How critical is Mosfet matching in power amp output stages?

Matching tends to count most when you edge toward high levels of output and the stressing through massive transients in a mixed harmonic, well, the micro distortions become macro and then you pop an output transistor or two.

In the rest of the given listening times, the created distortions from minor mismatching are generally masked to anything but electronic distortion analysis.

To check on that, one can look at amplifier and device reports from reviewers at stereophile and then note when John Atkinson finds anomalies or high distortion in one channel or mono block, vs it’s given compliment.

The distortions from sets of matched transistors in given amplifiers that you may encounter, in the manner that you seem to be dealing with...will probably be less than what the reviewers and John Atkinson have run into, those given odd times.

Matching at the levels you are looking at are more than likely will be within or below your given hearing threshold, regarding hearing discrepancies.

Again, it mostly comes out as a problem only under extreme stressing. Ideally perfectly matched transistors will all blow simultaneously(never gonna happen) but in the real world, one pops - the one odd man out, even if just out by a hair... and then takes the amp down with it. (these days we can check that under extreme loading and using a $300 thermal camera)

It sounds like you might be able to pull your transistors and check on them and match them better among themselves, if you find some minor discrepancies. The way you speak is that soldering is not a stranger to you.. and you can use a basic multi-meter, and that includes some basic transistor matching via said meter.

I’ve done the same in repairing audio amplifiers, ie, need 12 devices, buy 16-20, and use the best matched 12 out of the group (all same batch)....and leave it at that.

The best you can do is primitive matching (which is usually adequate, except for our audiophile anal fretting about ’best’ and whatnot), unless you’ve got a curve tracer.

Check for other alterations in the circuitry, base resistors and voltage/current drive in the predriver in the monoblock may be different, and so on. Look very very closely.

Thanks Teo, I appreciate your advice! I did a detailed comparison of the remaining amp board in the monoblock with the two stereo versions of the amp that I have on hand, checking resistor color codes, etc. while looking at the schematic. I'm going to do it again before I fire up the soldering iron, though, just to be sure. Unless I made a mistake it does seem like all B&K did to mono the amp out was to remove the right board, leaving the mosfets on the mounts as they were, and jumper some twisted pairs over from the remaining board. I wonder if these long and probably relatively low quality wires are negatively impacting the sound--just a side issue. They also changed the power supply to put all the power into that side, of course. Anyway, I read a thread on Audiokarma about inexpensive component testers and ended up buying one for about $31 that will allow me to test gate to source threshold voltage, which according to an online article from Pass is the essential test for matching and to ensure the mosfets are all working properly. I'd also be able to test gate capacitance, drain to source on resistance, and forward voltage of protecting diode if any of those turned out to be worthwhile tests for this application. In terms of matching I'll probably call it good after that and rely on A/B comparisons with my other amps, including the stock, factory reconditioned ST-140 to detect any problems. Hopefully it will be a success story, we'll see. Again, thanks for your help, Teo.