, I received 2 phone calls today praising my designs and the listening pleasure it brings them and asking do I have any more upgrades. How about you?

I never advocated for the entire audio universe, just stated what works well in my gear. Which, given the OP was posting about a mic pre which probably won't be larger than 4-5 square inches, and would be a similar situation, in my opinion.

See what I did there. I took ownership of my statements as I have all along. You may disagree but try to find a more respectful tone for discussion.
 
Last edited:
Back
Top