Best grounding scheme

[tubeswell]

In the brass plate scheme, is the plate then soldered to the chassis? If so, at what point(s)?

No its usually just bolted to the chassis with the force of the threads on the control pots and input sockets and switches

[tubeswell]

tubeswell, is that jack then grounded to the chassis directly near by, via soldering a ground wire?

I use sleeve-grounding jacks and I have the pre-amp star grounding at one of the jacks ground lugs. (I have all the input jack's ground lugs also wired to that same jack's ground lug)

[mbratch]

I use sleeve-grounding jacks and I have the pre-amp star grounding at one of the jacks ground lugs. (I have all the input jack's ground lugs also wired to that same jack's ground lug)

Hmm. I'm not sure if I have "sleeve grounded jacks".

[tubeswell]

Hmm. I'm not sure if I have "sleeve grounded jacks".

I mean ordinary grounding jacks where the sleeve contacts the chassis (as opposed to being isolated). The ones I use (switchcraft) a have a ground switch which connects the socket tip to ground when there is no plug in the jack.

[mbratch]

I mean ordinary grounding jacks where the sleeve contacts the chassis (as opposed to being isolated). The ones I use (switchcraft) a have a ground switch which connects the socket tip to ground when there is no plug in the jack.

Ah, got it. Yeah those be what I got.

[mbratch]

Well.... thanks to the inputs from a couple of good people here, my problem mainly appears to be my 12AY7. I replaced it with a different tube (a 12AT7 since I have no spare 12AY7) and nearly all the hum disappeared! I guess it didn't occur to me that it could be a bad tube. There's a negligible but detectable amount of hum remaining even with volumes all down, but it's very, very slight. So my original grounding probably isn't optimal, but it's about 99%.

Thanks everyone for all the input. I'll still be studying this grounding stuff and the inputs have been really helpful in learning.

[defaced]

I use a modular approach in my amps. The power supply, preamp and PI/power amp all have their own turret boards, and they are close to their associated tubes. I ground each board right at the mounting point in the chassis, soldered to brass busses I have soldered into the chassis. The heater winding is grounded right next to the power transformer (also to a buss bar) and the B+ winding a few inches downstream from that. This ground distribution scheme decouples the stages nicely and keeps everything quiet and stable, but I didn't happen upon it overnight. It took a month of solid work to debug it. This is just MY approach, but it might be helpful.

Do you have any pictures of this setup you could share? I'm more of a visual learner, and am working on a modular build that I've been pondering how I'm going to ground. I've got a rough idea as to how I'm going to go about it but am always willing to learn.

[finom1]

Well I just built a 5F2A, and I don't know how similar it is to your build, but MJWB suggested that I ground the preamp (i.e. everything in the preamp including vol and tone knobs, inputs etc - including the preamp-filter cap) to the input ground. The power amp (i.e. the power tube cathode the two power amp filter caps) went to the only other ground along with the HT CT ground, PT shield, heater CT, and Mains ground (on one of the PT bolts). And it works just dandy.

Thank you for sharing your wisdom with everyone!!!

I am just starting out on my first build, so thank you for the details when explaining what is happening. If anyone has Layout Diagrams please post them or send them to me. I am diffidently a visual learner.

Big Thanks to everyone who has contributed to this Grounding Project!!!

[Gingertube]

This is a good reference on grounding, covers safety grounds and signal grounds, with pictures
I generally use bus grounding. The important thing is to make sure there are no power supply return currents flowing down sensitive signal grounds near the front end of the amp.

http://www.valvewizard.co.uk/Grounding.pdf

Cheers,
Ian

[finom1]

Ian, thank you for helping me get started!!!

[Chuck H]

This is a good reference on grounding, covers safety grounds and signal grounds, with pictures
I generally use bus grounding. The important thing is to make sure there are no power supply return currents flowing down sensitive signal grounds near the front end of the amp.

http://www.valvewizard.co.uk/Grounding.pdf

Cheers,
Ian

Need two likes I didn't even visit the link but I don't have to. It's enough that you recognized the posters dilemma and gave him what I'm sure is useful info on this subject (so obscure to many). Helping when you could because you could. Making the forum proud!

[Roberto]

What Gingertube suggested is a must to start with.

I personally use this grounding scheme (yes, bias caps need to be reversed): each part of the amp grounded to the related cap, and the general grounding where the signal to noise ratio is worse.

[finom1]

Thank you for sharing your wisdom with everyone. What type and gauge Buss Wire do you use for your amp builds?

Thanks again!!!

[Gingertube]

Wire type is not critical. I generally just use what I have in either black or green insulation but have also used tinned copper wire (TCW).
I would not use lighter than 22 AWG, perhaps 16 AWG at main filter caps and rectifier where circulating currents are highest.
Cheers,
Ian

[R.G.]

The key to grounding is to remember that (1) V = I*R and (2) all wires are resistors.

I've counted numerous things that "ground" is used for. It can be an RF shield, a return current wire, a reference source of 0.0000V, a "sewer" return of used electricity back to the power supply, and others. What matters is to remember that the only way to have zero volts across a real world wire is to have the current through it be zero. So if you want a ground to really be at zero volts and make that zero volts available at some distance away, you can't put any current through the wire carrying the "zero volts" because if you do, it won't be zero volts when it gets there.

This all gets mishmashed up because it's common for people who designed classic era equipment to be ignorant of the issues, or to ignore them in favor of getting cheap and good enough.

So let's think about the chassis. Big, thick metal, ought to be good for grounding, right? Wrong. You critically, critically want the chassis to be an RF shield. You don't want to run currents through it, especially speaker return currents or heater return currents, or AC currents, because those currents force the voltage at different places on the chassis to be different.

Input jacks can be connected to the chassis, right? Maybe. If the chassis has currents flowing through it, the input jack "ground" voltage is wobbled around and that is fed right to the input circuits and amplified. Well, how about the input jacks being the only connection to the signal ground? That's OK as long as you don't care about three wire ground and safety ground and don't let the output jack contact the chassis either. Ideally, the chassis would be connected only to the third wire safety ground and perhaps to input jack ground by small RF caps, and the input jacks isolated from chassis. Now no current flows in the chassis and no voltage is developed across it to be amplified. Notice that reverb sends and returns and effects loop sends and returns have the same issues, but may have a lower sensitivity because their signal level is higher.

What about controls? Their "ground" contacts can be simply soldered to the control bushing/enclosure, right? Again, only if the remaining currents in the chassis are very low, because they then become a place to inject any chassis AC voltage in the amplifier.

What about power supply ground? That one you do want connected to chassis, but you have to be careful about where on the ground net you tie the wire. Most effective is the negative lead of the first filter cap.

And that first filter cap negative lead. That's where the CT of the PT goes, nowhere else. Otherwise, the pulse currents from the rectifiers will jerk "ground" around and you'll never get rid of the 120Hz hum.

How about circuit grounds from the various amp sections and decoupling caps? You have some choice here, at least. Ideally, power ground from each circuit section would have its own ground wire back to the power supply, so current from other stages doesn't travel back down that wire, producing an offset voltage on the wire to be amplified. The decoupling caps? They do two jobs: one of which is to decouple signal from the power supply voltage because the typical tube amplifier stages have zero rejection of any power supply noise, and also the decoupling of any ripple on the power supply. The closer to the power stages you get in the typical cascade of decoupling you get, the more this is decoupling power supply ripple, turning the 120hz sawtooth voltage into 120Hz sawtooth current headed back to the power supply ground point. The closer to the input circuits you get, the more the ripple has already been smoothed, and the more you want those decouplers to shunt any AC currents from the amplifier stages sucking current back to the power supply without elevating the local ground reference. Both of these would be best served with a separate ground wire for the decoupling cap, but near the preamp end you can get away with joining the decoupler ground to the local circuit ground.

This just goes on and on. To design grounding schemes, you have to know what kinds of currents flow toward the power supply ground, and then make your grounding scheme carry it in a way that injects the least objectionable amount of that current noise into your voltage amplification.