Advice on new design

[marcoloco961]

Here is the schematic to a Bass amp I am trying to design. I have a pretty decent quality signal and tone right up to the point where I took the Pre-amp out. Somewhere between there and the speaker output I am gaining a raspy distortion when turned up over half volume and when I really dig into the strings. At half volume or less it is only noticeable when you dig in really hard, but as the volume comes up it happens sooner. When used through the pre amp out and a separate power amp, the tone is awesome. I have added voltage dividers between stages to try and stop it, I also removed the NFB which seemed to help the most, so I am leaning toward and oscillation problem? I am noticing a lack of grid leak resistors, yet experimenting with them has not seemed to make a big difference.

I do have to admit the lead dress is horrible as this amp has been changed so many times, so I am thinking that might be where the problem lies. I was wondering if anyone here might be able to spot an obvious problem that I am to green to spot. The last thing I changed was the voltage divider directly prior to the PI. I actually raised the voltage some and it seemed to help, but that seemed to go against what I was trying to do so I was hesitant to raise it any more. Am I thinking right on this?

[ChrisM]

Just a thought dont you need a cap off the fourth triode to block the HV DC? Your voltage divider is dividing the HV DC.

[tictac]

the grid of the last triode before the PI needs to have some sort of ground reference; that .001 coupling cap and the .02 PI caps seem small values for a bass amp...

Also with a 46k tail resistor two 100k plate resistors would be more appropriate to get the correct balance for this circuit...


TT

[marcoloco961]

Just a thought dont you need a cap off the fourth triode to block the HV DC? Your voltage divider is dividing the HV DC.

Ahhh good point, I will try moving it back in front of the divider to see if that helps, thank you.

[marcoloco961]

the grid of the last triode before the PI needs to have some sort of ground reference; that .001 coupling cap and the .02 PI caps seem small values for a bass amp...

Also with a 46k tail resistor two 100k plate resistors would be more appropriate to get the correct balance for this circuit...


TT

The caps are small due to dealing with what I believe was a blocking distortion, although I have slowly been bringing the values back up in other areas without experiencing the blocking distortion again, so I might have changed something elsewhere since I installed the smaller value caps that has eliminated the BD.

For a ground reference on the grid before the PI, are you suggesting a grid leak? What kind of value(s) might be proper in this spot?

I have been looking at different bass amp designs and it seems common to use 330K resistors in place of the 1M in the LTP and more like a 22K in place of the 46 K tail. Do you think it would be better to just change the one 82K plate resistor to a 100K as opposed to all the others? Or would just changing the 46k to a 22k tail put the 100/82 plate resistors more in line?

Thanks for the input everyone.

[jjman]

The 82k before the grid of the 3rd stage would be an ac and dc reference-to-ground but the .001u after it is blocking the dc reference. This would let that stage float up to saturation. I don't think that .001u should be there for any reason.

[Firestorm]

If it sounds good taken from the pre out (before the 4th stage) into another power amp, maybe you want to try bypassing that 4th stage and going straight into this power amp. Just to see.

[marcoloco961]

If it sounds good taken from the pre out (before the 4th stage) into another power amp, maybe you want to try bypassing that 4th stage and going straight into this power amp. Just to see.

I believe I tried this at one point. IIRC the gain was so low like this that it was too weak to really judge tone.

[marcoloco961]

The 82k before the grid of the 3rd stage would be an ac and dc reference-to-ground but the .001u after it is blocking the dc reference. This would let that stage float up to saturation. I don't think that .001u should be there for any reason.

Makes sense, I seem to remember adding that low coupling cap earlier in the process to fight what I thought was blocking distortion. Now that you point it out, I guess it is pretty much not needed directly after the TS. Would this cause the breakup I am having problems with? It sure seems like it might to this rookie.

Thanks again for taking the time to look at it and offer good advice. I had a hunch someone here would be able to spot some of my errors quickly.

Keep the opinions coming, it is amazing how much you can learn and still only scratch the surface. With each thing I learn, I find out just how much more I do not know.

[billyz]

You don't need thew 4th stage. Use that 1/2 for say a paralleled triode V1 to fatten it all up and reduce noise. Or a buffered line out, preamp outs like that really load down the signal when used. For a bass amp I would definitely go with a choke instead of the 1K/25w resistor. It will also firm up the low end. I would increase the first series filters to 220uf ea for 110uf total. I would have used a little more voltage on the secondary too, but you are probably committed to that PT. Whats with the 10k Grid resistors ?

[marcoloco961]

You don't need thew 4th stage. Use that 1/2 for say a paralleled triode V1 to fatten it all up and reduce noise. Or a buffered line out, preamp outs like that really load down the signal when used. For a bass amp I would definitely go with a choke instead of the 1K/25w resistor. It will also firm up the low end. I would increase the first series filters to 220uf ea for 110uf total. I would have used a little more voltage on the secondary too, but you are probably committed to that PT. Whats with the 10k Grid resistors ?

I had wondered about the Choke. I know that using a choke as opposed to a resistor in low pass filters had always sounded better to my ears. I will give it a try. Any suggestions on value?

What is the advantage of the higher filter caps, other than the lower ripple?

Yes to the PT, it is a build with spare parts project.

The 10K grid resistors were also a result of trying to tame the Blocking distortion problem I initially had. I had read somewhere that increasing these would help tame blocking distortion and oscillations. I had them as high as 20K at one point looking for a little Hiwatt/Soundcity mojo.

To be honest, I keep saying Blocking distortion, but I am not really sure what it was. An oscillation of some sort, I imagine. When I would hit the string hard, it would be completely silent for just a fraction of a second (not even a normal hiss) then the sound would come back in low and raise to a normal volume level.

Excellent ideas, all of them. I will have to give it a go. The damn thing is so close, but frustrating me horribly with the final tweaks. Love it as a pre. I can't believe how Bassy it is for as small of couplers that I used in a few spots. Thanks again guys.

[marcoloco961]

Alright, I have removed the un-needed .001 cap, inserted a .1 cap before the voltage divider prior to the PI and changed the 82k plate resistor of the PI to a 100k. It seems to have cleaned it up some as long as I stay below 1/2 volume, as soon as I go past that point it is still sounding like a square wave signal when the strings are attacked. Damn.

I also substituted an old choke I had here for the 1k 25w resistor and it seemed to make the issue worse so I put the resistor back. I really am not sure what value the choke was so I am not judging the results as final. I have tried lowering the signal strength via the voltage dividers at different stages, but that just seems to lower the gain and not eliminate the problem.

Let me approach this in a different manner, Can someone give me some approximate p-p A/C signal values at the input of the PI and say maybe at the input of the power tubes. At the pre tubes I have been aiming for somewhere between a 2 to 3V P-P signal to keep the headroom high and avoid going into grid limit or cut-off. I have been measuring with a true rms fluke meter and multiplying by 1.414 to get the 1/2 peak value then x 2 for peak to peak. Am I doing this correctly?

What would the "normal" or "common" value be for P-P A/C entering the PI. I did notice that last time I raised the signal coming into the PI it seemed to help some, but as I stated earlier, I was hesitant to raise it any further, maybe I should give it a try just for the hell of it?

[billyz]

Double check your Feedback loop. I see the bottom half but not the top half coming from your OT secondary. NFB will also clean up your bass.

[tubeswell]

And won't the LTP sound better with a 12AT7 set up like in a blackface amp? (or at least a pair of 12AU7 cathode follower/buffers to drive the 6L6 grids? - since this is a bass amp)

[billyz]

And won't the LTP sound better with a 12AT7 set up like in a blackface amp? (or at least a pair of 12AU7 cathode follower/buffers to drive the 6L6 grids? - since this is a bass amp)

I agree, 12at7 for the PI.
I think there is consensus on removing the 4th stage, too.
And rework your NFB loop.
After those are done I would revisit the Choke in place of the inductor/sag resistor.