first build...It works (!!)...but....

[MARK1]

It is difficult for us to give any advice if you start with a modified baseline of a super reverb without the normal channel, so my advice is, go back to your baseline and correct the mistakes in your schematic. If you are still not happy with the result, the guys over here can give pointers in what direction to go.

A few thing should return to basic. Most important, and one of the mistakes in your schematic is replacing the intensity 50K pot to ground with a 68K resistor and 250K pot, just replace the 50K intensity pot for a 50K resistor. Your next mistake is, you left out the remaining 220K mixing resistor to the coupling cap of the PI. I think If you change this and return to the basic schematic, your amp should sound like a SR.

I don’t think the speaker is of major influence, as you already replied, because deluxes and twins also use 12 inch speakers.

I have been reading here for some years, but this is the first time I replied.

Luck with your build,

André

[pula58]

It is difficult for us to give any advice if you start with a modified baseline of a super reverb without the normal channel, so my advice is, go back to your baseline and correct the mistakes in your schematic. If you are still not happy with the result, the guys over here can give pointers in what direction to go.

A few thing should return to basic. Most important, and one of the mistakes in your schematic is replacing the intensity 50K pot to ground with a 68K resistor and 250K pot, just replace the 50K intensity pot for a 50K resistor. Your next mistake is, you left out the remaining 220K mixing resistor to the coupling cap of the PI. I think If you change this and return to the basic schematic, your amp should sound like a SR.

I don’t think the speaker is of major influence, as you already replied, because deluxes and twins also use 12 inch speakers.

I have been reading here for some years, but this is the first time I replied.

Luck with your build,

André

Thanks Andre.
There aren't any mistakes in my schematic per-say, it is all intentional.

The 250K master-volume pot in parallel with 68K is roughly a 53K load to ground, very close to 50K trem-pot of Fender. And, with the master-volume pot set to 7/10 (roughly) the pot is cutting the signal by 50% (like the two 220K mix resistors in a AB763 circuit)...and....the pots output resistance, at that point, is 62.5K. So, while not not identical to Fender value of 100k (The two 220K mix R's equiv output resistance), it's very close.

My idea was that I'd set the master to 7 (or so) to get the Fender 50% signal loss (from the two 220K mix resistors) and get close to the Fender sound, but then, turn-up the master past 7 to get a little more signal than Fender would have had going into the PI

Maybe I'll try adding an extra 100K in-series with the output of the master volume pot and see what that does.

Meanwhile, I still wonder what the effect of using the (mostly) Xicon caps is? They are polypropylene, and I have heard people say that polypro caps are harsher, and that polyester is sweeter (for Fender style amps). So, I am wondering if the Xicon caps should be changed-out.....

[Firestorm]

The 220K mix resistors don't attenuate signal the way you think they do because there's no load involved. Well, there is a subtle filtering effect because the normal channel 220K is in series with a .047 cap and then in series to the AC ground of the power supply through a 100K plate load; audible, but very subtle. But mostly it's just a series resistance which will kill off some high end. Might be just what the doctor ordered.

[pula58]

The 220K mix resistors don't attenuate signal the way you think they do because there's no load involved. Well, there is a subtle filtering effect because the normal channel 220K is in series with a .047 cap and then in series to the AC ground of the power supply through a 100K plate load; audible, but very subtle. But mostly it's just a series resistance which will kill off some high end. Might be just what the doctor ordered.

Thanks, I may try adding some series R

but, if you think about it, which I have, and I simulated it too (!), the normal channel and the 220k mix resistors kills about 5-6dB of gain for the other (vibrato) channel in a Fender AB763 amplifier

It's because the last amplifier before the phase inverter (in the "vibrato" channel of an AB763 amplifier) drives through its own 220k mix resistor, then into another 220K and the plate of the normal channel second triode stage. The normal channel and its 220K mix resistor is like the mirror image of the last stage (before the PI) in the "Wet" channel. So, 6dB of loss here. I am sure about this....simulated it using SPICE believe it or not.

[Firestorm]

6dB loss would be huge. I've disconnected the Normal channel 220K and got nothing that sounded close to that. Ignoring the .047 'cause it's pretty big, the normal channel 220K plus 100K are in series to AC ground, in parallel with the input impedance of the PI. 220K working into that looks like it would be more in the 2-3dB range, but not equally for all frequencies because of the .001 input cap. Did you, by chance, run a SPICE model of attenuation by frequency?

[pula58]

6dB loss would be huge. I've disconnected the Normal channel 220K and got nothing that sounded close to that. Ignoring the .047 'cause it's pretty big, the normal channel 220K plus 100K are in series to AC ground, in parallel with the input impedance of the PI. 220K working into that looks like it would be more in the 2-3dB range, but not equally for all frequencies because of the .001 input cap. Did you, by chance, run a SPICE model of attenuation by frequency?

Hey Firestorm, Here is a SPICE plot of gain (solid lines, in db) and phase (dotted lines) (I tried to eliminate phase from the plot, but,could not figure-out how to do it!). The red trace is gain of the vibrato channel as seen at the input of the phase inverter (grid of the PI tube, the one with a 82K load), the blue trace is the same circuit (i.e., the gain of the vibrato channel), except, that the normal channel has been added-in.

The vertical axis of the plot (on left) is in 6dB increments.

See what I mean.

[Firestorm]

Fascinating. I'm going to have to try and get measurements off an actual AB763 circuit w/ and w/o the Normal channel and see how well it matches. It's not that I don't believe you, it's just that I don't believe you. The hump just over 100Hz and the dip at around 500Hz seem odd to me given the .001 input cap (You did model with that value, right? What source impedance did you use?). Feels like it should be closer to shelving, but not sure at what frequency.

I'm interested because builds that attempt a single Fender channel never do sound quite right. As they say, you change one thing, you change everything.

[pula58]

The hump just over 100Hz and the dip at around 500Hz seem odd to me given the .001 input cap (You did model with that value, right? What source impedance did you use?). I'm interested because builds that attempt a single Fender channel never do sound quite right. As they say, you change one thing, you change everything.

The hump, and dip, is the tone stack I believe. I have the T, B and M and Volume set to: 6,4,7 and 4 respectively. All values are stock AB763.

Surprising, isn't it. I don't believe me either....

[tictac]

the 820/50 ohm feedback network is not correct for 6L6 output, you'll get less feedback which means a bit more gain but less headroom, should be 820/100 ohm if you're going for a stock Super Reverb tone.

Also some people find metal films to add brightness to an amp which is why some prefer carbon comp resistors on the signal path and plate load resistors.

My $.02
TT

[pula58]

the 820/50 ohm feedback network is not correct for 6L6 output, you'll get less feedback which means a bit more gain but less headroom, should be 820/100 ohm if you're going for a stock Super Reverb tone.

Also some people find metal films to add brightness to an amp which is why some prefer carbon comp resistors on the signal path and plate load resistors.

My $.02
TT

Thanks...but..I believe that the Super reverb uses 100ohms/820ohms with a 2 ohm load. I am using a higher ohm (8 ohms) load. The Blackface single speaker Fenders (Vibroverb, and the non-reverb Pro AA763) used 820/47 ratio. That's why I picked that. The Deluxe reverb also uses 820/47. And, when I first played the amp with 100 ohms it did not seem like it had the "right" volume I have come to expect from this type of amp at vol=4. The 50 ohms sounded/felt much better to me.

Thanks for keeping me on my toes though!