Randall RG80-112SC Hum - Solved!

this thing has a rep for bad 120 hz hum,

pwr section needs decoupling to the diff pair, no?

voltage chart and an attempt to decouple attached>

> decoupling to the diff pair, no?

Not really. Signal across R41 to + rail drives Q10 to + rail, all local reference. This works because a collector is "infinite" impedance and can be swing up and down with nearly no change in R41 current.

Overall it looks like a ton of Fishers I used to fix, and they didn't buzz.

I would start with the ripple-path through rectifiers and main filter cap. If ground for the rest of the system is tapped at a bad place, 120Hz is injected all over. This usually means a massive ground bus between the two mail filter caps, with PT CT on one side and all-else grounding on the other side.

Your proposed 330r dropper is in-effect a 150% change in the DC value of R41. DC NFB will "correct" this by reducing Q8 current (more to Q9). While this is tolerable for an instrument amp (THD will rise but nothing horrid), R41 was (for once) nicely selected to keep Q8 Q9 balance (for nominal supply voltages) and it would be a shame to change it 1.5X. (If it were changed 2X, Q9 would turn-off, Q8 emitter sees 10K, gives super-low gain, amp may never bias or pass signal).

C31 is moderately importance and if it is as old as this drawing and design, is due for freshening.

C35 could fail without obvious problem, but less audio gain/NFB and poor peak output drive.

HAH! There is no audio NFB through R45 R46, the usual audio NFB path. Essentially all audio NFB is through speaker and R63. If you don't have a few-Ohm speaker connected, voltage gain is insane and NFB does not reduce supply ripple. He was obviously aiming to imitate classic tube amps with high to very-high output impedance (damping factor like 0.1).

yeah decoupling resistor made the problem worse,

put in the buss bar, not much help, put in some new tip32c trasistors, no more hum, but nno more gain, pwr supply ripple way down due to reduce load of cutoff trasistors,will trim this thing up for the new transistors ad see what gives, notice the thermistor PTC across one of the diodes in the ref string,

still troubleshooting the 120 hz hum in the output circuit,
found a slightly different schematic from a Randall 100ES,

mainly the resistor to the base of the 2nd diff transistor thru the cap to ground (circled in red)
this RG80 amp has a 330 ohm
the 100ES has a 4.7Km don't know if changing 330 to 4.7 k will improve hum, will give it a try,

resistors that do the most as far as setting up the operating point are the 1K and 1.5K circled in green, they determine current flow down the string that sets up base voltage/current of the tree transistors m
the 250 ohm pot//47 ohm resistor simply vary the current to trim to bias the drivers, output transistors are cut off til a good signal comes thru, drivers run warm til a signal comes thru and gets them to conduct better, the thermistor regulates the heat of the drivers by increasing R and thus reducing current thru the bias string,


that follow, changing the 100 ohm emitter resistors to different values does not do much as most of the current simply flows thru the e-b junction of the next transistor,

changing the 680 ohm resistor to the diff pair does not do much as the tip32 is jammed into saturation unless the 680 is reduced to a much

do not know the purpose of the 22 ohm across D8 o the RG80, it is not used in the 100ES

so the most important resistors in the whole output stage are the R47 and R48 1 watt resistors (RG80)


here is the newer schematic, uses MJ15015 driver and outputs but has the same TIP 31 and 32 amps,

OT: Don Randall worked with Leo Fender way back when,

> tip32 is jammed into saturation unless the 680 is reduced to a much

It's all interconnected. NFB from the output should shift the input pair until current in 680r is correct for TIP32 to pull its collector halfway between the rails.

If it don't, it suggests an output device is dead. (Or one of the input pair.)

ok finally an answer to the humm probelm on this amp, one of the speaker jacks was missing the fiber washer, which means that the jack was grounded to the chassis (red) which means that the feedback loop, aparently a hum canceling loop, was effectively bypassed.

just those few millivolts was enuff to cause the problem, as the nfb gets iserted into a high gain node,

now the amp is quiet, we also increased the value of C31, the cap on the 4.7K on the input pair which might have decreased hum even more,

learned a lot about direct coupled amps so the pain was worth it, thanks for the help,

Interesting do you think the amp was worked on and the washer was lost?

C34 is interesting it removes the signal feedback but keeps the DC feed back.  So it acts like a current feedback amp.
PRR posted before about this with R45 and R46
I wonder if this amp could be adjusted for mixed feedback?
Was this built before the other guitar amps that use current feedback?

links from a very good website
http://sound.westhost.com/project56.htm
http://sound.westhost.com/z-effects.htm

yes, somebody else had worked on the amp , looks like there will be ac feedback on the same circuit as the hum cancel circuit once you start playing,

the neat thing about this amp is you can mix the two channels (with the foot sw unpluged) and also the compresor when you pull the treble knob out,

thanks for the links! lots of good stuff onn that guy's site!

cj