New circuit based in NP CSX

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Zen Mod recommended:" dady - start new thread for your amp

this way we are just going to get lost in these 1300 posts

then I'll be able both to follow and comment

After this wisdom counsels of Aleksandar and because I am still curious, I started this thread that Mr Nelson Pass just answer my question in the previous #1290 of the thread Article Sony vfets part 1.

Dear partners of this Pass Labs threads.
Few months ago I started with the CSX1 project. Then, I got from Acroman several pieces of the pairs 82/28 and with another pieces of another friend and member, jama, we found carefully their curves envelope and matching to obtain four identical pairs.
Then looking for more damping I planned in parallel them.
I tried they and they are matched fine, with a DC of 0.01volt and a difference in the polarization of 0.2volt between the negative and positive rail.
Yesterday I start to think in the project again and trial several drawings of symmetrical connection between the devices, finally I found a schema that I will show with the object of be judged for more experienced members. Please I need your counsel if is possible to make it without troubles.
Best Regards in advance and thank for your opinion.


  • Balanced Pushpull SIT.jpg
    Balanced Pushpull SIT.jpg
    762.1 KB · Views: 332
  • csx simetrico.jpg
    csx simetrico.jpg
    909.6 KB · Views: 330
Yesterday, I tried in the workshop at the basement the voltages and the currents, the DC in the out is more difficult in stabilize, 0,2volt.
This afternoon I put in my living room, I was intrigued and I tried with one channel. The sound, very good. The loudspeakers are full range 15ohm EX4Lowther in a DIY back loaded horn (Mauhorn). I tried with bach solo cello sonata, something with King Crimson, and a little of Mahler with bestial symphonic passage. Suddenly the SMPS died,
But the sound was amazing, coherent and dynamic.
I guess I have to stabilize the DC, put a fan in the SMPS. Anybody of the people here could give an idea about the improvement?

Today I rise the current until 1.5A with 23,5 Volt.
45ºC, I trying in calculate the input capacitance...
Before I tried with parallel windings on the primary of the transformer with louder sound.
Today I attached and preamp but the windings in series gave better sound.
Exploring one channel with different music today, after 3 hours the DC rise a little, the SMPS with a 12 volt fan working with 4,5 gave enough air for maintain under good temperature.
The clue? somebody can counsel to me. Mr Aleksandar ZEN?
Mr Mode?
Best Regards Gentlemen
Official Court Jester
Joined 2003
Paid Member
for parallel connection of primary windings , you need properly low driving impedance - so either beastie preamp/buffer or input buffer implemented in amp itself
that's explanation for better sound when using primaries in series

find enclosed schematic upright - it's always easier to think with head upright :rofl:

what you're having - balanced or unbalanced preamp (stage preceding amp) ?


can you clarify - what you exactly mean with parallel and what with series connection of primary windings ?


  • Balanced-Pushpull-SIT.jpg
    91.5 KB · Views: 295
Thanks Aleksandar, in the schematic the primaries are in series.
Respect to my preamp, I have none Jeje, but I trying a unbalanced BTW, however I diagram it like balanced connection because the circuit is all balanced.
At the first time I think in a double CSX with four transformers for individual polarization, further I made more simple but I have some DC on the output.
Do you know how much is the maximal DC that I could use with my fullrange 15ohms?
Today I listen my Aleph X again, and I remembered the sound of the simetrical with more bass.
At the moment I not measure the amp.
Thanks for your comments my friend.
Best Regards
Official Court Jester
Joined 2003
Paid Member
sorry , I simply can't see how primaries are in series , if connected as in picture

you can either feed them balanced , each winding fed from one phase of source , or unbalanced , where ideally non-modulated winding is closed to ground with impedance identical to output impedance of source feeding other winding

huh - complicated to write ..... but say it in other way - feed one winding with 300R source , close second winding to ground via 300R resistor ....... for SE connection

so , better to sketch what exactly you are thinking with series vs. parallel

regarding DC at speaker ...... consensus is that 100mV is OK for PA amps , while for domestic ones ( feeding gentle fullrangers,for instance) 50mV is sane limit
Last edited:
Hi amigo Dady

So you say me it's noise cancelation if well matched vfet's in your balanced topology amplifier.
They deliver 2 times RMS power like square laws class A with 4 ohm drivers ?

Any future psu build project with 400 or 500 watts r-core transformers or smps is ok and not worse the efforts ?

Greetings :)
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.