50w Single-Ended BAF2015 Schade Enabled

Triode connected mosfet with input transformer Papa short but with lot's descriptions about Schade 50W
in BAF2015 video made by Linear Audio from 25 minutes to 25 mn 40 seconds.

Oh yeah 2 harmonic character with little third in the mixture
I feel input transformer choice are fun i keep mystery on my candidates.
I like compare them to Jensen in near future.

ps. curious 2SK77B hockey puck photo ( prototype ? ) package from Maxonic catalogue.
 

Attachments

  • 01FD4848-05E4-4D20-BEE1-DF13A899EA4B.jpg
    01FD4848-05E4-4D20-BEE1-DF13A899EA4B.jpg
    260.4 KB · Views: 847
  • 06F583D2-094B-4473-BE49-B9379D930462.JPG
    06F583D2-094B-4473-BE49-B9379D930462.JPG
    132.8 KB · Views: 841
I simulated the Baf2015 in ltspice but without the input transformer and I used an IXFN120N20 model from Ixys. The amp has a closed loop gain of 20db and the distortion at 1 watt shows about 0.001% and at 10 watts about 0.005%. I simulated the F6 and the ba-3 and their distortion was close to what was measured. So I don't know whats going on here. It sure would be great if it was accurate.
 

Attachments

  • Baf20151.png
    Baf20151.png
    10.2 KB · Views: 712
  • Baf2015.png
    Baf2015.png
    11 KB · Views: 691
Well I can't find anyone selling it.I see that Ixys does now show the 50n120sic as an active part. And, the datasheet is incomplete. I just found another sic part "APT70SM70J" from Microsemi for $27.30
Picodumbs, what is its transconductance of it? I noticed you had a picture of it from a link in one of your previous posts.
 
Last edited:
The 140N20P datasheet shows the transconductance for a given Id but for the small current being used for the amplifier it's to hard to see what it is. If you use ltspice to graph out vgs to Id for the current range used and calculate the transconductance from the graph how close would it be to reality?
 
The 140N20P datasheet shows the transconductance for a given Id but for the small current being used for the amplifier it's to hard to see what it is. If you use ltspice to graph out vgs to Id for the current range used and calculate the transconductance from the graph how close would it be to reality?

It should be reasonably close eg within 20%, if not you can tweak the spice model.