Though this phono preamp is about 10 years old, I have decided to open this new thread, based on discussions on JFET phono preamps, noise, and MM cartridge impedance effect on the resulting noise. The preamp page is
Discrete preamplifier for MM cartridge
and there is not much seen of the circuit design, also the measurements are of the older data. So I am attaching a blind schematics (I still do not want to disclose everything) to have some topic of discussion. The preamp has complementary differential JFET input stage followed by MOSFET output stage. It is supplied from shunt regulators that are driven from constant current sources.
Measurements of output noise, calibrated in dBV, are attached. One measurement is with the preamp input shorted by 50 ohm, the other one with the input connected to Technics SL1200 with Ortofon 2M blue cartridge, so these are the real life conditions.
Simulations of noise were made and they almost perfectly correspond to the measured values.
P.S.: DC servo not shown in the schematics.
Discrete preamplifier for MM cartridge
and there is not much seen of the circuit design, also the measurements are of the older data. So I am attaching a blind schematics (I still do not want to disclose everything) to have some topic of discussion. The preamp has complementary differential JFET input stage followed by MOSFET output stage. It is supplied from shunt regulators that are driven from constant current sources.
Measurements of output noise, calibrated in dBV, are attached. One measurement is with the preamp input shorted by 50 ohm, the other one with the input connected to Technics SL1200 with Ortofon 2M blue cartridge, so these are the real life conditions.
Simulations of noise were made and they almost perfectly correspond to the measured values.
P.S.: DC servo not shown in the schematics.
Attachments
Last edited:
PMA what is the spot noise at 1kHz with a 50 Ohm resistor and then with the cartridge (in nV/rt Hz).
I assume all your plots are dBV, yes ?
I assume all your plots are dBV, yes ?
Yes the plots are in dBV, but not in dBV/sqrt(Hz), so it is not a power spectral density directly. I will re-measure it to see PSD directly.
In simulation I get this and I am pretty sure it will be the same by measurements, as total RMS of measured and simulated noise is the same. The elevated plots are for L = 700mH.
In simulation I get this and I am pretty sure it will be the same by measurements, as total RMS of measured and simulated noise is the same. The elevated plots are for L = 700mH.
Attachments
Having the lowest noise preamp in vinyl or tape playback is pointless due to the medium (vinyl-tape playedback at a certain speed) background noise.Anything higher than -75db S/N in vinyl at 33RPM is useless.You won't hear the difference .Max headroom and input-output saturation behavior is more important and then there's DUST who aim at ruining your day 🙂 !
I thought you'd be talking more on this aspect as you're using some Lender like VAS...which is good for theoretical headroom, but not great for ouput offset unless the input stage is perfectly matched...which frankly...is quite an expensive thing to do with 4 complementary j-fets...I thought that op-amps were invented to deal with this aspect a long time ago.
Are you actually selling some matched j-fet pairs that people should cherish in order to hear the lowest background noise in the Universe 🙂 ? I think you can buy an old radio and tune it between channels for that.
I thought you'd be talking more on this aspect as you're using some Lender like VAS...which is good for theoretical headroom, but not great for ouput offset unless the input stage is perfectly matched...which frankly...is quite an expensive thing to do with 4 complementary j-fets...I thought that op-amps were invented to deal with this aspect a long time ago.
Are you actually selling some matched j-fet pairs that people should cherish in order to hear the lowest background noise in the Universe 🙂 ? I think you can buy an old radio and tune it between channels for that.
Last edited:
Regarding offset ...
Remember that PMA did mention this device includes a DC servo?
You can also see it on the stuffed PCB.
mlloyd1
Remember that PMA did mention this device includes a DC servo?
You can also see it on the stuffed PCB.
mlloyd1
I don't recall any servo discussion anywhere in this topic...cand you point it to me? It may be subject to further undisclosed discussions as promissed though 🙂
The servo has been mentioned in the post #1. The output DC offset is unmeasurable with DMM at 000.0mV range.
I am selling nothing. I stopped selling audio kits and PCBs in 2012 except for several PCB units of the openamp after this date.
Did I .... I thought so, but frankly, I was not able to recall or search it. I do not know it had its own thread, but I admit I have mentioned it somewhere. After about 10 years of using it regularly, I find it proven and good enough 😉
What a coincidence to see this preamp having the same noise as the catridge Johnson noise itself discussed in another topic yesterday!
https://www.diyaudio.com/forums/ana...-vs-bjt-input-phono-preamp-4.html#post6288650
The largest noise source in a moving-magnet amplifier is usually the input termination resistor. If you want the lowest noise possible, you have to realize the 47 kohm damping resistance using combinations of series and shunt feedback (electronic cooling, synthesized loading, whatever you want to call it). If you notice any difference at all, it will be in between records, otherwise it gets masked by record surface noise.
I saw that approach in D Self Small signal book a long time ago, but then i found so many dirt cheap commercial designs whose noise i wasn't able to hear at all that i simply skipped that.
Re-measured for you in dBV/sqrt(Hz), PSD. To get input noise, please consider that 1kHz gain is 36dB. The soundcard with considerably lower noise was used for this measurements, so the RMS value for 50 ohm has improved a bit.
One can see that noise resulting from cartridge inductance is dominating at middle and high frequencies and can't be improved by lower noise design. Last but not least, real life noise is determined by empty groove noise and is at least 20dB higher than the preamp noise, so the fight for lowest noise design does not make much sense.
Attachments
Last edited:
You are absolutely correct, Bill. Real life LP may have S/N up to 60dB(A) and this is sometimes better than tape hiss (without Dolby A) and mike preamp noise. Depends.