Magnatone 440A Problem. Early onset of Bad Sounding Distortion

[Tom Phillips]

Hi All,

I have done a thorough checkout and restoration of a Magnatone 1963 440A. Both the cabinet and the electronics had suffered lots of bad workmanship over its 60 year life. As I worked my way through the true point-to-point wiring, I found many cold solder joints and some inappropriate parts such as a 1000 Ω power amp cathode resistor in place of the stock 250 Ω part.

A copy of the Maggie 440 schematic version that matches this amp is attached. (The markup notes shown are as found on the web.)

Currently, as the title of this post states, the amp exhibits early onset of bad sounding distortion. I have isolated the problem area to the first triode section of the 12AU7 in the vibrato modulation circuitry. Scoping the signals at the 12AU7, I find that the grid pin 2 signal is clean and the plate pin 3 signal is distorted. Re-amping and listening to those signals verifies the same crappy distortion that I hear at the output when driving a speaker.

I’ve tried three different 12AU7s with no change. All the components in that area have been verified to be good or replaced. I tried disconnecting the two varistors where they connect to the junction of the 820pF and 470pF caps. That made no difference at all with regards to the distortion problem.

All the voltages are reasonably close to those listed on the schematic EXCEPT the cathode (Pin 3) of the 12AU7 is 230V versus the 170V shown on the schematic. The plate pin 1 measures 264V vs. the 280V shown on the schematic. Re-checking the associated resistor values and checking the caps for DC leakage does not reveal any culprits. The 35% high cathode voltage is my only current clue.

I’m at a frustration impasse now and would appreciate any ideas to help solve this problem and/or clues to something I may be overlooking.

TIA,
Tom

[Helmholtz]

As the cathode voltage is high, the grid (pin 1) voltage must be high as well by about the same amount, as cathode voltage follows grid voltage.
You can't directly measure the grid voltage with a DMM.
Instead check two 4.7M bias voltage divider resistors as well as the 0.01µ coupling cap from 6EU7 pin 7.

[Tom Phillips]

...check two 4.7M bias voltage divider resistors as well as the 0.01µ coupling cap from 6EU7 pin 7.

I had checked those before. However, I checked them again because your comment made perfect sense. They still checked out good but, during the process, I found a wiring mistake that must have happened when someone replaced the rectifier tube socket before the amp came to me. The NC terminals were used as various wiring tie points and it turned out that both of the two 4.7M bias resistors were wired to the +B power rail. They are not physically near each other. One is at the 12AU7 tube socket and the other is at the other end of the chassis near the power supply.

Yay!

[Helmholtz]

Wow.
So the tube was biased into grid conduction. No wonder it distorted.

[The Dude]

Wow is right! B+ for tube bias.

[Helmholtz]

Fortunately the grid current drops the grid voltage to just above or the same as the cathode voltage.

[Tom Phillips]

Wow. So the tube was biased into grid conduction. No wonder it distorted.

The distorted sound was very similar to a blown speaker. Not at all pleasant.