Measuring noise and SINAD of MM phono preamps properly

[Heinrich]

If the source was only 5mV at a few hundred ohms, this wouldn't be a problem, but with 100mV or more at 31k it's an entirely different story.

From all that I see in measurements, FET is a viable option. I don't agree with your analysis, not considering the limited capabilities of analog in general. But it slips slightly off topic, and I would better leave it at that.

 

[undeakyetonde]

Yes, nonlinear junction capacitance is proportional to the square of voltage. But there are two other mechanisms at play here:

1. The nominal output increases rapidly with frequency, the cartridge being a velocity-to-voltage converter, so we can expect at least 10 times more voltage than 0.5mV for a given displacement at 10kHz vs the quoted 1kHz. Surface clicks and pops also produce high transients. If we want 20dB of overload margin, our preamp must be capable of handling at least 250mV at 10kHz.
2. The output impedance of an MM system rises rapidly as cartridge inductance dominates, contributing up to 31 kilo-ohm at 10kHz for 500mH, further exacerbating the non-linearity. Well over 100 times greater than a typical generator output.

If the source was only 5mV at a few hundred ohms, this wouldn't be a problem, but with 100mV or more at 31k it's an entirely different story.

So, we pay for lower noise by distortion in high frequencies? Assuming careful circuit design, etc.

Thank you for your answer. As somebody with much less experience, I was curious why you chose NE5534 for your MM Pro, which seems to have worse noise characteristics than JFE2140. Seemed like a way of pushing the noise floor even lower, along with 'electrical cooling' or 'active loading'

 

[undeakyetonde]

To prevent nonlinearities and Miller capacitance, you can of course greatly complicate the circuit with cascode bipolars, but it is much simpler and more effective to cover the input stage with a general feedback, as I have done [ https://imrad.com.ua/ua/fonokorekto...vok-zvukoznimacha-z-ruhomim-magnitom-mm-50452 ] [ https://www.patreon.com/posts/otsenite-tekst-90783025 ]

"Графік залежності коефіцієнта гармонік від вихідної напруги"

I see this is THD @ 1 kHz at different voltages. Any chance you could provide similar measurement @ 5 kHz, 10 kHz, 15 kHz, 20 kHz?

 

[EJ3]

Yes, I'm aware of that. But the V15 (type V), which pma (presumably) referred to as the MM with the lowest inductance, that he was aware of, is anything but a new cartridge either.

And of course my examples weren't supposed to mean, that I'd vote for a lower inductance to be applied for MM input SNR measurement purposes. In fact I'd rather vote for using just the same input termination dummy values as already suggested in the good, old IHF-A-202 standard.

Greetings from Munich!

Manfred / lini

I have several & they are my favorite carts. (and in the T4P style [which is my primary use]) Personally, I also have a couple of Technics and a new GRADO P-MOUNT PRESTIGE GOLD3 still in the box.

 

[StevenSGT98]

Has the opa 1612 been tested with a cartridge load? Or calculated? I’m curious where it lands in the Fosi x5 for noise when loaded. Thank you.

 

[pma]

Has the opa 1612 been tested with a cartridge load? Or calculated? I’m curious where it lands in the Fosi x5 for noise when loaded. Thank you.

Yes. OPA1611 (single version of the dual 1612) has 2.27x higher noise (22Hz - 22kHz band) than the old good NE5534, with Shure M35X cartridge. With Ortofon 2m blue the difference is even bigger (higher inductance). OPA1611/12 choice is an incompetent design choice for MM, due its high current noise. Input current noise is 1.7pA/rt(Hz) compared to 0.6pA/rt(Hz) of NE5534 (both at 1kHz). A look into datasheets and understanding circuit design rules would tell the designer not to use 1611/12 for MM.

Noise with M35X is 72uV for NE5534, but 163.7uV for OPA1611. Almost same proportionality as is for input current noise. Also, please note that the frequency interval above 1kHz is much more important for integrated noise than the low frequency range below 1kHz (because of the 47k resistor shunted by rising cartridge inductive impedance). This is still overlooked by some persons.

Simulations and real life results fit perfectly. And listening as well. You hear sharp hiss with the opamps like AD797, LT1028, ADA4898 and OPA1611, when connected to the cartridge. And total silence when connected to 20 ohm or 50 ohm, but that is the irrelevant test method.

 

[morillon]

Yes. OPA1611 (single version of the dual 1612) has 2.27x higher noise (22Hz - 22kHz band) than the old good NE5534, with Shure M35X cartridge. With Ortofon 2m blue the difference is even bigger (higher inductance). OPA1611/12 choice is an incompetent design choice for MM, due its high current noise. Input current noise is 1.7pA/rt(Hz) compared to 0.6pA/rt(Hz) of NE5534 (both at 1kHz). A look into datasheets and understanding circuit design rules would tell the designer not to use 1611/12 for MM.

View attachment 433394

View attachment 433395

Noise with M35X is 72uV for NE5534, but 163.7uV for OPA1611. Almost same proportionality as is for input current noise. Also, please note that the frequency interval above 1kHz is much more important for integrated noise than the low frequency range below 1kHz (because of the 47k resistor shunted by rising cartridge inductive impedance). This is still overlooked by some persons.

Simulations and real life results fit perfectly. And listening as well. You hear sharp hiss with the opamps like AD797, LT1028, ADA4898 and OPA1611, when connected to the cartridge. And total silence when connected to 20 ohm or 50 ohm, but that is the irrelevant test method.

aie....

 

[pma]

Ici nous parlons anglais .

 

[morillon]

Ici nous parlons anglais .

;-)
ouch...

 

[StevenSGT98]

Yes. OPA1611 (single version of the dual 1612) has 2.27x higher noise (22Hz - 22kHz band) than the old good NE5534, with Shure M35X cartridge. With Ortofon 2m blue the difference is even bigger (higher inductance). OPA1611/12 choice is an incompetent design choice for MM, due its high current noise. Input current noise is 1.7pA/rt(Hz) compared to 0.6pA/rt(Hz) of NE5534 (both at 1kHz). A loo

Noise with M35X is 72uV for NE5534, but 163.7uV for OPA1611.

Simulations and real life results fit perfectly. And listening as well. You hear sharp hiss with the opamps like AD797, LT1028, ADA4898 and OPA1611, when connected to the cartridge. And total silence when connected to 20 ohm or 50 ohm, but that is the irrelevant test method.

Thank you very much for your explanation. All the information was extremely useful.

Just so I understand:
The 5534 SINAD is 75.3 db with the Shure M35x according to the chart on post 78
The ad797 which had measured at 2 pA for noise in was 71.3 db
So for the opa1611 it’s around 72 db??

I’m trying to understand the voltage comparisons and what exactly this means in real life. A record is capable of 70 db at best, correct? Most being way below this.

Does the inductance that is influencing the rise in the noise floor above 1k primarily also influence hf linearity? Meaning if the test scenario that doesn’t include the cartridge shows a linear response, once the cartridge is added will it have an affect on linearity?

Bottom line, is the noise hurting anything or is it outside of the vinyl experience other than people plastering their ears to a tweeter with the gain set to 0db?

 

[morillon]

the noise is added to the noise of the source! it is therefore all the more unfortunate....! same for thd etc
always astonishing to read these strange shortcuts with the "levels of noise""" ,"""distortion"", attached to the support...

 

[StevenSGT98]

the noise is added to the noise of the source! it is therefore all the more unfortunate....! same for thd etc
always astonishing to read these strange shortcuts with the "levels of noise""" ,"""distortion"", attached to the support...

I’m really trying to understand this. Signal has to sit above noise to be heard, correct. So, if one hears the groove, everything else is above it, correct?

 

[morillon]

I’m really trying to understand this. Signal has to sit above noise to be heard, correct. So, if one hears the groove, everything else is above it, correct?

the phono stage sees what is come to it as being just a ....signal... ;-)

even warned , a """1khz of music""" sounds a lot like a 1khz of "noise" ...
music is noise... ;-)

 

[StevenSGT98]

The 5532 is used in the phono stage in the Wiim Ultra. It sounds pretty decent.

 

[pma]

For those who might be interested, I have prepared a real world samples of the noise recorded from the phono preamplifier output when the MM cartridge is connected to its input.

A simple RIAA preamp with one opamp and 40dB/1kHz gain has Shure M35X cartridge at the input. The output noise is recorded with 2 different opamps, LT1028 (lowest voltage noise bipolar input) and OPA 627 (JFET input, acceptable voltage noise). Bellow in the zip file are 2 files with recorded noise. 48kHz sampling, 32-bit float wav mono files. The FS level corresponds to 500mVrms.

It is a demonstration of the audible influence of the input current noise of the bipolar opamp (LT1028 here), which is in no way revealed in a usual test with the low impedance generator. With the low impedance generator, the noise (and thus SINAD) is lower with the bipolar LT1028 opamp, just the opposite of the real world scenario with the MM cartridge. This is the reason why tests with low impedance generator (as in the reviews here) are pointless and do not reflect the real world operating conditions.

So, give it a chance, listen and you will hear and understand what is a "sharp" noise that is a result of phono preamp input current noise.

P.S.: some readers at diyaudio could not play the 32-bit files, so I am also posting the 24-bit versions below.

 

[narshorn]

Hello,
Sorry if a little off-topic.
I own (and used to use) an Ortofon MC3turbo coupled to a D. SELF NE5534 RIAA preamp.
Some folks told me that this was bad for noise (that a LT1028 would be far better), however I couldn't hear a hiss with that combo ?
So far, the 5534 worked fine for this cartridge, albeit being about 100 ohm output
(if my memories are right, but I don't remember the particular inductance value of the cart).

Best,

n.

 

[AnalogSteph]

The MC-3 Turbo is a high-output MC cartridge (3.3 mV, almost as much as Ortofon's MM carts at 4-5 mV - that's kind of the entire point of HOMCs). Given that MM phonopres with voltage noise levels approaching 3 times those of a 5534 still tend to be fine in practice, I'm not surprised that you didn't notice any issues.

 

[narshorn]

For those who might be interested, I have prepared a real world samples of the noise recorded from the phono preamplifier output when the MM cartridge is connected to its input.

A simple RIAA preamp with one opamp and 40dB/1kHz gain has Shure M35X cartridge at the input. The output noise is recorded with 2 different opamps, LT1028 (lowest voltage noise bipolar input) and OPA 627 (JFET input, acceptable voltage noise). Bellow in the zip file are 2 files with recorded noise. 48kHz sampling, 32-bit float wav mono files. The FS level corresponds to 500mVrms.

It is a demonstration of the audible influence of the input current noise of the bipolar opamp (LT1028 here), which is in no way revealed in a usual test with the low impedance generator. With the low impedance generator, the noise (and thus SINAD) is lower with the bipolar LT1028 opamp, just the opposite of the real world scenario with the MM cartridge. This is the reason why tests with low impedance generator (as in the reviews here) are pointless and do not reflect the real world operating conditions.

So, give it a chance, listen and you will hear and understand what is a "sharp" noise that is a result of phono preamp input current noise.

P.S.: some readers at diyaudio could not play the 32-bit files, so I am also posting the 24-bit versions below.

Thanks PMA.

The LT1028 version seems about 6.6 dB noisier on analysis.

With instant FFT comparison it becomes clear where noise is spectrally higher.

Regards,

n.

 

[morillon]

Thanks PMA.

The LT1028 version seems about 6.6 dB noisier on analysis.

With instant FFT comparison it becomes clear where noise is spectrally higher.

Regards,

n.


View attachment 436885

(excuse the off-topic..what analyzer fft are you using?) ;-) )

 

[StevenSGT98]

Thanks PMA.

The LT1028 version seems about 6.6 dB noisier on analysis.

With instant FFT comparison it becomes clear where noise is spectrally higher.

Regards,

n.


View attachment 436885

I ask a simple question I believe. What is the highest dynamic range vinyl carries? I’ve heard of a sample “somewhere” of about 70 db.

At what point does noise become a diminishing factor?