Memphis 16-MCD500

[cajones65]

Hi y'all, I am looking for help on this Memphis, Q255 and Q256 are getting extremely hot. (IRF640). I have replaced the driver transistors on the driver board. If I leave the outputs out of the board, I don't get the high current draw, and have a good signal. By the way I have replaced the 4 9640's and 2 940's.
I have also replaced the gate resistors for all the fets.

 

[Perry Babin]

Did you check the drive signals at the outputs before installing them?

 

[cajones65]

Yes sir I did, signals were good, low side and high. Pulled one end of inductor out to see if it was past the outputs. Still pulled high amps.

 

[Perry Babin]

Are those outputs the ones you installed or the originals?

Do you have any IRFB31N20Ds?

 

[cajones65]

They are the ones I installed, and I do have the IRFB31N20D's in stock. I replaced with the same part number as the ones that I removed. I originally had one shorted in each bank.

 

[Perry Babin]

I would try the 31n20 in place of the 640. I was wondering if the 640 should be a 640N.

You should try to use FETs with matching date codes (for those in parallel) for better reliability. The ones in the photo are a mix.

 

[cajones65]

Perry the picture I posted is of the originals ones. The ones I have in now are from same batch. I take pictures when I open them up for the first time and place in a folder for future references. I will replace the 640's with the 3120's. It does look as though this amp has been worked on before.

 

[cajones65]

Perry I replaced the 640's with the 3120's and I have the same results.

 

[Perry Babin]

I'd suggest removing the outputs and re-checking the drive with a cap connected from gate to source. The value isn't critical but I'd try something like a 0.01uf cap.

 

[cajones65]

Perry I have done that. I am getting a signal, it is smaller and not drawing excessive current.

 

[Perry Babin]

What does the signal look like?

 

[cajones65]

Perry, the first picture is signal with out the cap, and second is with the cap. These are the 31n20's

 

[Perry Babin]

That's not good. The difference should be minimal. A slight slope on the falling edge is all I'd expect to see with the cap. Do the same on the high side to see what you get.

 

[cajones65]

Perry this is the high side, I am using a .1uf cap by the way, not a .01uf.

 

[Perry Babin]

0.1 may be too much. Do you have something between a 0.001 and a 0.01?

 

[cajones65]

Ok this is a .01uf, high side and low side.

 

[Perry Babin]

Do both the high and low side go back to within about 1v of their respective rail?

 

[cajones65]

Perry are you talking about with cap in or pulled out?

 

[cajones65]

Perry I rechecked the cap I had, wasn't no good. I still have a signal like the first ones I showed you. With the cap out I have a rail of -49.6 with it in I have -50.4. High side without 51.2, with cap 52.8.

 

[Perry Babin]

I don't know if I have a similar amp in the shop but I'll check tomorrow and see if I can find a reliable test. This one should have been but I'll try to confirm.