Vibrochamp

[lowell]

This amp came in and would compress and fart out when cranked. I found the cathode resistor and bypass cap to be the issue there. I replaced them. I checked the bias and the 6V6 is at 100ma at 290v... which equates to 30W. All voltages seem quite low throughout the amp. The main issue though is that the plate is at 290v and the screens are at 303v. I replaced the 1k dropping resistor to the screen supply and still - 303v. I am suspecting the OT but I measured the primary DC resistance and it's 240ohms... not sure if that's right but either way it's way less than that 1k dropping resistor to the screens so I'd think the screens should be lower.

[nosaj]

What amount of AC do you have on the load resistor when feeding a sine wave into it?

jason

[Tom Phillips]

... I replaced the 1k dropping resistor to the screen supply and still - 303v. I am suspecting the OT but I measured the primary DC resistance and it's 240ohms... not sure if that's right but either way it's way less than that 1k dropping resistor to the screens so I'd think the screens should be lower.

The plate is drawing a lot more current than the screen. Therefore, there is more voltage drop across the lower resistance. (Ohms law rules)
Pull the 6V6 out and the voltages should rise to above normal operating point. Next you need to determine why the 6V6 is pulling do much current at idle. Check the cathode resistor and cap. Even new parts can go bad or be incorrect values. Also check the coupling cap to the 6V6 grid. Is it leaking DC?

[lowell]

I meant to post back earlier but right after I posted I followed Tom's line of thought. I was wondering how the 6V6 was drawing so much current. So I checked for DC voltage on it's grid and it had 2-4volts. I then followed over to the 220k and its ground path... bingo the ground wire to the brass plate had completely lifted off it's solder pad. All good now. Thanks guys.

[Bruce / Mission Amps]

I meant to post back earlier but right after I posted I followed Tom's line of thought. I was wondering how the 6V6 was drawing so much current. So I checked for DC voltage on it's grid and it had 2-4volts. I then followed over to the 220k and its ground path... bingo the ground wire to the brass plate had completely lifted off it's solder pad. All good now. Thanks guys.

Ha ha... that's called zero bias!! Good catch.

[Jazz P Bass]

I agree.
Good catch, Lowell.

[Tom Phillips]

...I was wondering how the 6V6 was drawing so much current. So I checked for DC voltage on it's grid and it had 2-4volts. I then followed over to the 220k and its ground path... bingo the ground wire to the brass plate had completely lifted off it's solder pad.

Some readers may wonder what was really going wrong with the circuit in this case.
When it is working properly with no signal applied the 6V6 cathode is a little over 20 Volts positive with respect to ground and the grid is at about 0 Volts with respect to ground. The tube is properly biased because the grid voltage is negative with respect to the cathode voltage.

When the 220 kΩ grid leak resistor lost its ground connection the grid voltage floated up and the tube ran away into an under biased condition. The actual voltage on the grid was probably much higher than the 2-4 Volts that Lowell measured. That’s because the meter’s input impedance (Probably 1 Meg Ω) provided a grid leak path when it was connected. This is one of those cases where the test equipment can change the value you are trying to measure. If the plate current was being monitored as the voltmeter was connected to the grid I’ll bet that a reduction in plate current would have been observed.

Anyway...the problem was fixed but understanding the fault mechanism can help with future troubleshooting and design.

Regards,
Tom