Just shooting the breeze Carvin 400FET

[Enzo]

The output devices are current amps, not voltage amps, so that is a small concern in my view. If you want the amp balanced, apply white noise (or pink) mono to the left and right, then adjust so the sound is centered.

[greenmeanh1]

So the IRFP254 mosfets weigh exactly 6 grams. Also the lettering is etched into the face of the device. It will be interesting to see how much the china versions weigh when they come in. My understanding is the Digikey IRFP254 replacements are from china as well. I have decided to go with these instead of the STW33N20 originally spec'ed with the units mostly cause the IRFP's are what was populating the boards.. All the STW33N20's i have been able to find anyway are pulls with shorter leads.
I have gone through one of the amp boards and found some things.
Gate resistors on the good working side are all sort of within the 150 ohm spec. One is reading a little low at 146 ohms all the others are 149, 150, 151, 150, 149. On the other bad side one resistor is at 150 and the rest read 78 ohms probably due to all the shorted mosfets. I will check the readings again after pulling the fets. All diodes check out. All the values seem to check out as per the schematic thus far. I need a little advise on identifying the driver transistors on my schematic. I am assuming they are the eight 2n5400's and the two 2n5550's. I am also not sure how to test them and could use some advice on that.

[greenmeanh1]

Also looking at the datasheets to compare the differences between the IRFP254 and the STW33N20. To my untrained analysis they do not appear to be close at all. Input capacitance numbers are quite different,, but what do i know. Thoughts??
The IRFP254 http://www.vishay.com/docs/91214/91214.pdf
The STW33N20 HTTP 301 This page has been moved

[J M Fahey]

They are reasonably close, so don't worry, unless you try to mix them in the same amp.

Both are overspec'd for the actual amp needs.
Amp has around +/-70V rails, or 140V end to end, those MosFets stand 200 and 250V respectively so you are well covered with any of them.

As of current, they will pass around 9A peak each, with 4 ohms load.
One is specified "33A" , the other "25A" but that nice big bold number is at 25C case temperature, ideal conditions.

At realistic 100C case temperature, one is rated 15A , the other is not shown but should be some 20-22A , again both suitable.

The only main parameter which makes them different, still by not that much, is that one has minimum 0.085 ohms chip resistance, while the other has 0.14 ohms, but tht woulkd be very important if used as switches at the maximum current possible, not the case here and no that big deal.

Of course, it means different types can not be paralleled but you already know that.

Input capacitance is way too large in all switching power MosFets, but if the drive stage is well dimensioned, it is accounted for.

My MosFet of choice is IRFP250, but once I only had a few IRFP450 left, much higher voltage than needed *but* also doubled chip resistance , I used them anyway and asked my Customer (who HAD to play that night) to come back next Monday, when I would have replenished my IRFP250 stash .

He never did because "the amp works fine" ... and keeps working fine 3 years later, a practical confirmation

FWIW here in Argentina we have a new President and he devaluated the Argentine Peso 30% in a week, so Importers are holding to their stock until Economy stabilizes, so I only got some IRFP240 ... work like a charm.

[greenmeanh1]

Transistors came in today from Digikey. I only got 14 pieces so hopefully i can get 6 to match and get one amp done. Still waiting on the caps. Should be here this week. I was a little alarmed to find the new transistors weigh a gram less than the original 254's which were 6 grams while these are 5. At this point i'm not going to worry about it. I ordered some cheap Chinese for a buck each so i can compare when they get here.

[greenmeanh1]

Ten Rubicon 4700uf 80v caps showed up in the mail today so i'm ready to start surgery.