AB165 Bassman crackle in V1 and V3 stages??

[martin manning]

Is there any voltage across that 470k resistor going from the 0.01u/V2 grid to ground? There shouldn't be, unless the 0.01 is leaky.

[lespaulnmarshall]

Is there any voltage across that 470k resistor going from the 0.01u/V2 grid to ground? There shouldn't be, unless the 0.01 is leaky.

Oops, I just noticed that Paul's post was about the 470k grid leak resistor. I had assumed it was the 470k feedback resistor he was talking about instead. That is the one I changed and that solved the issue... The grid leak is still a carbon comp.

[martin manning]

Ok, that makes more sense. That resistor has some DC current flowing through it. Some people report lower noise after switching the supply node for the second stage of the normal channel to be the same as the one feeding the first stage. The LNFB should really have a blocking cap.

[lespaulnmarshall]

Ok, that makes more sense. That resistor has some DC current flowing through it. Some people report lower noise after switching the supply node for the second stage of the normal channel to be the same as the one feeding the first stage. The LNFB should really have a blocking cap.

I've had it hooked up to the other node when it was still crackling, might want to try it again just to hear the difference. Trying a blacking cap for the LNFB also seems like a good thing to experiment with, are there any values you'd recommend trying?

Also, just when I thought this amp couldn't be any more troublesome, it started making horrible noises again this morning. Luckily I solved it relatively quickly this time as I could track down exactly where it was happening, the second triode of the normal channel 12ax7. Turned out to be either a bad cathode bias resistor or a bad bypass cap. Replaced them both just to be sure. Hope this is the last issue I'll have with this one for a while. It was used for 6-8 hours a day, every day of the week for years on end. At this point it's about fully rebuilt.

[martin manning]

Trying a blacking cap for the LNFB also seems like a good thing to experiment with, are there any values you'd recommend trying?

I think a 22n should be large enough. I'd be interested to hear about what that does if you try it.

[lespaulnmarshall]

Trying a blacking cap for the LNFB also seems like a good thing to experiment with, are there any values you'd recommend trying?

I think a 22n should be large enough. I'd be interested to hear about what that does if you try it.

Will give it a try! Not quite sure when as I'm already just really glad the thing just works as it should at this stage, but I will post an update once I get bored and decide to give it a try!

[lespaulnmarshall]

Trying a blacking cap for the LNFB also seems like a good thing to experiment with, are there any values you'd recommend trying?

I think a 22n should be large enough. I'd be interested to hear about what that does if you try it.

Will give it a try! Not quite sure when as I'm already just really glad the thing just works as it should at this stage, but I will post an update once I get bored and decide to give it a try!

Okay I guess I get bored pretty quickly. I just tried both removing the 27k/8uf node and adding the 22nf to the LNFB loop. Removing the extra node added a bit of noise in fact. Adding the 22nf cap did not make a huge difference IMO.

[martin manning]

Okay I guess I get bored pretty quickly. I just tried both removing the 27k/8uf node and adding the 22nf to the LNFB loop. Removing the extra node added a bit of noise in fact. Adding the 22nf cap did not make a huge difference IMO.

Well, none of us have many places to go right at the moment...

The supply node change involves lifting the supply voltage lead running over the board from the normal channel second stage, and jumping the first stage supply lead over to the second stage. Did you do something different?

I guess I'm not too surprised that adding the blocking cap didn't make much difference once you had already put a quiet 470k resistor in, but I'd argue that is something that should be there.

[lespaulnmarshall]

Okay I guess I get bored pretty quickly. I just tried both removing the 27k/8uf node and adding the 22nf to the LNFB loop. Removing the extra node added a bit of noise in fact. Adding the 22nf cap did not make a huge difference IMO.

Well, none of us have many places to go right at the moment...

The supply node change involves lifting the supply voltage lead running over the board from the normal channel second stage, and jumping the first stage supply lead over to the second stage. Did you do something different?

I guess I'm not too surprised that adding the blocking cap didn't make much difference once you had already put a quiet 470k resistor in, but I'd argue that is something that should be there.

Regarding the node, yes that is pretty much what I did, only I did it at the other end, under the doghouse, but it comes down to the same thing.

As for the blocking cap, I didn't really have very neat way to install the cap so I took it back out. It might be a little irrational, but to an extent, I do like doing things the way fender did them, as much as possible. That said, many of the changes from AA864 to AA/AB165 are really not so desirable for guitar purposes so I'm not sure if there is a point to sticking to that rule with this one. Plus, it's already got a few mods...

Also I noticed there is still a tiny little bit of a 'frying bacon' type sound when I run the bass and volume way up and the treble on 0 (aside from the obvious noise that you'd expect from doing that of course) on the bass channel, but perhaps I'm hearing the resistors that are still carbon comps? That channel now has two .022's, a 56k (carbon comp) slope resistor, 25k mid pot and a 250PF in the tonestack like a 5F6-A would have. The .1uf on the deep switch is removed. The problem only happens with that specific tone/volume setting. The tone caps aren't leaky. Is this anything that needs to be checked out you think? It looks fine on my scope, no weird oscillations or anything, amp sounds great otherwise and you really have to listen closely to hear the bacon noise. This is really the only Bassman I've ever worked on so I don't know what's 'normal' for them really. I feel like they are slightly different beasts than AB763 designs which I am very familiar with, rarely ever have noise issues with those. Also, maybe I have gotten really a bit obsessive with the noise levels in this amp?

I must add, adding an aligator clip lead to ground from the bass cap takes care of the problem, but then again, so would turning the bass control to 0.

[martin manning]

Regarding the node, yes that is pretty much what I did, only I did it at the other end, under the doghouse, but it comes down to the same thing.

If you move the lead under the doghouse, I believe that would leave nothing on the "A" node. The intent is to move the second stage of the normal to the "B" node (with V1a and b) and leave only the mixer stage on the "A" node.

[lespaulnmarshall]

Regarding the node, yes that is pretty much what I did, only I did it at the other end, under the doghouse, but it comes down to the same thing.

If you move the lead under the doghouse, I believe that would leave nothing on the "A" node. The intent is to move the second stage of the normal to the "B" node (with V1a and b) and leave only the mixer stage on the "A" node.

Ah, gotcha!

I thought you meant modding it to essentially be an AA864 power supply. Will give it a try tomorrow!

[lespaulnmarshall]

Regarding the node, yes that is pretty much what I did, only I did it at the other end, under the doghouse, but it comes down to the same thing.

If you move the lead under the doghouse, I believe that would leave nothing on the "A" node. The intent is to move the second stage of the normal to the "B" node (with V1a and b) and leave only the mixer stage on the "A" node.

Tried it, great advice, definitely less hiss now! Thanks!

Regarding the bacon noise, it's really only with the amp flat out, so I suppose it is normal. I can reduce the level of noise with different tubes too, so it probably isn't anything odd going on.

[Paul G.]

... The intent is to move the second stage of the normal to the "B" node (with V1a and b) and leave only the mixer stage on the "A" node.

I was curious about this one, so I gave it a shot. Noticeably quieter than before, even on an amp that started out pretty quiet.

I'm looking at the schematic and I'm just wondering why this makes so much of a difference in the noise floor?

[martin manning]

I believe it is due to the different voltages causing DC current to flow along several paths, including through the 470k LNFB resistor, which is connected to the grid of the mixer stage. Those voltages will change with different tubes, too. Switching the supply node must be reducing the DC current at the mixer grid.

[Paul G.]

I see it now, it's a junction of the plates of 3 different tubes with no decoupling caps anywhere. DC everywhere, going anywhere.