Help fix amp - rail voltage gone crazy

[Perry Babin]

The highlighted components in the attached image show the part of the circuit that's supposed to roll back the pulse width when the negative rail voltage becomes excessive. If all of the grounds are OK, these would be the next components that you need to check.

[I Am An Idiot]

Can you post a picture of the bottom side of the circuit board? The power supply end and pay special attention to the 2 corners of that end of the amp.

[manina]

I swapped the inductors and, surprise, the negative went down.

Now, with the VR951 all the way down, the voltages are +25,5V and -34,4V.

If I increase the positive to the expected working voltage of +30 the negative goes up to 39.

Now the gap is "only" 9V and the negative voltage is influenced by the VR951.

@Eva - I tried to send the PDF but it bounced back. I'l try again later in the morning.

Pictures will come shortly.

[Eva]

Great! Try to measure these inductors. The one producing a too high rail voltage probably has shorted turns or the core is defective. Look carefully for scratched enamel due to rubbing of adjacent turns.

[Perry Babin]

Why not just replace them with a straight wire to remove them from the equation?

[Eva]

Because the supply is regulated. The reansformer produces +/-50V or so without them.

[manina]

I have taken the inductors from another 922 amp and applied the 47 Ohm/10W load to both rails.



+34,3v and -34,4V

Much better

I measured all the 4 inductors (2 of this amp and 2 of the twin one) with an LC meter and they all read the same value - can not understand it but as long as it works, is OK.



I know it's not of much help however the pin 9&10 voltage of the IC951 is now 3.84V, in line with the other working amp.

The bad news is that the VR951 is all the way down so I can not reduce the voltage to the desired 30V - something is still not 100% OK but way better than before. I think I'll re-mount the Sanken and see if it improves.

[Eva]

Intermittent shorted turns? When you measured the inductor the short could have disappeared. What was the value shown by the meter?

A brute force solution is to rewind the bad one with new magnet wire.

[Perry Babin]

If it's regulated, it would regulate it with or without the inductors.

It's likely that the negative half of the regulator isn't working properly. If Q953 isn't conducting, it's allowing R967 to pull the voltage on pin 16 down which would tend to make the regulated voltage higher than it should be. I'm assuming that the pot should be relatively close to the mid point when the rail voltage is 35v.

The service manual states that the rail voltage is supposed to be ±35v.

[manina]

Originally Posted by Perry Babin

If it's regulated, it would regulate it with or without the inductors.

It's likely that the negative half of the regulator isn't working properly. If Q953 isn't conducting, it's allowing R967 to pull the voltage on pin 16 down which would tend to make the regulated voltage higher than it should be. I'm assuming that the pot should be relatively close to the mid point when the rail voltage is 35v.

The service manual states that the rail voltage is supposed to be ±35v.

Right it's +/- 35V. I had in mind 30V - don't know why

I'll have a look to the section you highlighted in the green square later.

Eva - the inductors are 0,135 mH (more or less). I like the brute force solution