Old Soul

I'm tired, long day

So, not too much babbling , for change :clown:

Objective: make it before Pa, and (normally) make it more complicated than Pa

What: amp using Germanium Biguns


edit: while I was dreaming about and working on this amp, I was often thinking about Pa's Joe and Jack

friendship, love and appreciation are very important parts of life

edit on 24. April 2021: THD measurements with 1T813 ( USSR!) instead of 2N1099 , post #68 Old Soul

edit on 15.09.2021 - few general setting procedure tips and tricks; be sure to read to the very bottom of post:

-will not teach ya how to solder everything properly - as always -Assembly of Japanese Bicycle Require Great Peace of Mind

-prior to powering up, be sure that you have trimpots set properly:

P102 (P202) - irrelevant
P101 (P201) - set to max; confirm with ohmmeter across R104 - in vicinity of 28R
P103 (P203) -irrelevant
P104 (P204) - set to min ; confirm with ohmmeter - improvise where to put probes

- jumpers JP101/201 - open/not mounted

-as always - best to work with one channel connected to PSU at time; in fact - first step just one channel , in second step - you can freely leave powered/connected already set first channel

- one VMeter across output ( DC offset) , second VMeter across any of current sense 0R22 resistors; use mini-clips/probes

- no load on output, input shorted to GND

- power up thingie; observe DC offset - it will rise slowly and everything up to several volts is normal in this moment

observe VMeter across 0R22 resistor - it will be Dodo reading for some long time; after 60-90-120sec - if you don't see any substantial reading , start fiddling with P104 (P204), one turn at time; of course - you did remember where you rotate itr to set to min value, now rotate in counter direction; - don't ask me which - I don't care, I can't remember, simply because I'm always trusting in DMM, not in my memory how to orient damn trimpot to have screw in proper place ....... and my logic what needs to be decreased and what needs to be increased is always ooked, vs. overall consensus :rofl:

allow 30sec or so between 1-turn steps, to observe when and how much mV you're getting across 0R22

goal is , , say for 1A of Iq - 1A * 0R11 = 110mV

keep fiddling to get mV creeping up, and once when you see steady rise for each turn, set da thingy to - say 80mV

(if you're using Si Bigun in place of lower output, feel free to go to 1A5 as target, pre-set to 1A3)

once when you're there - fiddle with P103 to get as close to 0mV of output DC offset

that would be preliminary setting of OS, now to input buffer setting :

- power off , put mVmeter across R106(206) test pins ; put second mV meter on jumper pin marked "B" and GND

power on, fiddle with P101 to set 20mV across R106(206) ; done - fiddle with P102(202) to set as close to 0mV for buffer output offset ( second
mVmeter) ; when done - power off, close jumper

*************** connect second channel to PSU, repeat entire procedure******************

when done with both channels , power on , re-set Iq of both channels to have 1A ( thinking mostly of life of venerable Ge - but don't fret - Square Law OS is crazy thing) somewhere in temp equilibrium, check output DC offset

slide of both Iq and DC offset - cold vs. hot - is really minimal, but I'm mentioning thermal equilibrium more thinking of importance of final temperature of heatsinks - observe, observe also temperature of outputs , that will give you info/confirmation how good and uniform is thermal interface between parts and heatsink


when done, connect signal, speakers, flip da switch, and enjoy

don't forget Da Porn - no Porn, no Glory



pinout of 2N1099 is in post #299



M101/201 not having type on schm - it is IRF9510



  • Old Soul sch.png
    Old Soul sch.png
    49.7 KB · Views: 2,714
Last edited:
LuDEF out of T-Bed, Old Soul on T-Bed


  • IMG_20210410_191029.jpg
    88.1 KB · Views: 516
  • IMG_20210410_191012.jpg
    87.5 KB · Views: 408
  • IMG_20210410_190215.jpg
    82.6 KB · Views: 267
  • IMG_20210410_190205.jpg
    72 KB · Views: 258
  • IMG_20210410_190158.jpg
    97.3 KB · Views: 314
  • IMG_20210410_190151.jpg
    73.6 KB · Views: 297
  • IMG_20210410_185321.jpg
    145 KB · Views: 281
  • IMG_20210410_185314.jpg
    221.1 KB · Views: 313
  • IMG_20210410_185307.jpg
    113.7 KB · Views: 414
  • IMG_20210410_185302.jpg
    154.8 KB · Views: 500
hard data, 8R load

rails 24Vdc, Iq 1A (yes, just 1A; ZM Chicken, want to preserve that funny Ge, to have long and prosperous life)

a seen on pictures, Ge used 2N1099


  • Old Soul THD vs. Freq. 8R.png
    Old Soul THD vs. Freq. 8R.png
    48.5 KB · Views: 177
  • Old Soul Freq. 8R.png
    Old Soul Freq. 8R.png
    48.1 KB · Views: 215
  • Old Soul 25W 8R.png
    Old Soul 25W 8R.png
    36.9 KB · Views: 207
  • Old Soul 20W 8R.png
    Old Soul 20W 8R.png
    35.4 KB · Views: 181
  • Old Soul 15W 8R.png
    Old Soul 15W 8R.png
    34.4 KB · Views: 177
  • Old Soul 10W 8R.png
    Old Soul 10W 8R.png
    34 KB · Views: 208
  • Old Soul 5W 8R.png
    Old Soul 5W 8R.png
    33.5 KB · Views: 316
  • Old Soul 1W 8R.png
    Old Soul 1W 8R.png
    32.3 KB · Views: 559
Last edited:
hard data, 4R load, same conditions as previous


  • Old Soul THD vs. Freq. 4R.png
    Old Soul THD vs. Freq. 4R.png
    49.1 KB · Views: 115
  • Old Soul Freq. 4R.png
    Old Soul Freq. 4R.png
    48.7 KB · Views: 126
  • Old Soul 30W 4R.png
    Old Soul 30W 4R.png
    37.7 KB · Views: 112
  • Old Soul 25W 4R.png
    Old Soul 25W 4R.png
    37.7 KB · Views: 98
  • Old Soul 20W 4R.png
    Old Soul 20W 4R.png
    37.5 KB · Views: 87
  • Old Soul 15W 4R.png
    Old Soul 15W 4R.png
    37.2 KB · Views: 91
  • Old Soul 10W 4R.png
    Old Soul 10W 4R.png
    36.7 KB · Views: 115
  • Old Soul 5W 4R.png
    Old Soul 5W 4R.png
    36.5 KB · Views: 157
  • Old Soul 1W 4R.png
    Old Soul 1W 4R.png
    32.4 KB · Views: 204
before anyone asks, to note:

biasing and drive - everything explained in LuDEF thread , posts #32-37, LuDEF , applied here too

only that level shifting source follower put in front of Ge in bottom half; due to fact that Ge needs 180-250mV EB voltage, looking at schm B lower than E, there is no need for separate CCS load for source follower - it's simply tied to output/Ge emiter node

(R122) 22R resistor value (in bracket) is sole change needed for 2SA1943

in case of Ge SF current is 180mV/10R

in case of bjt SF current is 650mV/22R