Compound (Sziklai) pre

Some progress. It works from the start. Powered by ~59VDC, with sine output ~960mVpkpk, 327mVrms, it has distortion 0.063%, likely better since my supply is not the best at the moment.
Very interesting.
Personaly I detect 2nd harmonic at 0,63% and 3rd at 0,22% during an extensive test of sensitivity of different distortion components.
You are very close to that relationship but a magnitude lower so I suspect that it can handle some complex music too before being noticed as colouration.
Hope for a very neutral, musical and transparent sound character with that harmonic distortion pattern.
Will follow your progress...
 
Last edited:
  • Like
Reactions: Vunce
I do not plan other experiments, its installed in the box playing beautifuly. Its clean and musical.
0.6% distortion may be confusing on complex music. That reminds me when i started my experiments with hamtone pre. Its ideal as mic preamp for one voice. But as for complex music its too much distortion. Some hamptones depending on the jfet i used got very clean and great sounding down to 0.02%. Worth exploring.
 
  • Like
Reactions: Mooly
A simulation of a variation on the same theme...
Similar distortion character. Very high Zin.
Zout 600 ohm.
Gain 4,5
I don't favour global negative feedback, but in this case it seems to work well. Very low distortion in simulation.
Not really global though.

Skärmbild 2025-06-09 233400.jpg


Skärmbild 2025-06-09 233640.jpg


Skärmbild 2025-06-09 234042.jpg
 
Last edited:
Similar thread...
 
  • Like
Reactions: flex2
I built this exactly. Except i used bipolar capacitors on the input and output (two double the value back to back). And a1145/c2705 complementary pair.
Yes, simple unregulated. But i used lots of Cs.
 

Attachments

  • zd22 output stage (1) (1).JPG
    zd22 output stage (1) (1).JPG
    46.5 KB · Views: 61
Last edited:
The simplicity of this circuit was enough to motivate me to build a one and check it out.
I followed the attached circuit, minus the 1K output resistor and added sockets for easy gain changes. (10uF caps, Q2-BC550, Q1-BC557, 48V supply)

Adason,
The DC offset takes about 30secs to get below 50mV before I turn the power amplifier on.
Measured current draw for the circuit is 2.5mA.
Do you get similar results?
 

Attachments

  • Like
Reactions: adason
Yup, offset settling time is not out of the ordinary. Power amp last on/first off is a sure way to avoid any 'nasties".

No setup for distortion measurments.

The current draw is so low I wasn't sure if I messed something up.

Is the 1K output resistor needed?

Actually, I am quite happy with what I'm hearing with this mono channel using a bench DC supply, Class D amp and small speaker. Super quiet, I can't hear anything with my ear up to the speaker.
I will proceed building the second channel.

I'd like to pair this preamp up with a 10K ohm relay/resistor attenuator in my stash.
 
Last edited:
@Vunce
Seems like you have mixed resistor values between several of the circuits Adason presented.
Circuit with 3,3k + 1,5k in the feedback line don't have those values of resistors for bias.
2,5 mA is too little so something is not correct as far as I can tell.
 
I have typically ~5mV DC on the output. Settling time depends on what is connected to it, but its typically 2-3 seconds.
DC on output climbs to few volts only briefly, as its being discharged by 100k resistor. Or volume pot on next stage.
Yes, power the pre first, then amp. Or leave pre on all the time.
Listening to it now, sounds fantastic. Even on test speakers.
 
  • Like
Reactions: Vunce and flex2