I missed the words "only film"....
Have you folks checked out THAT's 300 series of matched low noise transistors? I just noticed they have NPN's and PNP's together in the same package. They seem to sell direct in relatively small quantities. By my figuring the 2SA1084 is speced at .5nV @ 10mA Ic which is ~16 Ohms rbb. THAT's are 30 Ohms so at reasonable currents they're almost at parity.
@PigletsDad: what is the reason you don't use also +-24V for the frontend? Adjusting especially R8/R9 and R3/R7 should bring back the currents to their original values. But I guess I'm missing something?
Or do you just want to keep resistor noise low (+-24V are anyway not needed for additional headroom of the frontend)?
Cheers, Hannes
Or do you just want to keep resistor noise low (+-24V are anyway not needed for additional headroom of the frontend)?
Cheers, Hannes
Resistor noise is a factor but the big one is avalanche noise in the front end transistors - if you put more than about 10V of Vce on a bipolar, an additional noise process starts to intrude.
If you wanted to run the front end at +/- 24V, the front end transistors would have to be cascoded, which has some other benefits, but makes the circuit rather more complex.
As I said initially, there are lots of obvious things that can be done to improve performance in various ways, at the expense of complexity.
If you wanted to run the front end at +/- 24V, the front end transistors would have to be cascoded, which has some other benefits, but makes the circuit rather more complex.
As I said initially, there are lots of obvious things that can be done to improve performance in various ways, at the expense of complexity.
I understand!
Do I see correctly that it is in principle possible to use that phono pre also with moving magnet cartridges? One needs simply to adjust R5/R4 such that all the gain is wasted.
A brute force solution for sure, but I prefer having both possibilites if feasible at all.
Thank you for your replies! All the best, Hannes
Do I see correctly that it is in principle possible to use that phono pre also with moving magnet cartridges? One needs simply to adjust R5/R4 such that all the gain is wasted.
A brute force solution for sure, but I prefer having both possibilites if feasible at all.
Thank you for your replies! All the best, Hannes
For other cartridge types, you can use a similar circuit, with currents etc adjusted.
For medium impedance, so called high output MC (say 100-200 Ohm, 1 or 2mV), just raise R5 (and maybe R4, to give a bit less distortion). I would go for R4=47 and R5=1K. Gain is now 20 or so.
For MM, the impedance scale is much higher, and you really want to use much lower operating currents in the input stage to optimise the noise match. At these levels, the Renesas transistors have no great advantage as Johnson noise from the source is dominant anyhow. In fact, BC550/BC560 have rather lower interelectrode capacitance, so I would swap over to using those for MM for Q1 to Q4.
To optimise for MM, R8 and R9 should be 15K, R1 and R6 go to 2K7, R3 and R7 should be 100, and I would go for R4=220 and R5=2K2 - gain of 10.
For medium impedance, so called high output MC (say 100-200 Ohm, 1 or 2mV), just raise R5 (and maybe R4, to give a bit less distortion). I would go for R4=47 and R5=1K. Gain is now 20 or so.
For MM, the impedance scale is much higher, and you really want to use much lower operating currents in the input stage to optimise the noise match. At these levels, the Renesas transistors have no great advantage as Johnson noise from the source is dominant anyhow. In fact, BC550/BC560 have rather lower interelectrode capacitance, so I would swap over to using those for MM for Q1 to Q4.
To optimise for MM, R8 and R9 should be 15K, R1 and R6 go to 2K7, R3 and R7 should be 100, and I would go for R4=220 and R5=2K2 - gain of 10.
SY said:
in scales with the operating currents (until you get to very low currents, where thermal leakage dominates).
For MM operation, you want to match a range of impedances from about 400 Ohm (near DC) to 47KOhm (at 20kHz or so). The noise figure contours say that operation at about 200uA will do this very nicely, and most of the range lies inside the 0.5dB contour - the highest end maybe just gets close to the 1dB level, but that is much less audible, and rolled off by the RIAA curve anyhow.
In my previous post, I suggest values that will be good for MM. I also suggest swapping to BC550/560, as it has lower interelectrode capacitance. The non-linear Miller capacitance is irrelevant with MC, as the source impedance is so low, but matters much more with MM, as the source impedance rises across the audio band. The complementary design removes some of the nonlinear effects, but the capacitance of PNP and NPN devices typically differ by 50%, so it is not a big win at HF. A more complex, cascoded design does better here.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.