Single Ended Master Volume Hum

[loudthud]

Bass and treble controls have no effect on the hum, only the Master.

Your explanation makes sense, now why is the hum entering there and what do I do about it? The grounding of the MV is not the issue, I've moved it all over the chassis with no effect. Wire from the MV to the power tube is shielded.


Hugo

There has to be a difference between the ground for the MV and the ground at the KT88 cathode resistor. You may be able to measure it with a DVM set to AC volts. It's probably only .001 or .002 VAC.

[Hugo]

There has to be a difference between the ground for the MV and the ground at the KT88 cathode resistor. You may be able to measure it with a DVM set to AC volts. It's probably only .001 or .002 VAC.

First of all, let me say that I really appreciate you guys trying to help me out.

If I seem to always come back with "that's not it", I'm not being a smart ass...


So... that's not it. No discernible AC voltage difference between the MV ground and any other ground in the amp. Just to try, I connected the MV ground to the output tube cathode ground. No difference.


Am I too critical? With reasonable filtering and no choke I get about 55mV of hum peak to peak at the output. Is that "normal" for a SE tube guitar amp?

I still do not understand how I'm getting 60Hz hum that is not coming from the heaters, why MV settings affect the hum level, and why no changes in the grounding scheme or transformer placement make any difference at all.


Hugo

[Steve Conner]

Main filter and HV CT are grounded by the power transformer, again, on a dedicated ground lug with a star washer.

That's wrong. The HV CT should have its own wire running directly to the main filter negative. And then the main filter negative goes to the lug.

Yes, but the hum is 60Hz.

Says your meter. I don't necessarily believe it. Try that grounding change and get back to us.

[Hugo]

That's wrong. The HV CT should have its own wire running directly to the main filter negative. And then the main filter negative goes to the lug.


Says your meter. I don't necessarily believe it. Try that grounding change and get back to us.

O.k., I tried your grounding scheme. Unfortunately it made no difference.

I'm curious as to why you think I shouldn't trust the Fluke meter reading of 60.02Hz?


Hugo

[Steve Conner]

Well, because frequency counters aren't very smart. I'm a practising EE, and I've seen all manner of cases where they read twice the actual fundamental frequency of a complex noisy waveform, or half, or some random number, or nothing at all, or you could get any answer you wanted by jiggling the probes around.

I'd want to see the waveform on an oscilloscope.

Either way, since changing that ground connection made no difference, I give up.

[loudthud]

OK, time for a reality check. When you flip the standby switch from Play to Standby does the hum fade out over a period a second or two or does it stop immediately? How about if you leave the standby switch in Play mode and switch the power off?

Disconnect the heater CT from the KT88 cathode and ground the CT. Do you have about 3.2 VAC from each side of the heater supply to ground? Any change in the hum?

[Steve A.]

Just my two cents, Hugo: I would cast my vote with Loudthud's initial reply here: "What kind of grounding scheme did you use? It seems from the description so far that the Master Volume's ground is the source of hum. Have you tried moving it?"

Been there, done that.

The fact that the hum goes away when the MV is turned up pretty much points to a ground problem, at least in my mind. While the star ground is highly touted you have probably noticed that the old Fender and Marshalls did not use that grounding scheme and their amps were fairly quiet.

I learned the hard way that in some circuits you want to keep some grounds separated a bit. My recollection was the tail of a 5E3A-style tone stack and the master volume that was immediately after the treble control. However, that might not have been a hum but an oscillation which would vary in pitch as I adjusted one of the pots. I guess that may not be relevant here... However the idea of trying out different grounding schemes could result in a solution to your problem.

Here is a simple experiment that you can try. Disconnect the ground from the MV and then use a jumper wire to connect it in turn to each the grounded terminals of your front panel pots. If none of that helps then try connecting the MV ground to various ground points on the board. If that doesn't help either, then I suspect that your amp is possessed by Satan and will need exorcism from your local priest.

Without dropping names I learned a whole bunch about grounding schemes from exchanging emails with one of the boutique amp builders here. His goal was to run a single wire from a collective ground on the board to the chassis, but he would put certain grounds together on the board and then run them to the common ground. He would try to connect the grounds of each tube stage together- so the grid load ground and the ground after the coupling cap would go to the same point on his tag boards. And then he would run a lead from each of those ground points to the common ground on the board.

One problem that a star ground will solve is having current paths on the chassis which can introduce all sorts of noise. So what is really important is having a single wire connect the grounds from your board to the chassis. As for isolating the input jacks that sometimes helps but other times is not really needed.


As I said, just my two cents as the perennial hobbyist and founder of electronic mysticism.

Steve Ahola

P.S. You said that the preamp was Marshall style. Were you referring to the original 4 input Marshalls (basically a tweaked 5E3A) or one of the Marshall MVP amps with the added gain stage ahead of the cathode follower (like the 2204)?

FWIW I would usually find the solution to an amp problem here by following one of the suggestions that I rejected right off the bat with "no, its not that".

Good luck!

[Hugo]

I'd want to see the waveform on an oscilloscope.

Here's the hum with a speaker hooked up:



Here's the hum into a load resistor (8 Ohm resistor/8 Ohm output):

OK, time for a reality check. When you flip the standby switch from Play to Standby does the hum fade out over a period a second or two or does it stop immediately? How about if you leave the standby switch in Play mode and switch the power off?

Disconnect the heater CT from the KT88 cathode and ground the CT. Do you have about 3.2 VAC from each side of the heater supply to ground? Any change in the hum?

Flipping either the power or the standby switch kills the hum immediately, it does not "fade out".

Heater CT grounded directly gives around 3.23V per filament side. There is zero difference in the hum level, no matter where the CT is grounded, incl. directly to the output tube cathode.

Just my two cents, Hugo: I would cast my vote with Loudthud's initial reply here: "What kind of grounding scheme did you use? It seems from the description so far that the Master Volume's ground is the source of hum. Have you tried moving it?"



Here is a simple experiment that you can try. Disconnect the ground from the MV and then use a jumper wire to connect it in turn to each the grounded terminals of your front panel pots. If none of that helps then try connecting the MV ground to various ground points on the board. If that doesn't help either, then I suspect that your amp is possessed by Satan and will need exorcism from your local priest.

I jumped the MV ground all over the chassis, including points where it definitely should not be... With zero, and I mean not a fraction of a milivolt, difference in hum level.

The exorcism option is gaining attraction by the hour...

I've used this exact grounding scheme on a number of amps I've built, and they all have been dead quiet. Even ones with much more gain than this one.

But, this is my first single ended amp. Which brings me back to the question if this might be "normal". For those of you who have built SE amps, does 55mV peak to peak of hum at the output sound excessive? But why oh why does the MV affect it?

P.S. You said that the preamp was Marshall style. Were you referring to the original 4 input Marshalls (basically a tweaked 5E3A) or one of the Marshall MVP amps with the added gain stage ahead of the cathode follower (like the 2204)?

The 2204 style with added gain stage.


Thanks guys for not throwing in the towel on me just yet. This is driving me nuts.


Hugo

[RWood]

Have you tried grounding the pot casing?

Tried a different pot?

Eliminated that control from the circuit?

[Hugo]

Have you tried grounding the pot casing?

Tried a different pot?

Eliminated that control from the circuit?

The pot casing is grounded to the chassis via it's star washer and nut, I have not ran a separate ground.

I have tried a different pot, to no avail.

If I eliminate the MV from the circuit, the hum stops. I guess that answers my own question of whether this amount of hum is "normal" in a SE amp in the negative.


Hugo

[Steve Conner]

Can you explain more by what you mean by "eliminate the MV", then? As in disconnect the power tube from the rest of the circuit and leave it dangling, or as in connect the power tube's grid straight to the output of the preceding stage?

If the latter, then grounding the CCW end of the MV's track to the same point as the cathode of the tube whose plate drives the MV, should get rid of the hum.

[MWJB]

The schem looks wrong! Your maximum load for the power tube grid can't ever be what it should, because of the tone stack loading down the volume control, at low settings the power tube grid might even be seing a comparative short.

Ground a 220K to 470K resistor from pin 5. Now add an additonal coupling cap between that resistor and the volume pot wiper.

[loudthud]

Does it hum before the tubes warm up?

If the standby switch cuts the hum off, that points to something in the power supply grounding. Almost all of that will be spikey 120Hz stuff. The 60Hz sine wave Hum is usually from the heater supply. It will fade out. The only explanation is demonic possession.

I don't understand what the difference is in the two scope photos. What changed? Why does it look like two traces in the first photo?

The Schematic shows a 5.6K grid stopper on the KT88. Where is it located?

[Hugo]

Can you explain more by what you mean by "eliminate the MV", then? As in disconnect the power tube from the rest of the circuit and leave it dangling, or as in connect the power tube's grid straight to the output of the preceding stage?

The MV is fed off the wiper of the treble pot. From the MV it goes into the power tube grid. "Eliminating the MV" means connecting the wiper of the treble pot straight to the PT grid, bypassing the MV altogether. ->no hum.

If the latter, then grounding the CCW end of the MV's track to the same point as the cathode of the tube whose plate drives the MV, should get rid of the hum.

That's exactly where I had it grounded in the first place.

I've moved this ground all over the chassis, with no effect. I'm fairly confident that it is not the culprit here.

Does it hum before the tubes warm up?

No.

I don't understand what the difference is in the two scope photos. What changed? Why does it look like two traces in the first photo?

The first picture shows the amp being plugged into a speaker.

I think the two traces are due to the fact that the speaker generates kickback voltage as the cone travels backwards. So you're seeing two signals. At least that's what I think, somebody correct me if I'm wrong.

The second picture is the amp plugged into an 8 Ohm 100W load resistor. I don't understand why the wave form looks the way it does. I was expecting a sine wave like in the first picture. Maybe somebody can explain it...

The Schematic shows a 5.6K grid stopper on the KT88. Where is it located?

Right on the output tube socket, with a shielded wire going to it.

The schem looks wrong! Your maximum load for the power tube grid can't ever be what it should, because of the tone stack loading down the volume control, at low settings the power tube grid might even be seing a comparative short.

Ground a 220K to 470K resistor from pin 5. Now add an additonal coupling cap between that resistor and the volume pot wiper.

Wow, you might be onto something here. That makes a lot of sense. I can't try it now, but I will tomorrow and report back. This could be it.


Thanks everybody,
Hugo.

[Hugo]

The schem looks wrong! Your maximum load for the power tube grid can't ever be what it should, because of the tone stack loading down the volume control, at low settings the power tube grid might even be seing a comparative short.

Ground a 220K to 470K resistor from pin 5. Now add an additonal coupling cap between that resistor and the volume pot wiper.

Ok, something bizarre happened that I do not understand.

I first added a 470k resistor from pin 5 of the power tube to ground. This had absolutely zero effect on anything, hum was exactly the same.

Then I added the coupling cap you suggested. At that point the hum got louder and louder and after about 15 seconds the output tube catode bypass cap violently exploded like a fire cracker.


Hugo