volume rise with frequency problem

[nightanole]

I just built one of these:
Hammond AO-44 Reverb Amp to Hi-Fi Amp Conversion


However with speakers with little to no crossover, the volume gets higher as the frequency rises. It starts around 2-3k and then is about 6db hot by 18khz. I can more or less correct this with my software eq, but id like to have a hardware fix. I was under the impression that 10k/1nf on the primaries was a zobel network. If its not would a standard 2.7ohm/.1uf zobel on the secondaries fix it?

 

[Miles Prower]

However with speakers with little to no crossover, the volume gets higher as the frequency rises. It starts around 2-3k and then is about 6db hot by 18khz. I can more or less correct this with my software eq, but id like to have a hardware fix. I was under the impression that 10k/1nf on the primaries was a zobel network. If its not would a standard 2.7ohm/.1uf zobel on the secondaries fix it?

It looks like something funky is happening in your gNFB loop. This thingy probably doesn't have enough phase margin, and you're getting a rising characteristic with frequency. You'll need to do some o'scoping while running the input with square waves to see where the ringing is worst. Since this is a reverb amp, you can bet that its OPT ain't exactly quality iron, and its leakage inductance is probably higher than what you'd expect from a true Hi-Fi OPT.

Then you'll need to parallel the 82K NFB resistor with small capacitors to see what will stop that, and remove that high frequency roll-on.

Also, something doesn't look right about that Zobel. 10R / 0.1uF across the primary doesn't look right. That probably should be connected in parallel with the secondary, not the primary. You might want to try that before doing anything else.