blown channels

[jringels]

Next problem...
I have my AMP9, build#2 with paralleled outputs. While hooking up the outputs, I missed my screw terminal on chanel 2 and had the (-) wire under the board, touching the output capacitors on the bottom. I did not see this, until I turned it on. Uh, no sound can out from those 2 channels. Since I had ch1 and ch2 paralleled, the output from ch1 would have reached the capacitor under the board. I hear nothing from these channels, unless the pot is at full volume, then I just barely hear the sound. Ch3 and Ch4 work fine.
Question, what did I likely fry, the caps, the chip, other? Any suggestions for repairing this? Or do I now own a 2 channel amp9?

Thanks!

[V-bro]

What voltage do you measure on the outputs?

It seems only the chip could be the cause, the diodes only clamp the outputs to prevent overshoots, resistors usually look and smell burned when they go, caps same story. You obviously have 5V....

You could measure for zero ohms between the pads of the diodes on the bottom (chip side...) and the speaker outputs. They should conduct between there, they are basically aligned in the same way...so easy to check one by one. When they all conduct it's most likely that all the components in the output section are ok...

------X-max. can be several meters on any driver, too bad it can be done only once...------

[V-bro]

Here's the idea:


Basicallly you measure from the diodes to the outputs, but I also branched the arrows towards the chip pin holes...

Lucky I just received a new AMP9

------X-max. can be several meters on any driver, too bad it can be done only once...------

[jringels]

Hey, thanks. I do measure 0 Ohms between the pads as you describe, so the output components must be ok, must be a problem with the chip. If C131 or C132 were destroyed, would that cause my symptoms? It is strange to me that I hear nothing on those channels unless the volume is at max, then I only hear weak sound. I do not know how to measure the volts output, or the Watts, though I would like to. I do have 1kHz wave on my ipod.

[V-bro]

The output offset voltage can be measured with a DMM on mV setting. You should short the inputs for this measurement. It should read lower than 150mV, preferably 0,00V...

I had a dead chip (dead output) that produced railvoltage on the output...(27Vdc!) So be carefull with your speakers! You can remove C131/132 as long as you keep a load on the output, but it's easier to measure them for shorts (unlikely...) They decouple the outputs, so they connect to ground, normally the overcurrent protection in the chip would cut in if such a thing happened. This doesn't happen because all channels would cut out...

The faint sound you hear would rather indicate that there's something wrong on the inputs, but this would be unlikely regarding what happened....

Next time you'd better solder the input and the output bridges on the paralleled channels, the chip also heats up from a loose contact on one of the outputs...

Do you have the input clamping diodes installed on this one?

------X-max. can be several meters on any driver, too bad it can be done only once...------

[V-bro]

Oh and when measuring DC offset make sure you choose DC mV setting, not AC!

------X-max. can be several meters on any driver, too bad it can be done only once...------

[jringels]

I do have the input diodes in.

So, I should short all the inputs (+) and (-) to AGND, then turn the amp ON and measure the mV on each pair of outputs, J1-4? I should see 0 to 150 mV on each output.

[V-bro]

Yep...

------X-max. can be several meters on any driver, too bad it can be done only once...------

[V-bro]

It is even better to hang some load on the output, like a >8 ohm resistor or just a lousy speaker.

------X-max. can be several meters on any driver, too bad it can be done only once...------

[jringels]

Build2: Well, I made the measurements but I don't know what they mean... This is the Build where I think to blew channels 1 and 2 (were paralleled).
I unparalleled all channels and I did the Hum test and each channel hums when I touch it's input pin. I also measured the mV on each channel output. Here is my results...

Inputs Ungrounded
Ch1, 14mV (then 19)
Ch2, 18mV (then 22)
Ch3, 9mV (then 18)
Ch4, 12mV (then 22)

I then Grounded all the input (+) together, and connected them to AGND and input (-). Here are my results:
Inputs Grounded
Ch1, 23mV (then 28)
Ch2, 26mV (then 31)
Ch3, 20mV (then 22)
Ch4, 22mV (then 24)

Again, each time I measured, the value increased, maybe related to the chip heating up. I did have an appropriate heat sink on the chip, and no speakers or load on the output.

So, does this info help diagnose the problem? Is the chip bad? Thank you.