Major Homebrewed amp problem

[GainFreak1984]

Hi all, this is my first post on here and this seems to be THE place for tech assistance for tube amps! Anyway, here's my problem...

I've got a two channel home brewed amp over here. Basically it's a dual rectifier high gain channel and a 2204 low-er gain channel. There's also a TUT parallel effects loop as well. I'm using a magic parts 50w marshall replacement tranny for the output and a Super Reverb tranny for the power.

Now this is my problem. Amp works and sounds good with the exception of this weird ghosting/non-related overtone, but only on the high gain channel when you play on the higher frets. It does have some hum in the high gain channel but i've heard a LOT worse from some commercial amps.

I got it on a scope and couldnt find any signs of oscillations or anything. However, the main power supply rail is jumping around all over the place even when you dont apply a signal AND when there are no tubes in the amp.

As far as the overall waveform it's a textbook sawtooth pattern with about 5v of ripple peak to peak and 480vdc at idle. The preamp caps are Nichicon 33uF and the power supply caps are UCC 220uF series stacked for a total of 110uF in the HT and screens...all fresh stock.

Here's what I've tried so far.
Swapped tubes = no change
Replaced filter caps = no change
Played with different grounding schemes = no change

Sorry for the long winded post but wanted to give you guys as much complete info as possible. I'm completely out of ideas and any help would be greatly appreciated!

[stokes]

Get rid of the jap filter caps.If your power supply is eratic with no tubes it only leaves your caps in the circuit.Get some good quality Sprague caps.Those jap caps suck,even if you get some that filter properly,they sound like crap compared to Sprague or F&T's.

[Amp Kat]

Even if the Nichcons don't sound good they should still filter correctly which makes me believe you still have a problem like something touching that's not supposed to be like a Cathode cap or anything touching a plate resistor or any high voltage source. When this happens I disconnect the HV supply from the circuit and just leave the filter caps hokked up without no load. If it does it then you have a power supply problem which only leaves the power supply,filter caps and rectifier and the standby and power switch.

[stokes]

Not likely anything to do with the cathode caps,since he already said he has the problem with the tubes out.If something was touching the plate resistor to ground he would likely blow a fuse,With no tubes,there is no load, the only thing that could draw current would be the caps.The rectifier is still in the circuit,but with a rectifier it is either good or bad usually.The ghosting is a sure sign of bad caps,and the fact that the problem is there with no tubes.

[GainFreak1984]

AmpKat, so basically what you're saying is to disconnect EVERYTHING so that all that's hooked up is the caps attached to the rectifier?

I'll do that now and post back in a while with what shows up. If it still acts weird then what should I start looking at...replacing the caps again? What would be the odds of having a whole batch of bad caps? I suppose I could pimp the big caps out of my dual showman temporarily just to see what happens.

btw, how could a standby switch cause problems like this? The ones I'm using are the 6A Carling SPST.

[Enzo]

OK, you have 480VDC with 5v pp on it. That ripple is less than 2v RMS, and that is not bad at all for an unregulated supply. If the tubes are out, then the voltages mean even less. Tubes need to be in there before we pay much attention to voltages. Without tubes, no current is drawn beyond leakage, so the B+ will just sit there in the pipe... bouncing. I think when you add the tubes, the B+ will fall into the proper voltages, and I also think you will find that ripple will only exist on the power tube plates. And there it will cancel out. Look back at the preamp B+ nodes - WITH TUBES - and see how clean they are, or are not.

480v is four times 120v. Every volt then that the mains change will result in a 4 volt change on your B+. Monitor the mains for a while to see just how unstable that is. B+ bounces all over the place normally, as do your mains.

But who says hopping B+ is his problem anyway. Correlation is not causation. His problem is ghost notes, and his problem could not possibly occur with the tubes out. No tubes, no notes. He decided the B+ voltage was his problem, but I would not accept that diagnosis myself. I tend to doubt his problem is his filters, wherever they came from.

The symptom is only present on the high gain channel and only up the neck. I would be more inclined to think his lead dress had an issue, or something related. COUld even be the speaker. Have we tried other speakers? And of course othe guitars? Even if it is in the amp, the higher gain channel may be distorting the waveform enough to cause distortion artifacts that don't show up in the lower gain channel.

Got a dummy load? Load the amp and listen through headphones, or just monitor at lower levels through another amp.

And while on that track, apply a hot signal to the loop return to see if this noise occurs in the power amp. And send the loop out to another amp to listen to it. Isolate the problem to part of the amp.

[GainFreak1984]

ok, I did as was recommended by ampkat and the bouncing B+ is still there. ONly difference was that there was virtually no ripple at all...2v maybe! But the dc voltage went way up to something on the order of 495v. Also, I remeasured all the dc resistances on the caps and every one of them was way over 100k. I also temporarily swapped in a pair of 100uF@350v atoms and there was no difference except maybe a little more ripple but not much. So maybe there isnt a power supply problem at all.

WRT enzo's comments....
I dont have another cabinet to try it with. Just a marshall 4x12 with V30's.
I've also tried it with two guitars...Jackson Soloist and a Fender strat. Same thing both instruments.

As far as the other suggestions, it seems like they're more for confirming a problem than diagnosis. So I guess the questions now would be what else could cause this, and how would I go about finding the actual problem is and getting it fixed?

[stokes]

"However, the main power supply rail is jumping around all over the place even when you dont apply a signal AND when there are no tubes in the amp."

I am assuming in this quote he is saying the dc voltages are "jumping around".With no tubes the only thing is the rectifier or the caps,no?
Gainfreak,confirming a problem is diagnosis.

[GainFreak1984]

actually no....the dc voltage is fairly constant. it hops around on the meter by like 2 volts, but that's an average value right? What's hopping around is the waveform on the scope. Its bouncing around vertically by about 2 or 3 divisions on the 5v/div setting and there's no periodicity to the movement. so it seems pretty random

[Enzo]

If you turn up the sensitivity of any measuring instrument - meter, scope, whatever - you will eventually find something to watch.

Think in percents. At 480VDC, 5 volts is only 1%, while 2 volts is half a percent. That is darn clean DC for a tube amp.

Do you have any friends who play? SOmeone who would let you play your amp through their cab just to compare the sounds?

[Amp Kat]

I've seen that scope jumping several times but can't remember exactly what it was. Enzo is on the right track with isolating the problem from preamp to power amp. Send the high gain signal to another power amp or return to another amp and see if it freaks on you at the high frets. Realize your frustrated but without the amp right in front of us it's sometimes hard to diagnos,troubleshoot or whatever it's called especially a home brew so hang in there.

[Bruce / Mission Amps]

My scope does that once in a while when the probe ground connection is loose or bad, or I am not using it at all and the ground of the amp chassis is not hard core secure.
The ghost noting under higher volume loads, not related to tubes or speakers, is most frequently a main filter cap ground point issue or an actual poor capacitor.
I've used every cap out there and eventually, one of all of them will do it.
Sprague, Mallory, F&T, Ruby, Xicon.... no matter, I always seem to find one that checks OK on the Sencore but still allows the amp to make a sounds that are not part of the music. This can drive a builder crazy.

[GainFreak1984]

Update...

Ok, I took the amp to a friends house and we ran it through his 2x12 and another 4x12 cab and it always does it so i know its definetly an amp problem. as for ghost notes at high volume, its there all the time and you dont even need to have the amp turned up very loud so its not a high volume thing.

Now here's something that might mean nothing like enzo said if you turn it up high enough theres always something to look at. But anyway, if i go to the 10x horizontal setting i found what i think is called a damped oscillation coming from the output of the third stage(the one with the 39k cathode). it's about 50mv on the 5uS/div scale so that's some seriously high frequency shit! I get 3 cycles of a sine wave then one smaller then it's straight for one time division. Then it repeats itself. Could this be something or am I just chasing shadows?

[Enzo]

No, THAT sounds like parasitic oscillations. Try moving the wires around while it is doing it.

The only part I wanted to ignore was the B+ moving around within a few volts.