Twin Reverb Silverface 100W 1976: missing a 0.01uF capacitor and a 220K res

[Lambertini]

Hi there,

I'm doing a full recap of a venerable 1976 Twin Reverb Silverface 100W.
I ordered all caps based on the schematic provided with the amp, which has been proven correct until now.

I've done and redone my calculations. I've ordered a 0.010 uF capacitor in excess.

I read again and again the schematics. And I discovered that a part of the circuit drawn on the schematic does not exist in my actual amp.
A coupling of which I have not the slightest idea of the usefulness between the negative voltage circuit of the bias and the phase inverter is missing.
I circled it in orange in the attached file. Neither the 0.01uF cap nor the 220K resistor are in the actual amp...

And by the way, the 470R cathode resistor of V6 is actually 680R in my amp.

What does this mean?!

Thanks in advance!

[BrownIsound]

I believe it to be additional negative feedback for the PI. I do have this in 2 1971 amps (a quad reverb and a SuperSix). I have not found much discussion on this on previous searches, but maybe someone else will comment on this thread.

As for the cathode resistor, I believe that will lower the current of the PI over the stock 470. Does it appear to be factory?

[BrownIsound]

BTW, are you replacing all the caps for any particular reason?

I ask because my 71 SuperSix is almost all original (other than PS and bias supply filter caps, trem roach, and a screen grid resistor), and it sounds righteous, though not much different than my quad, that has had all of the electolytic caps and most of the brown coupling caps replaced.

[stratomaster]

Interesting, looks like local negative feedback but going to the 2nd input of the PI, which would make it constructive feedback.  An attempt to increase gain from the PI?

I'm interested in hearing more about your circuit.

I'm also interested in why you're replacing every capacitor in the amp. 

[BrownIsound]

Ahh, yes you’re right, positive feedback for some range of treble frequencies. These amps definitely have a certain treble spank to them.

[Lambertini]

For a little background, the amp was 100% stock from 1976 and had a rough touring life...

I purshased it few weeks ago, not as a collector but to play it. Sounding airy and deep altogether, good mids, with unfortunate drawbacks: low volume, bad pots, very noisy and hummy, almost all tubes microphonic, bad reverb, almost no vibrato, very low sensitivity to guitar signal...
Once open I found the circuit greasy and dirty, with filter caps in awfull conditions. I changed the filter and bias caps, replaced all tubes, screen resistors, pots, the LDR, and the reverb tank. I also got rid of the meaningless push/pull boost and finally completely remove the master volume.
The amp got much more sensitive and powerfull but still had an important 50 cycles hum, a lot of parastitic noises and an incredible hiss.
Changing the caps helped a lot. The amount of weird noises reduced significantly, without fixing all of them (still have some random small cracklings) and with low effect on hiss.

I'm currently working on the 50 cycles hum. I rearranged the heater wiring, also replaced the wires between inputs / volume pots and tubes with new shielded ones. This was a rather good success. Still need to replace the hum balance pot which is crackling and has a unexpected behavior.

Next step is to replace most of the carbon resistor to tackle the hiss.

Now I have an unused cap, I can leave with it, but I'm very surprised of a circuit part missing in the amp vs the schematic. Should I try to rebuild it? There's an empty eyelet close to where the components should be.
The existing 680R instead of expected 470R really looks stock, as dirty as the others ;)

If you want to follow the journey, here's the photo album I'm feeding.

Thank you for the info on the positive feedback / treble frequencies. Need to evaluate this.

[pdf64]

…
Still need to replace the hum balance pot which is crackling and has a unexpected behavior.
 …

For the heater circuit not to cause buzzy hum, it should be balanced and have a low impedance path to circuit common.
A bad humdinger pot can cause problems with both those requirements.

Just use a couple of 100 or 220R resistors to balance the heater circuit.

[PRR]

If it is feedback, it is insignificant. The 12AT7 gain is like 40. The 220,000:100r divider is 2,200:1. There's no effective feedback.

It *may* be an attempt to improve the PI balance, but the 12AT7 plate resistance is like 20k so a 220K load is only 10% change, and a little imbalance is not bad, may be good for tone.

There is also the chance the draftsman took a long weekend in the middle of this drawing and lost his place. Or Leo was tired of copycats and left a deliberate error. 

[Lambertini]

Just use a couple of 100 or 220R resistors to balance the heater circuit.

Like completely removing the hum digger pot (pic1) and adding a 100R/220R resistor on each lug of the pilot lamp holder (pic2), going to ground?

If it is feedback, it is insignificant. The 12AT7 gain is like 40. The 220,000:100r divider is 2,200:1. There's no effective feedback.

It *may* be an attempt to improve the PI balance, but the 12AT7 plate resistance is like 20k so a 220K load is only 10% change, and a little imbalance is not bad, may be good for tone.

There is also the chance the draftsman took a long weekend in the middle of this drawing and lost his place. Or Leo was tired of copycats and left a deliberate error. 

What's still surprising me is that the current coming this way is negative voltage. The aim would be injecting some negative voltage at V6B grid to compensate the signal already amplified by V6A coming from the V6B cathod?

[stratomaster]

The cap blocks the DC, positive or negative.  The point where the feedback is injected into the PI is pure AC.  By taking it after the PI output cap you can use a lower DC voltage rated cap to do the same job.  It only needs to block ~50v instead of ~400v.

Dumble local negative feedback of a triode was at plate voltage before the cap.  Often using 10M to 44M of resistance, so local feedback ratios are usually very low in practice from what I can gather.

[Lambertini]

The missing parts I ordered arrived. Would that modifications fit the gap with the schematic in the first post?

[sluckey]

Probably goes like this...

[Lambertini]

Indeed. Thank you Sluckey.