Weird issue with amp rejecting negative feedback.

[eddie25]

I have a recent 6V6 Marshall style build that had some distortion on the low end (mostly only noticeable when clean). One of the first things I did was remove the negative feedback and the amp sounded great. I then switched the OT primary wires and hooked up the NFB again and the amp sounded even worse than before.

I've traced through and through (not ruling out that it's not a simple error on my part), I also tried some different NFB schemes. I have a hard time suspecting any specific components because the amp sounds great with no NFB. Could this be an OT issue? I suppose I could either leave the NFB out because it still sounds really good, or start a more exhaustive troubleshooting process, but I am incredibly busy lately, so the real question is has anyone ever experienced this before? Thanks guys.

[Stevem]

I like amps better with no feedback due to the fact that the cross over point in the amps output range between clean and clipping is wider, and give a far better range of crunch tones before full on clipping!

If you like the way it is better with no feedback then just make the two PI plate load resistors the same highest value if it's a long tail type, and call it a day!

[martin manning]

If you like the way it is better with no feedback then just make the two PI plate load resistors the same highest value if it's a long tail type, and call it a day!

I disagree here, Steve. The miss-match in loads is to improve the balance of the LTP PI, which is not significantly affected by the NFB (or the lack of it).

It could be that there is a lead dress issue that is causing a problem when the primaries are connected correctly.

[VacuumVoodoo]

I I then switched the OT primary wires and hooked up the NFB again and the amp sounded even worse than before.

Revert back to original connection of PT primary. You created PFB which is a sure fire way of turning a good amplifier into an even better oscillator.

[eddie25]

I like amps better with no feedback due to the fact that the cross over point in the amps output range between clean and clipping is wider, and give a far better range of crunch tones before full on clipping!

If you like the way it is better with no feedback then just make the two PI plate load resistors the same highest value if it's a long tail type, and call it a day!

The problem is that I'm not exactly sure if I like it better because one way is distorting and I would also like to experiment with different NFB types if I can get it to work. I could use slightly more headroom as well. I will say it sounds damn good so I'm not completely stressing about it.

It could be that there is a lead dress issue that is causing a problem when the primaries are connected correctly.

I suppose this is where I should look first. I have used different versions of this layout many times though. The NFB wire and the power tube control grid wires go above the board straight to the tubes/jacks. The OT secondaries do go right under the phase inverter wires, but how many other thousands of amps have that type of situation?

Revert back to original connection of PT primary. You created PFB which is a sure fire way of turning a good amplifier into an even better oscillator.

Thanks, I'm aware of that and they are back to the correct way. I'm using a Hammond and they clearly label their ends. Actually, I was surprised at how much it did not oscillate when I reversed them, it just made more distortion. Does that tell me anything? Shouldn't it be an intense squeal? To this day I have never set up an amp with reversed OT primaries.

[pinkphiloyd]

Shouldn't it be an intense squeal? To this day I have never set up an amp with reversed OT primaries.

Not necessarily. Like you, I had never encountered the issue, until my last build. What I had was a ~45 Hz hum. After some probing around with my meter I disconnected the negative feedback and the problem went away. Switched the control grid problems to my power tubes (they were easier to get to than my OT leads), hooked the neg. feedback up, problem solved. What the conditions are that would determine whether you got a high pitched squeal or the deep rumble that I got, I'm not sure...

[tubeswell]

What feedback does to your amp also depends in part on the way your OT is constructed, including the overall inductance and the losses from the core (which affect the bandwidth from primary to secondary), as well as the interleaving of the windings in your OT.

In particular, more interleaving helps deliver more bandwidth and better balance between the two sides of a PP OT, but also promotes more inter-winding capacitance and more phase shift at higher frequencies. If there is 'enough' (i.e. too-much) phase shift at higher frequencies, then this can result in positive feedback on higher frequencies when you hook the NFB loop up. In which case, you need to watch how much feedback you apply.

Not saying its what's causing your issue, because its not the same in all OTs and all amps. Also of course other things affect feedback besides the OT and the closed loop gain of the circuit, but the particular construction of the OT (in terms of the number of interleavings) can explain why some OTs react differently to feedback in different circuits than others. Something to consider.

One possible 'fix' for this suggested by Merlin Blencowe in the 1st edition of his 'pre-amps book, which I have found useful in some amps, is to use 100-250pF between each output tube's grid and cathode. This is quite effective at eliminating HF instability resulting from NFB.

[Stevem]

Without the feedback your PIs output level is likely driving the outputs into clipping too early .
I would bet on the simple change over to a 12Au7 will put you very close to what you want for clean output level, or if you have the parts on hand stick in a cross type master volume to pin down what drive signal level you need out if the PI.
Unless I am building a amp strictly for clean usage, or Hi FI use I do not like or use feedback, in fact where it not for the fact that early guitar amps where direct copy's of PA and theater sound systems I do not think that guitar amps would have ever come to have feedback circuits in them at all!
With the quality of parts and the what can be had these days with the frequency responce of PTs the added high end of having a feedback circuit in place can make building a high gain amp that does not become its own transmitter all that much harder!

[Littlewyan]

Have you got a photo of the inside of your amp so we can see how your wires are routed?

Also what Marshall circuit have you used? With the NFB hooked up the bass should be tighter and with the presence control you can add high end.

[eddie25]

Well guys, it turned out to be a wiring error for the OT secondary ground, which meant it wasn't grounded at all. Classic bonehead move, I'm surprised it ever sounded good. Took about 5 minutes to find. After really thinking that the wiring was fine (if not my best work yet) I started playing around with the grounding and violo! Amp sounds amazing, settled on a superdrive NFB scheme (with fixed value rather than presence pot) after focused testing.

[tubeswell]

Ah yes - that'll do it. Without a ground reference for the secondary, if you have NFB tapped off the secondary, the signal tries to 'reference' itself to the NFB voltage divider, which will be subject to the pre-amp signal at the NFB insertion point, with the result that the secondary will be all over the place.