Ampzilla Outputs Using MJ15022

I've never tried matching the MJEs, really nothing to match. There are no same gender pairs working together and matching NPN to PNP will take MANY of each and you MIGHT get lucky. The drivers look to be working together but with each getting a different input signal does matching even matter.
 
I finally had time to test the finished products. I ran a bunch of tests and everything performs well. I'm not loading a dozen images, but everyone likes to look at FFTs. So here are two. An image at 75W (both channels) and then another image just before 1% distortion, which was 245W (both channels).

I used my HP 8903B as the signal source and not the software generator just because it was a bit easier to work an actual instrument vs software. My 8903B has a unique hump around the fundamental that low in the noise floor. Yes, the software sinewave is much cleaner. I'll probably use the software in the next rounds of testing. I also have GPIB control for my 8903 which allows me to run some automated tests.

Ampzilla 1 @75W.jpg


Ampzilla 1 @245W.jpg
 
To wrap this thread up, I rebuilt 3 Ampzilla amplifiers (original). Some were kits and one was a factory built amplifier. I used MJE15028 / 29, MJE15034 / 35 and MJ15003 / 4 for outputs. I was able to obtain very similar numbers on all of them. Similar noise floors, distortion, frequency response, max output, etc.

Both channels driven, 1% distortion was at 238W. Harmonics started to increase at 220W. Single channel driven, 270W at 1%.

I can do IMD, but I didn't set it up and measure.

All three passed my listening test. One still has its working jeweled LED. The one thing I could not find any longer.

Thanks to those that contributed. I struck up a nice conversation with user @llwhtt over the course of the rebuilds.