Thinking about a transimpedance MM phono amplifier
The first stage uses a 6k2 series resistor to give a 75us pole, to be followed by the 3180/318 pole/zero
Adding capacitance does not give the usual cartridge LC resonance, so what sort of frequency response error is this going to cause?
The first stage uses a 6k2 series resistor to give a 75us pole, to be followed by the 3180/318 pole/zero
Adding capacitance does not give the usual cartridge LC resonance, so what sort of frequency response error is this going to cause?
Theres a whole long thread on this. Search 'aurak'
edit thread is here mechanical resonance in MMs don't let the title confuse you, but all the good stuff on transimpedance loading is together there so makes sense to add your questions there.
Some discussion broke out here as well MM Input Stage working in inverted mode - not to find; why ?? but has gone quiet whilst LD mulls negative impedance.
edit thread is here mechanical resonance in MMs don't let the title confuse you, but all the good stuff on transimpedance loading is together there so makes sense to add your questions there.
Some discussion broke out here as well MM Input Stage working in inverted mode - not to find; why ?? but has gone quiet whilst LD mulls negative impedance.
Last edited:
There's some good history for sure. IMO, the way to fly is to present as close to true transimpedance, ie s/c, to the cartridge as possible then deal with correction. Aurak does this with some tidy tricks convenient in the first stage.
I've used Aurak as my main listening preamp for years now.
Just recently, an idea cropped up on another thread for a negative R input impedance preamp, so obtaining true 1/f amplitude sensitive response or getting the lf RIAA roll off from it. Then dealing with correction of course. IMHO that is well worth exploring...………..
LD
I've used Aurak as my main listening preamp for years now.
Just recently, an idea cropped up on another thread for a negative R input impedance preamp, so obtaining true 1/f amplitude sensitive response or getting the lf RIAA roll off from it. Then dealing with correction of course. IMHO that is well worth exploring...………..
LD
Last edited:
Please don't make the mistake that synthesized negative resistance has anything to do with cancelling noise (as some do) there is no free lunch. The "zero" input resistance transimpedance amplifier still has the current noise of the transresistance at the input. My conventional pre-amp with my Grado cart has the input referred noise of the 400 Ohm series resistance over the entire 20 to 20k range. You can't improve on this.
Last edited:
I think that a few doubts have been answered since a few of us started looking at transimpedance and lower input resistance amplifiers several years ago.
That the loading will not cause measurable mechanical damping
and
That the inductance does not seem to be level dependent and is sensibly constant over the band.
That the loading will not cause measurable mechanical damping
and
That the inductance does not seem to be level dependent and is sensibly constant over the band.
Not much you can do about level dependent inductance anyway except operating completely unloaded (maybe, just thinking here).
For a sample size of one (G1042) I have found the inductance to be constant from -80dBU to -20dBU. As the biggest signal I have ever measured from that cartridge was -20dBU from a scratch, this is more than enough
Bob used a synthesized "cold" resistor to optimize SNR. The transimpedance connection optimizes the frequency response at the expense of SNR (in the case of MM with 400+ mH L and transresistances significantly less than 47K).
Also note the "ultimate" SNR measured without the LP surface noise is probably irrelevant.
well the cooled load wasn't in the LA article, but is here. http://www.cordellaudio.com/preamplifiers/vinyltrak.shtml In case anyone wasn't already totally confused 🙂
Bob created two articles called vinyltrak, I forgot. As it was pointed out you can't fix the frequency response and optimize the ultimate SNR this way. Funny too Marcel thinks the cantilever wiggles when you drive the pins.
No, noise performance wasn't the motivation. Yer cannot defy the 2nd law. More the elegance and simplification of obtaining the 3180uS pole and a 1/f response directly from the cartridge. Then simplification of remaining correction, it's all very tidy. And different...….!
LD
Having had a fair look, there's no evidence quite the opposite, so it probably isn't there. But if it did exist, it would also show up unloaded, because only the inductive part of the coil can generate.
LD
I just asked because I was playing with a simulation of all three loading options, TI, 47k||100pF, and open circuit, and plotting the coil current (I used the model with the 100kHz self resonance we were playing with). I used a weakly coupled flux corrected for constant velocity response as a stimulus set so the output was 5mV for the 47K load. The current into a short was ~1.8uA and open ~7nA, the scale below is in dB re: Amps. I was just wondering if this was something that might cause a measurable artifact, if not that's OK.
BTW any ideas on a "cold" negative resistance that does not make an oscillator out of the TIA?
Attachments
Last edited:
- Status
- Not open for further replies.