Measuring noise and SINAD of MM phono preamps properly

[Heinrich]

Yes I will. I found 2 pcs of NE5534 in my stock. They should be good because of their quite low current noise. I think that SNR with cartridge should be at least 70 dB not to affect the best vinyl presses. SNR calculated as 20log(500mV)/total noise). 500mV as 5mV x 100 (Openamp gain at 1kHz).

@pma, I agree in regard to some due distance between the noise originating in the pickup/amp aggregate in comparison to the surface noise of masterly pressed and new records, weighing in noise from microphones, tape (we're analog) and mixing consoles and tape overlay and cutting (even more analog) and ... . It could be expressed as a noise figure also ;-)
It's a pitty that 7kHz is a quite promiment frequency range when it comes to noise, see the D-weighting curve. (wiki: "The large peak in the D-weighting curve is not a feature of the equal-loudness contours, but reflects the fact that humans hear random noise differently from pure tones, an effect that is particularly pronounced around 6 kHz. This is because individual neurons from different regions of the cochlea in the inner ear respond to narrow bands of frequencies, but the higher frequency neurons integrate a wider band and hence ..."). We may want to adopt to it additionally.

But what do you think of switching to noise figures once the signal generator's contribution, its thermal noise is evaluated? Alas, I can't do that, because my very old LTSpice doesn't support the needed calculations.

 

[pma]

I have checked total noise at preamp output with input terminated by 50 Ohm and with the 50 Ohm generator. There is no difference ….
A kind of my know-how, but OK. As I need only 5mV/1kHz at the preamp input and Topping DAC has 2V output, I use shielded series resistor to send current to the preamp input that is terminated by splitter and 50 Ohm coaxial terminator. Thus, DAC output noise is about 50x reduced to negligible level and the only noise source then is a thermal noise of the 50 Ohm terminator, which is again negligible with respect to amplifier noise specs.

 

[DonH56]

Comparison of simulated and measured output noise density and total output noise for several op-amps in the Openamp phono pre. Shure M35X cartridge is at the input. Please note good correspondence of simulated and measured results. Please also note the dominant role of the input current noise density at higher frequencies.

View attachment 427462

Nice. For decades we (you, me, designers) have known bipolar input op-amps are poor choice for any high-impedance circuit due to bias current noise, and the cancellation (injection) circuits to reduce the input bias current increase the internal current noise (so make it worse). CMOS chopper circuits to reduce input offsets are also a problem, as if MOSFET 1/f and flicker noise is not already an issue at audio frequencies. JFETs have been the preferred choices for a long time and this shows one reason why.

I have related before the story of a friend who tried to use a mW GaAsFET as a phono preamp, back when GaAs was the Big New Thing, only to discover that a device with 1/f noise corner >100 MHz was a poor choice for audio. Yes, I did say "I told you so", since I had pointed that out before he tried it. It didn't even make a good line-level buffer.

 

[pma]

As you know very well, Don, experience is not transferable. And so is the knowledge. In case of knowledge, there is a chance that one would study, however, experience is still not transferable .

 

[Heinrich]

Seems interest is low. Anyway, my LTSpice generated some simulated data no RIAA applied. I cannot translate it for non-tech/ people to easily understandable graphs. Personally for me the situation is quite clear in many aspects. Dunno how the phono folks would deal with the issues. I display the stuff uncommented as a proof of work, and leave it at that without going into the details.

Current noise (dBv, referring to 1V!!):


Self noise:


From hand-made calculations I learned that self noise is easily the major contributor whn using contemporary JFET opamps.

Please (re)consider the old fashioned noise figure, and maybe you are willing to investigate on the D-weighting as cited from the wiki above.

Have fun!

 

[DonH56]

Using a 1 A source and converting to dB is not terribly useful to me as an audiophile or engineer when such high impedances are present and thus high noise levels are reported. Useful for a relative comparison, but unrealistic without considering the RIAA curve if you want to do that, and of course 1 A through a phono cartridge would probably destroy it (plus no cart would generate that much current). You can scale the source in SPICE if you want more realistic numbers.

D weighting has not been used for years (nor B); it was to measure high-level noise (from aircraft engines, after doing a search), and has a large peak that is really not proper for a phono preamp or normal audio measurement and would provide an inaccurate noise level compared to normal measurements (and simulations). Today A and C are common weights, but for tests like these I would just use wideband noise for the spectrum and integrate it to compare to the signal level from the cartridge rather than rolling off the noise. Actually, specifically for a phono preamplifiers, if I were to do anything it would be measure after RIAA equalization and generate the output or input-referred noise from that as @pma did with his preamp circuit. You can calculate the noise figure if desired but that metric is mainly an RF term not typically seen in audio. The folk who know it can undoubtedly calculate it themselves, and for audiophiles it is a meaningless metric.

 

[AnalogSteph]

I found 2 pcs of NE5534 in my stock. They should be good because of their quite low current noise.

Do note that supposedly there is a noteworthy difference between A and non-A parts, with the former basically having BC559 dies as input transistors. These days, heaven knows. Stating the exact manufacturer and type should keep away any possible confusion.

 

[Heinrich]

Using a 1 A source and converting to dB is not terribly useful ...

D weighting has not been used for years (nor B); it was to measure high-level noise (from aircraft engines, after doing a search), and has a large peak that is really not proper for a phono preamp ....

Considering the complaints people like to lend their ear TO the speakers in 30cm = 1 foot distance registering "hiss". The pre amp is crap, what to say? I found it interesting to learn about quite off-center expectations. That was it. I'm looking forward towards what you guys achieve.

My graphs were not for the first-timers, as I said. Just a "proof of work", as is seen rarely in audio discussions )

 

[pma]

I also do not consider noise figure and noise factor to be useful in case of evaluation of MM preamps.

 

[pma]

Seems interest is low. Anyway, my LTSpice generated some simulated data no RIAA applied.

Post #1 of this thread, simulated voltage noise density of Shure M35X, loaded by 47k//150pF. You can do it for any other cartridge as you wish. Please use V/rt(Hz) units instead of dBV or dBr when showing noise plots, to keep it FFT length independent.

 

[Heinrich]

I also do not consider noise figure and noise factor to be useful in case of evaluation of MM preamps.

A statement like "The pre amp / pickup combo is 3dB noisier than is theoretical possible." In my days way back in time it seemed to be useful. The engineer's 'good enough'.
The D-weigthing may address peoples' frequent complaints--also with active speakers, about 'hiss'. The engineer's 'good enough'?

This weekend I had an anecdotal conversation on the urge of today's analog / phono scene for total perfection. It was fun

 

[JSmith]

properly

Why use this word in the thread title?


JSmith

 

[pma]

If you read the post #1 of the thread you would know the answer. And I am not going to participate in off-topic discussions.

 

[narshorn]

@pma, I agree in regard to some due distance between the noise originating in the pickup/amp aggregate in comparison to the surface noise of masterly pressed and new records, weighing in noise from microphones, tape (we're analog) and mixing consoles and tape overlay and cutting (even more analog) and ... . It could be expressed as a noise figure also ;-)
It's a pitty that 7kHz is a quite promiment frequency range when it comes to noise, see the D-weighting curve. (wiki: "The large peak in the D-weighting curve is not a feature of the equal-loudness contours, but reflects the fact that humans hear random noise differently from pure tones, an effect that is particularly pronounced around 6 kHz. This is because individual neurons from different regions of the cochlea in the inner ear respond to narrow bands of frequencies, but the higher frequency neurons integrate a wider band and hence ..."). We may want to adopt to it additionally.

But what do you think of switching to noise figures once the signal generator's contribution, its thermal noise is evaluated? Alas, I can't do that, because my very old LTSpice doesn't support the needed calculations.

Hi Heinrich (and the others )

In the field of MM carts and MM RIAA, regarding weighting curves and humans random noise hearing, please find attached this article by Marcel Van de Gevel.

https://www.diyaudio.com/community/attachments/noise-and-moving-magnet-cartridges-pdf.1087276/

Marcel Van de Gevel, "Noise and Moving-Magnet Cartridges", 2003, Electronics World.

Best regards,
.

 

[MaxwellsEq]

Hi Heinrich (and the others )

In the field of MM carts and MM RIAA, regarding weighting curves and humans random noise hearing, please find attached this article by Marcel Van de Gevel.

https://www.diyaudio.com/community/attachments/noise-and-moving-magnet-cartridges-pdf.1087276/

Marcel Van de Gevel, "Noise and Moving-Magnet Cartridges", 2003, Electronics World.

Best regards,
.

An interesting article, thank you, and welcome to ASR.

 

[restorer-john]

The issue in 2025 is of course the CD player and its test signals are grossly inferior to what we can achieve with a high quality D/A converter, in terms of noise, THD and linearity.

That said, I used a test CD and a "blameless" CD player for signal generation for many years.

 

[Heinrich]

Hi Heinrich (and the others )

In the field of MM carts and MM RIAA, regarding weighting curves and humans random noise hearing, please find attached this article by Marcel Van de Gevel.

https://www.diyaudio.com/community/attachments/noise-and-moving-magnet-cartridges-pdf.1087276/

Marcel Van de Gevel, "Noise and Moving-Magnet Cartridges", 2003, Electronics World.

Best regards,
.

Thanks for the link. Alas, even with 'electronic cooling' the improvement remains at 3dB or so. My main caveat, a personal one, targets the notion to assign the 47kOhm standard input to the amp. It should be seen as part of the pickup, because it is positively prescribed by it ;-) That increases the cartridge's contribution to that of a 12kOhm resistor, as explained in the arcticle.

What I liked to see is an honest discussion of the worth of exceptional efforts. Not the least, the combination of JFET and bipolar in that single differential frontend, so nice ... didn't dare to think of it before!

 

[pma]

This is the updated chart with SINAD and noise results for the Openamp MM phono preamplifier with various op-amps. Noise was measured with M35X cartridge at the input. SINAD was measured from 50 ohm low noise generator and 5mV/1kHz signal. SINAD with the cartridge was calculated using the formula posted in the post #1 of this thread.
4 op-amps with bipolar input and 2 opamps with JFET input were tested. Measured results for LT1028 are not shown, because I was not able to make it stable in the phono circuit yet.

When measured with a 50 ohm generator (which is almost same method as used at ASR with 20 ohm AP generator), the best SINAD result was for ADA4898, 85dB. But, when the cartridge was connected to the input, this op-amp had the highest noise and thus the worst SINAD in real conditions. This is for the reason that the generator method does not take into account the input current noise of the preamp. For this reason, the method with low impedance generator is only valid for low impedance signal source, but not for the high impedance source as is the MM cartridge loaded by 47k standardized impedance. This is what I have been trying to explain, with higher or lesser success, last weeks. JFET op-amps with with reasonably low voltage noise (<8uV/rt(Hz)) are much better choice for MM preamps as they are free of input current noise issue. Kudos to the old NE5534! (bipolar). As Douglas Self states, this is the preferred one from bipolar opamps.

Please also note excellent conformance of simulation and measurements for JFET op-amps. Not so perfect for bipolar op-amps, due to a bi unpredictable value of the input current noise of individual parts.

--------------

Oscillations with LT1028:


This results in highly elevated noise floor and SINAD (with generator) falls down to 67 dB.

 

[DonH56]

@pma This is one of the things that makes ASR so valuable, thank you!

 

[pma]

Thank you @DonH56 , the reactions from knowledgeable people like you make me thing that probably the efforts might make sense.