Sound Stream SA245 repair help

[grjr]

Hey guys, I'm working on an SA245 five channel amp. The amplifier blew some of the outputs and a MPSA14 driver along with a 10 Ohm resistor that connects the ground of the audio section to the PS ground. I've removed most of the outputs, along with the driver boards for them. One of the outputs turned a section of the PCB to charcoal.

In the power supply there are a couple of 75 ohm resistors that tie the gates of the switching MOSFETs to ground that have been discolored from heat. The SG3524 is switching the MOSFETs at around 57.5 kHz. According to the datasheet and with Rt = 1.91 k and Ct = 4.7 n the IC should be running at around 145 kHz. Could the IC be bad? What else should I be looking for? I've included some pics of the amp in the link below. Thanks


http://picasaweb.google.com/grjr2000/Sa245#

[Perry Babin]

I calculated an oscillator frequency of 131k. That would make the switching frequency ~65k. The tolerance in the components could easily make up the difference between 57k and 65k.

It appears that the traces where the board burned are directly on the opposite sides of the board. If so, you need to remove the copper from the top side until you have clean fiberglass. Then jump the connection from the trace to the terminal of the transistor. If you have copper on both sides of the burned area, the difference in voltage will pass current through the carbonized board which will result in further damage/burning.

[grjr]

I'm waiting to borrow a dremel bit from my brother to grind away the burned PCB. Perry did you see the waveform being fed to the switching MOSFETs in my pictures? What could be causing the distorted square wave?

[Perry Babin]

There is capacitive coupling between the drain and gate of the mosfet. When there is a sharp transition in the voltage on the drain, you can often see it on the gate. If you use both channels of the scope and monitor the gate with one channel and the drain with the other channel, the pulse and the drain's transition from low to high should be synchronized.

Since it only has the resistor to pull the gate down, the pulse is somewhat more than you'd see if it had a PNP transistor pulling it down.

Are the FETs the original parts?

[macknorris]

I have one of these that is working if you need any reference numbers.