JFET input phono preamp for MM

Jump to Latest
#81
Bonsai said:
Yo will need a lot of JFET’s to beat a Zetex 851/951 MC amp for noise. See the MC Head Amp thread on this forum for example.

😉
Click to expand...

True. But then, just to spoil the fun (as per Marcel), you don't need many to be already way lower than the surface noise of the record (as per PMA).

There is a difference between designing to the absolute limit vs. to the necessary level.

However, those Tin Can (? forgot the name) amps are that simple that they are per se super interesting.
 
#84
Hans Polak said:
I just would like to have some recipe to tell me when to use a BJT or a FET.
It can't simply be that a Fet is always better in all situations apart from the fact how much better.
But it's an interesting discussion so far.

Hans
Click to expand...

Regarding noise at audio frequencies, if you are prepared to use huge JFETs or huge numbers of paralleled JFETs when the source impedance is low, a JFET/group of JFETs can always outperform a bipolar transistor/group of bipolar transistors. That's simply because the product of voltage and current noise is lower for a JFET than for a bipolar transistor, because a JFET has far less gate shot noise than a bipolar transistor base shot noise. You can reduce the voltage noise at the expense of more current noise or vice versa by paralleling more or less devices (assuming the drain current per unit device is kept constant), so in the end the product of voltage and current noise is what sets the limit.

Whether the difference is worth the trouble is another matter entirely.
 
Last edited:
#86
Hans Polak said:
I just would like to have some recipe to tell me when to use a BJT or a FET.
It can't simply be that a Fet is always better in all situations apart from the fact how much better.
But it's an interesting discussion so far.

Hans
Click to expand...

Agree 100%. For ultra low noise, Lo Z inputs bip is my preferred choice. For Hi Z then JFET.

RFI is an issue but then take precautions as applicable. I never bought the ferrite bs (imagine telling that to Bruno!).

An AD797 is hard to beat up until about 2.5 k Zsource.
 
Last edited:
#88
dreamth, that Marantz preamp in post 66 has a .01 cap to ground and apparently nothing else. there must be a shema mistake there especially for MC operation.

what is the FET vs BJT distortion of a 5mv @1khz signal gain of 15?
then again at 50mv @ 20khz.
I am using 1ma gain of 15 2SC1815 single transistor.
I am guessing .01% at 20khz however this transistor is really quiet.
 
Last edited:
#89
PMA said:
That's fine, however this thread is about MM preamps, and MM cartridges are high impedance sources. Plus there is an indisputable benefit of higher JFETs immunity to RFI over BJTs. So, what's wrong, what is addressed and what is the contradiction here??
Click to expand...

Nothing - I also think JFET's will offer the best performance but I do not discount bipolar opamps for MM - the old warhorse 5534A is still one of the best wrt noise and I have never had an RFI problem with one (or the 4562 in line level apps where I have used them I can confirm)

Further, RFI problems will arise if you don't take precautions. I see far too many small signal amplifiers with no input RF filtering and quite a few power amps as well.


Limiting an amplifier's bandwidth so that it reproduces signals faithfully over the audio bandwidth is just good engineering practice - but you already know this.

🙂
 
#91
@PMA
There must have slipped a typo in your specs, regarding the 87dB unweighted S/N ref 5mV @1Khz, see image below.
The S/N that I calculated based on the Orto 2M Blue and on your several graphs, give a S/N of 75.8 unweighted and 78.4dB(A) weighted ref 5mV@1Khz.
Maximum S/N with a completely utopian noiseless amp with the Orto 2M Blue Cart and ref 5mV/1Khz is 76.8dB unweighted and 79dB(A) weighted, so you are only short of 1dB from the theoretical maximum which is excellent.

Hans

P.S. What was your purpose of starting this thread ?
You told not selling since 2012, but you did not specify the components in your circuit diagram either.
 

Attachments

  • Specs.jpg
    Specs.jpg
    85.5 KB · Views: 435
#93
Only with the intention to learn and avoiding to polarize, I read part of the Blowtorch link that Sottomano gave.
Apart from (some?) micropower BJT amps with super beta transistors where fets were far superior as Scott also reported, results where inconclusive for more usual audio designs as far as RFI susceptibility concerns.
Especially Syn08’s opinion that I regard very high but also JPV made convincing contributions based on measurements.

Everybody has the option to choose whatever he prefers, but especially in these times where Fet’s are getting more and more scarce, we don’t have to live with inferior BJT’s as it seems.

Hans
 
#95
It now seems that the only purpose of starting this thread was to promote the product for genereating revenue.
Nothing wrong with that, but it should have been placed in the Commercial Sector.

But even then, when questioning on multiple apparent conflicting pieces of information, you expect some answers.

Hans
 
#96
I’ve put this up before but here it is again.

I am getting in exactly the same figures as you Hans and interestingly, the noise floor measurements using the QA401 are in line with predictions as well on MC inputs.

So Just How Quiet is Your Phono Stage?

For MM, using 2 LSK389B devices in parallel and a 110 Ohm source resistor for the gain setting resistor in an all active EQ arrangement, the SNR is 75.2 dB with the cart connected (unweighted). So theory and measurements seem to tie up very nicely.

I cannot give more details at this stage because the design will be published in a US Audio magazine later this year, but here is a taster:-

X-Altra RIAA Phono EQ Preamp

Yes, I am promoting it 😉

(disclaimer: I make no financial gain from this other than a small payment for article pages as published. This purely a DIY project for tho# that want to explore low noise MM/MC solutions)
 
Last edited:
You must log in or register to reply here.
Top Bottom