Valvestate 8040 neg vdc on output

[T9c]

I started repairing this amp by changing out the 12ax7 that was cooking the heater resistors. Everything fine except a slew of bad pots, so I ordered direct replacements & installed and put the pcb back in the chassis, ground wire connected to chassis, dummy load connected.

All of a sudden, the power amp chip TDA 1514 started cooking, so I flipped the variac off, but the damage had been done-obvious crack appeared on the chip's top.

Ordered some replacements from Dalbani and installed one. Now I have -23vdc on the output (pin 5) and -13vdc on the input (pin 1) as well as -23vdc on the tied together pins 2&3. The +/- rails are 23vdc/-23vdc. I checked thoroughly for loose connections, none found. Checked ZD3, the 6.8v zener, good. Replaced it with a 6.2v (all I could find today)...no change. Haven't put a different TDA chip in yet, but I'm curious why there's -13v on the input, and neg vdc on the output, of course. Dummy load is disconnected, of course till I can remove the dc on the output.

Anyone have any ideas?

[nickb]

Pull the TDA1514.
Remove dummy load
Measure the resistance from pin 5 to 0v: should be about 23K - if not find out why.
Power up and check all the pads of TDA1514, 1=0, 2 & 3= -17V, 4= -23V, 5 = 0V, 6=+23V, 7=+23V, 8 = 0V. If not right, find out why.
Replace TDA1514 making sure you have it in the correct way round - metal tab goes to heatsink. Don't connect dummy load. Connect meter pin 5 to 0v and slowly bring up power with variac. I would expect the meter to read close to 0 at all times. Only other thing I can think of is the mute circuit on pins 2 &3 not operating resulting in a DC transient level on the output. If no transient and you have 0 on the meter at full volts then you should be good to go.

[T9c]

Thanks for your help Nickb. With pulled TDA, I initially have +12vdc on pin 5 on pwr up. When I power down the voltage slowly drops to -12vdc, then slowly climbs back to zero. All other pin voltages are as you stated.

The resistance between pin 5 pad & circuit gnd is around 1meg. Why should it be 23k?

Pull the TDA1514.
Remove dummy load
Measure the resistance from pin 5 to 0v: should be about 23K - if not find out why.
Power up and check all the pads of TDA1514, 1=0, 2 & 3= -17V, 4= -23V, 5 = 0V, 6=+23V, 7=+23V, 8 = 0V. If not right, find out why.
Replace TDA1514 making sure you have it in the correct way round - metal tab goes to heatsink. Don't connect dummy load. Connect meter pin 5 to 0v and slowly bring up power with variac. I would expect the meter to read close to 0 at all times. Only other thing I can think of is the mute circuit on pins 2 &3 not operating resulting in a DC transient level on the output. If no transient and you have 0 on the meter at full volts then you should be good to go.

[Enzo]

With the IC removed, the only path to ground from pin5 is through the two resistors that make the line out from the speaker signal. I can't read them, but I'll assume they add up to that amount.

WHile you in there, check just to the right of the pin 5 on the drawing, dropping down to ground are R66 and C54 (if I read right). Make sure 10 ohm R66 is not open. Those parts are your stability network, and without them, the IC can oscillate to death.

Pin 5 starts out with +12v because of that cap just above it. As it charges, teh voltage drops back to normal.. The question is, does that 12v get down to about zero volotage after a bit. When you power down, the charge on that cap slowly ebbs away, and that is why you get the -12v.

[Jazz P Bass]

The resistance between pin 5 pad & circuit gnd is around 1meg. Why should it be 23k?

I believe Nick is refering to the line out series resistors. R36 & R37. (41K)

[T9c]

Thanks Enzo. Yes, pin 5 does get stable to ~100mv after 30-45 secs. The 10 ohm 1/2w is good, checked it 1st as it's a little browned from previous heating...I went ahead and replaced it with a flameproof 10 ohm.