Magnatone weirdness

[Valvehead]

Working on a early 50s ? Varsity 197. Its been re-capped and im seeing that all the resistors are in spec except 1. Right after the 6sj7 theres a 500k on the scem. I will attach a pic . Out of the amp it tests 234k and has red/yel bands (220k ? ) . So , the amp sounds ok but gets farty/ flubby when you crank it up. I thought i would replace this resistor...put in a 470k. This completely kills the signal. Double checked everything . Any ideas? This part looked like it was factory. ??

magnatone_M-197-3-V_PM.pdf

[Jon Snell]

I would, without even testing anything, replace the 10M on pin 4 of the 6SJ7. High value older resistors are notorious for going even higher in value.
A 6SJ7 won't be happy with 500k anode load. 220k is much nearer the mark. Re fit the 220k. 234k is well within spec of a 10% tollerance.
Check the voltages look about righ around the 6SJ7 and 6V6.

[Enzo]

Looks like red/black/yellow to my eye. SO 200k. The amp worked with the 200k, I agree, put it back in, or really use a new 220k.

I am OK with the 10meg for grid leak bias, even if it goes higher, it still should allow the grid to collect a voltage.

Check the plate and screen voltages on the 6SJ7.

On an amp of this age I expect the three electrolytics to be worn out. And the four "film" caps are likely paper caps and are almost assuredly leaky. I'd worry about those caps way before I worry that a resistor is off spec.

[Valvehead]

The schematic calling for 500k is what caught my eye ...maybe a different version of the amp? The model numbers match

[Enzo]

Beware "model numbers match". I have one Gibson model that has THREE different published schematics, plus from examples through my shop I have THREE hand drawn schematics that are all different and different from the published ones. SIX different schematics for one model name.

I suggest schematics should be considered approximations. And remember these are just guitar amps, not precision equipment.

[Helmholtz]

DC voltages from both ends of the plate resistor will show if the stage is working properly and which resistor value results in best biasing.

[Valvehead]

Beware "model numbers match". I have one Gibson model that has THREE different published schematics, plus from examples through my shop I have THREE hand drawn schematics that are all different and different from the published ones. SIX different schematics for one model name.

I suggest schematics should be considered approximations. And remember these are just guitar amps, not precision equipment.

I was looking through very similar pre 1953 Magnatone schematics and some use 47k , some use 100k at the inputs of the amps. Mine has 47k 's

http://magnatoneamps.com/schematics.html

[g1]

I don't think that is an 'official' schematic. I would think if someone was doing a clean up on the original schematic, they mis-read the 200 as 500.

[Valvehead]

Im not sure what voltages i need here. Here are some measurements ---

6sj7 plate is 82vdc , and 230v on the other end of the 220k i just installed

the 6V6 has 260v at plate

B+ out of the rect is 310v

bias calculates to 39ma for the 6v6

all the caps are new , all the resistors are new except the 1k/7w , 250/5w and 10k/2w , which test ok

[nosaj]

Beware "model numbers match". I have one Gibson model that has THREE different published schematics, plus from examples through my shop I have THREE hand drawn schematics that are all different and different from the published ones. SIX different schematics for one model name.

I suggest schematics should be considered approximations. And remember these are just guitar amps, not precision equipment.

Skylark by chance?

nosaj

[The Dude]

I'd monitor B+ (preferably on a scope) and see if it's sagging substantially when the amp is driven hard.

[Helmholtz]

6sj7 plate is 82vdc , and 230v on the other end of the 220k i just installed

Means 148V drop across the plate resistor, corresponding to a plate current of 0.7mA.
No wonder you get no decent signal with a 470k resistor as that will drop most of the supply voltage, leaving no useful plate voltage.

82V plate voltage is rather low/unfavorable. A lower resistor value (dropping maybe 100V) would increase headroom.

[Enzo]

Seems to me it was GA17 or GA19, if failing memory serves me...

[Helmholtz]

bias calculates to 39ma for the 6v6

Is that plate or cathode current? What is the 6V6 cathode voltage?
You could lower the cathode resistor to 220R, aiming for a cathode current of 50mA.

all the caps are new

Does that include the cathode bypass cap?

Best measure output power.
Should be around 4W.

[Valvehead]

Heres something...check out how this 6sj7 is wired compared to the Varsity scem. Should i try this? Mine is close but pin 5 is going straight to ground

http://magnatoneamps.com/schematics/dickerson_1946.pdf