About possible Babelfish J interest

They should be both N Channel Mosfets. Just in case you made a mistake.

Distortion seems higher than it should be, as well as harmonic profile looks wrong.

Dropping to 0.22 Ohms Source, increasing bias to between 1.8A to 2A and changing to IRFP250 or 150 will cerainly bring significant improvements.

The results seem off though. Check all components around Aleph current source.
You might need to adjust the current gain on the aleph current source.
Did you do that?
 
Founder of XSA-Labs
Joined 2012
Paid Member
I changed the source resistors to 0.33R 3W carbon composition (vs Panasonic metal thin film), and the bias is now about 1.63 to 1.68amps (the bias pot seems to have little to no effect (7 to 8 turns no effect) on actually changing the bias, only changing Source resistors did anything, is this normal?) The distortion went dow to 0.14% to 0.15% THD and the profile is predominant H2 and H3 about -15dB lower and nothing else - so this is good. Should sound very lush and vivid. Still well above what I think it should be at (circa 0.04% or less?). Should I just leave it alone now or pursue trying to see if it can go lower in THD?

attachment.php


Zen Mod: what did the Spice simulations predict for the THD and the spectrum?
I am taking the data in a different location so the mains peaks and harmonics a little different.

Left Channel:
attachment.php


Right Channel:
attachment.php
 

Attachments

  • babbelfish-j-amp-0.33R-testing.jpg
    babbelfish-j-amp-0.33R-testing.jpg
    415.4 KB · Views: 439
  • Babelfish-J-FFT-1.68amp-2.83vrms-8ohms-FFT-Left-Ch.jpg
    Babelfish-J-FFT-1.68amp-2.83vrms-8ohms-FFT-Left-Ch.jpg
    142.2 KB · Views: 441
  • Babelfish-J-FFT-1.63amp-2.83vrms-8ohms-FFT-Right-Ch.jpg
    Babelfish-J-FFT-1.63amp-2.83vrms-8ohms-FFT-Right-Ch.jpg
    143.1 KB · Views: 1,335
Official Court Jester
Joined 2003
Paid Member
TP3+R24 combined value is what's influencing Iq , besides voltage across R29

so - lesser ohms of TP3+R24 ,Iq lower ..... more ohms there , R29 is dominating

regarding THD ........ I would say that's good enough , not taking in account methodology of measurements ........ what's most important is that you confirmed K3 15db down vs. K2

Spice is ( as we know always enthusiastic) predicting something as 0.015% , 8Vpp@8R , with 0R33 biggies and app. 1A6 Iq

more oomph and THD goes even lower , but who cares
 
Founder of XSA-Labs
Joined 2012
Paid Member
I got the smaller ones mostly as an experiment to see if the ground loop noise could be fixed, Did not want to plunk too much down before I knew that was the fix. I can use these for other projects now and will probably order a pair of 200VA's.

I have tried this with a 96dB/1w full range dual PA130-8 with an AMT super-tweeter and it does indeed come alive with more efficient speakers. It's really not sufficient for the 10F/RS225 FAST which really needs a 100w/ch amp to shine.