Nick Sukhov Su-XXI MM Phono stage -85 dBA SN ratio...

Damn...I wish I had access to these sources 20 years ago...
Marcel proposed an original circuit Fig5, but the non-differential signal pickup from the input stage adds noise of the current generator transistor BC550 and the 422 Ohm resistor in its emitter. It is too noisy and his fine idea on active inpur resistance is far from being fully implemented.
 
Account Closed
Joined 2010
Actually it's not that noisy...but you can use 2sc1845 instead...I think that a Wireless World article need to adress a wider audience and outline the principles of operation. That differential is actually a true differential as in the famous APT Holman just not symmetrical if judged by input impedance . As for the 422 ohms running 6.5 mA through it sounds better than 330 ohms run at 17mA...
 

Attachments

  • Screenshot_20220623-211017_Samsung Internet.jpg
    Screenshot_20220623-211017_Samsung Internet.jpg
    303.3 KB · Views: 232
  • Screenshot_20220623-211755_Word.jpg
    Screenshot_20220623-211755_Word.jpg
    297.7 KB · Views: 245
Marcel proposed an original circuit Fig5, but the non-differential signal pickup from the input stage adds noise of the current generator transistor BC550 and the 422 Ohm resistor in its emitter. It is too noisy and his fine idea on active inpur resistance is far from being fully implemented.

There is an error in figure 5, one section of the 30 dB/40 dB switch is drawn in the wrong state.

The noise of the tail current source is effectively in parallel with the channel noise of the JFET. I designed it in the mid 1990's and I don't remember all the details, but I'm sure I must have chosen the tail current degeneration such that the JFET dominates.

In any case, it's good enough for me, but if you or anyone else wants to improve it further, be my guest.

The circuit is just my implementation of the Hoeffelman and Meys configuration from around 1978, so it is not that original. Electronic cooling even dates back to 1939.
 
Last edited:
  • Like
Reactions: 1 user
The circuit is just my implementation of the Hoeffelman and Meys configuration from around 1978, so it is not that original. Electronic cooling even dates back to 1939.

Marcel, can you suggest links to these interesting works of 78 and 39?

действие охлажденки.jpg
As a radio circuit engineer by education, the principle of cooling has been obvious to me since the time of studying the Theory of Radio Circuits, but it was your work in 2003 that gave impetus to the activation of its use in sound engineering. I also studied the cooling circuitry in detail, and came to the conclusion that it is not very effective for phono preamps due to the fact that the RIAA frequency response has a 20 dB drop at high frequencies, i.e. where the cartridge inductance impedance is at its maximum, input resistor noise current causes the most trouble, and cooling should have the greatest effect. But in reality, this same frequency response drop of 20 dB eliminates the benefit of cooling - the real SNR IEC-Awtd advantage is only 1.5...2 dBA, according to my observations. But as for the use of cooling in the playback amplifier of tape recorders, here the frequency response does not have a decline at high frequencies, but even some rise (compensation for the loss of the playback head gap) after alignment with time constants 120-70-47 µs. Therefore, the advantage of cooling is more significant - see the spectral noise densities of my 1987 playback amplifier [ http://archive.radio.ru/web/1987/06/034/ ] later attached with cooling on the screenshot (with х100 cooling resistor of 5.1 Meg vs standard 51 k). It is output spectral noise densities IEC-A weighted, input with BC560C. PB head equivalent 120 mH + 240 Ohm (Nakamichi S4720).
 
W. S. Percival, "An electrically "cold" resistance", The wireless engineer, vol. 16, May 1939, pages 237...240

Jean M. Hoeffelman and René P. Meys, "Improvements of the noise characteristics of amplifiers for magnetic transducers", Journal of the Audio Engineering Society, vol. 26, no. 12, December 1978, pages 935...939, see also Ernst H. Nordholt, "Comments on "Improvement of the noise characteristics of amplifiers for magnetic transducers"", Journal of the Audio Engineering Society, vol. 27, no. 9, September 1979, pages 680...681
 
  • Like
Reactions: 1 user
Account Closed
Joined 2010
Here's where I found a hint to 1984 patent which proved to be Mohr's:
http://waynestegall.com/audio/riaa.htm#mozTocId495677
He's heads and shoulders above my mental gymnastics abilities, but I hope you will enjoy his work more than I can.His spice models reveal the circuits values.
What's even more strange is that Mohr's circuit resembles the AD844 internal guts...while we know now that Win Palmer designed the AD844.
 
  • Thank You
Reactions: 1 user
Member
Joined 2014
Paid Member
How about cartridges with an inductance of the order of 500 mH and a recommended load capacitance of the order of 400 pF?
I do have a measurement by Hans of an ortofon S-120. This is a DJ cart, but interested me as it has a fancy suspension to improve tracking thats in the 700mH range and really needs a load in the range of 100k and 100pF to not drop like a stone above 10kHz.
Decades later, Steven van Raalte wrote an article about an improved version.
Steven's analysis is one I do not agree with and hoping to show some evidence he was incorrect in his conclusions
@billshurv is trying to measure these things on cartridges. The first cartridge he tried has a nearly flat electrical response and must therefore have a well-damped mass-spring response, as the response with a test record is also good.
To be clear, due to time pressures of life I've not been able to measure much at all but am being helped by some great friends who do. All the results are being posted in the mechanical resonance thread as I don't want to cross post too much or derail this thread other than to point out cartridges are a far more complex equivalent circuit than the literature until now might have suggested.
 
As for the 422 ohms running 6.5 mA through it sounds better than 330 ohms run at 17mA...

I do not think so. Because in Kenwood L-02 the current generator noise of Q30 Q28 is distributed as equally correlated currents to the input differential stage Q18 Q20 Q22 and then it is perfectly compensated as a common mode signal by the second differential stage Q38 Q40. In Marcel's circuit, the noise current of the current generator is not compensated in this way and is added to the input transistors noise.
 
The source impedances to each transistor, idle currents and respective transconductances are the ones determining the overall noise, not the resistors or the balanced distribution of currents through the differential stage.
I agree with Nick, the tail current source noise almost completely cancels in his circuit. I don't think it is a big deal, as you can dimension the tail current source such that its noise doesn't dominate anyway.
In my book of 1985 [ freely here: https://www.patreon.com/posts/67628599 ] I have strictly mathematically shown on pages 72 and 73 that the input-referenced noise of the differential stage with an active uncompensated (via next differential stage) current generator is 1.8 dB higher than that with a resistive supply of the differential stage
 

Attachments

  • Н.Сухов, В.В.Колосов. Техника высококачественного звуковоспроизведения (1985, djvu) () (z-lib....png
    Н.Сухов, В.В.Колосов. Техника высококачественного звуковоспроизведения (1985, djvu) () (z-lib....png
    1.2 MB · Views: 227
  • Н.Сухов, В.В.Колосов. Техника высококачественного звуковоспроизведения (1985, djvu) () (z-lib....png
    Н.Сухов, В.В.Колосов. Техника высококачественного звуковоспроизведения (1985, djvu) () (z-lib....png
    605.9 KB · Views: 218
  • Like
Reactions: 1 user
In my book of 1985 [ freely here: https://www.patreon.com/posts/67628599 ] I have strictly mathematically shown on pages 72 and 73 that the input-referenced noise of the differential stage with an active uncompensated (via next differential stage) current generator is 1.8 dB higher than that with a resistive supply of the differential stage
And there is also on page 73: the input-referenced noise of the differential stage with an active uncompensated current source is 4.8 dB higher than that with a compensated by the next differential stage.