Another JFET phono preamplifier

Hello guys! I remembered recently about a Romanian magazine since 1994 where was presented an interesting phono preamplifier based on FET transistors. As previously most listeners preffered to buy or construct either "Le Pacific" or "Boozhound" preamplifier, I decided to come in help for them with this diagram.

Short description:
"The preamplifier for the magnetic cartridge presented can satisfy a series of lately wishes such as:
-very good signal/noise raport thanks to using the FETs;
-very low THD for a high gain and a high output voltage;
-high capacitance for preloading the signal peaks.

Technical characteristics:
-Gain at 1KHz = 40dB;
-V out = 250mV;
-S/N = 82dB;
-overload for the normal output voltage (250mV) = 40dB;
-THD = 0,01%;
-Z in = 220K;
-minimum Z out = 1K.

The preamplifier is built after the pattern of an OA equipped with FETs at the input in a differential mode. The stage mixes or combines the advantages of the BJT (regard frequency and high amplification) with the low noise of the P-channel FETs.

Regarding to the type of cardridge, C1 can vary between 10-90pF.

Here are the diagram, the power supply, the part list and the PCB 45 x 55mm
 

Attachments

  • IMG_20160212_153844.jpg
    IMG_20160212_153844.jpg
    431.6 KB · Views: 985
  • IMG_20160212_153927.jpg
    IMG_20160212_153927.jpg
    438.3 KB · Views: 954
  • IMG_20160212_153955.jpg
    IMG_20160212_153955.jpg
    452.7 KB · Views: 893
  • IMG_20160212_154013.jpg
    IMG_20160212_154013.jpg
    424.2 KB · Views: 840
Last edited:
Looks like a std ckt used by a few japanese mfgs.
there are advantages of dynamic range if you use the +/-30V supplies. lowering the supplies to +/-15V, you may have to adjust some values in the pre-amp ckt.
I think you can change the supply, change the zener diodes to be 15V instead of 33V. may have to lower the zener resistors tro maintain the same current, determined where the zener knee is for the devices you spec. you do not need +/-48V so that can be lowered to save on power dissipation of the pass bjts. Something like no less than +/- 20V should do.
You can use almost any power transistors, usually one uses a to-220 packaged device. you have to determine the supply current, know the un-regulated supply voltage, determine pass bjt Pd and then figure on a heatsink.
 
The supply regulator looks interesting (the preamp not so very) There may be a resistor value typo. I'm thinking of running some sims and posting what may be an improved version when I can get to it. My main computer is down, so I can only play with PSpice when I'm at the job...

Posted with a puny netbook running Linux, with a big monitor hooked up so I can see what the #[email protected]@@&*! I'm doing.
 
Looking at the regulator schematic in post #1, the resistors ins series with the zener diodes should both be 3k3. This biases the zener diodes at ~10ma, an appropriate value for a 1W zener that gets the bias off the "knee" and into a region of optimal incremental impedance. A 33k resistor sets the bias for the zener at ~1mA - way too low. The 510R resistors at the collectors of the BC176 and BC256 set their bias currents at a little over 1 mA.
I will try running this regulator in PSpice with available components and see how it fares for transient response and output impedance vs. frequency. I propose replacing the BD237 and BD238 with mosfets, as well as a few changes to speed up the loop and reduce the noise contribution from the zener diodes. I'll post the modified circuit when I can get the simulations done.
 
Last edited:
It would be nice to see all of the regulator. It looks like the Russian schematic also addresses the startup issue, but I'd like to see the rest of the regulator to figure out what is going on.

BTW, I have a Nikko Beta preamp with a similar circuit in the phono stage. I'll have to dig it out and compare. The Nikko got bested by some simple single-ended solutions long, long ago. There were some nice circuits in some places but a fair number of corners cut elsewhere.
 
Last edited:
...............This biases the zener diodes at ~10ma, an appropriate value for a 1W zener that gets the bias off the "knee" and into a region of optimal incremental impedance. A 33k resistor sets the bias for the zener at ~1mA - way too low. ..................
I have recommened that Zeners should run at a minimum current that dissipates at least 5% of max dissipation to ensure the dynamic impedance slope is above the knee in the Zener characteristic.

But last year some knowledgable Member came back and said that is not the way Zeners work. They work down to very low currents and still stay above the knee.
I have not followed up with any more research to find out which is more true.
 
I have recommened that Zeners should run at a minimum current that dissipates at least 5% of max dissipation to ensure the dynamic impedance slope is above the knee in the Zener characteristic.

But last year some knowledgable Member came back and said that is not the way Zeners work. They work down to very low currents and still stay above the knee.
I have not followed up with any more research to find out which is more true.

It depends on the zener voltage. Low voltage zeners have a very sloppy knee, so it takes a lot of current to get past it into a useful/stable operating region, in the sense that the voltage drop does not vary dramatically as a function of zener current. Higher voltage zeners, as in this case, have a sharper knee. so it takes less current to get them to the region of relatively flat voltage vs. current. This is reflected in the data sheets, which show reduced test current for higher voltage zeners (though there are considerations also of power dissipation).

To be safe, I would run at least at the specified zener test current from the data sheet. This presumably will get you to the specified value of incremental impedance. Also, starved zeners can be noisy. I found out the hard way...
 

Salas

diyAudio Chief Moderator
Paid Member
2002-10-08 11:31 am
Athens-Greece
The supply regulator looks interesting (the preamp not so very) There may be a resistor value typo. I'm thinking of running some sims and posting what may be an improved version when I can get to it. My main computer is down, so I can only play with PSpice when I'm at the job...

Posted with a puny netbook running Linux, with a big monitor hooked up so I can see what the #[email protected]@@&*! I'm doing.

I had tested this DC stabilizer circuit I put together in early November. It measured 55dB PSRR (scope limit) and simulates at 27mOhm Zo wideband flat. It was stable on the breadboard with those transistors and 100 Ohm. I think its related simple topology, if any helpful to you.
 

Attachments

  • DCSTB_exp.png
    DCSTB_exp.png
    5.7 KB · Views: 195
  • IMG_20151104_115430.jpg
    IMG_20151104_115430.jpg
    280.7 KB · Views: 184