AB763 Concert/Bandmaster with pentode channel and Trem-o-nator

[bmccowan]

I have a couple of question, but first let me describe the project to ease the mind of any that think I hacked up an important amp.
I happened upon the chassis of a BF Concert. All iron was gone, the heater wiring was shorted and burned, the filter caps had been changed to some bizarre series and parallel combination of 10 caps, and the chassis appeared to have a lot of smoke damage. The pots and most of the circuit were intact, and the faceplates were nice. I was tempted to use scrounged iron, but broke down and bought new iron. I also located a Bandmaster head cabinet that fit perfect. I think the OT and the speaker config is the only difference between that series Concert and Bandmaster.
After cleaning it up, installing the new iron and tubes, I got it working stock first. All good except the trem which had a broken roach.
Mods:
Normal channel to 5879 pentode.
Dropped the voltage a tad with larger power rail resistors.
Sluckey's Trem-o-nator
Questions:
1. No surprise, but I found that the Fender tone stack has too high an insertion loss for a pentode with no recovery stage. So I wired the channel raw and tried various known 5879 configurations. The one I like best is one from one of the many Gibson GA-40 versions. Besides the 510K plate resistor typically used on GA-40 trem channel, it has a 1M screen resistor with a .25uf bypassed 250K resistor to ground. The cathode is typical 22uf/1.5K. I read a thread on amp garage about the unusual screen resistor configuration but did not quite understand. To stabilize the screen voltage? Can any shed some light on this?
2. Mixing resistors - I experimented with different values here. Matched and not. Right now I have 82K on the pentode and 180K on the triode channel. Trial and error worked fine, but is there a better way? As long as there is enough resistance to prevent unwanted interaction, is it important that the mixing resistors be the same value? I think not, but
3. Trem-o-nator I used the Vactrol and values recommended by Steve. I first wired it to the triode channel - sounded good. I then wired it post mixing resistors to get trem in both channels. Sounds good in both channels, but curiously the change improved the tremolo on the triode channel in addition to providing tremolo on the pentode channel. This tempts me to take the "are you insane" poll, but maybe there is a reason? the non-matching mixing resistors?
4. I like the raw pentode channel for variety and will keep it that way for a bit. But wondering about alternatives. I might try the Framus mid control which I have used in other amps - ideas? 
 

[DummyLoad]

well, with a 510KΩ plate R, what did you expect? drop plate R to 68K-82KΩ and try it. cool down bias with cathode R, it will sound better. keep Vg2 < Va for best overdrive tone. 


--pete

[2deaf]

Better check those input jacks wiring.

[bmccowan]

Thanks Pete, but I am not sure what you are referring to when you ask, "what did you expect?" As the post states, I tried many different typical 5879 configurations, and likely will try more as its basically an experiment right now. The one in the schematic is straight from a GA-40. I have tried multiple plate/screen/grid combinations, including in the range you suggest. None sound bad, the current one sounds very good, I was not complaining about the tone. I was, however, looking for some knowledge/help on that unusual screen resister and cap configuration, also about selecting mixing resistors, and some ideas for lower loss tone stack options.
2deaf - thanks - a lazy copy/paste on the input jacks.

[shooter]

the schematic is straight from a GA-40

unusual screen resister and cap configuration

In the '70s, "my guy" was a Nam vet, used to work at the Gibson plant, just saying.............

[bmccowan]

I hear ya Shooter. I have '72 J-50 - great tone, shoddy workmanship.

[DummyLoad]

have you tried using the LTPI as the mixer? you'd want to ditch the NFB if you do. thinking something similar to matchless DC30.

on the pentode channel, the plate to wiper of the volume pot (via .022uF)  is going to shave a lot of low frequency as the wiper sweeps CCW, esp. with a .022uF cap. 

there's a spare triode in V3? perhaps use it for a recovery amp for pentode channel? reverse the two - use the unused triode to drive TS, pentode is recovery, sum channels as you do now, or use the matchless DC30 LTPI plan?

typical summing R values range from 220K to 1M - yours work. less loading with larger summing R values. you are near 2:1 with 180K/82K - try 220K/470K? 270K/470K?


--pete 

[bmccowan]

Hi Pete,

there's a spare triode in V3? perhaps use it for a recovery amp for pentode channel? reverse the two - use the unused triode to drive TS, pentode is recovery, sum channels as you do now, or use the matchless DC30 LTPI plan?

Funny you should mention that. I wired up the pentode channel first, and then the Trem-o-nator. And then thought - oh look at that, I have a gain stage available. Since I am experimenting, I will likely try that. But the LTPI intrigues me too - that had not occurred to me. I built a Clubman earlier during my Covid amp building craze, and I love the tone of that amp.
Good info on the mixing sum values. I will try the combinations you suggest. I should read up on that in Merlin's book maybe? No electronics background here - Political Science degree - gee, I wonder if anything is going on in politics about now?

[shooter]

gee, I wonder if anything is going on in politics about now?

that's such a 20th century idea

[bmccowan]

Yup - I was in HS in 1968-70. Never thought I'd see anything like it again. We live in interesting times.

[tubeswell]

4. I like the raw pentode channel for variety and will keep it that way for a bit. But wondering about alternatives. I might try the Framus mid control which I have used in other amps - ideas?

You can get some sort of tone 'control' manipulating the g2 bypass by putting a 250k to 500k pot in series between the Cg2 and the ground - and decreasing Cg2 to 0.1uF or 0.05uF (depending on where you want the LF rolloff). The idea here is to alter screen current feedback at treble frequencies to create a variable treble boost.

If you're going to try this, you'd probably ditch the 220k and play around with the value of the 1M Rg2 (to get Vg2 into the 60-80V ballpark) to get a better rolloff from the tone control.


There are also other ways of fiddling around with screen current feedback frequency rolloff points. YMMV