I can implement this if folks are OK with it.
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Measuring noise and SINAD of MM phono preamps properly
- Thread starter pma
- Start date
lini
Member
Sorry, but no, it doesn't. See section 5.12 (weighted SNR), 4th paragraph - which suggests an additional 500 mH in series and 125 pF in parallel to the resistance-inductance-combo. So actually exactly the same as in IHF-A-202 section 3.12, 4th paragraph.
Greetings from Munich!
Manfred / lini
- Thread Starter
- #44
Updated version of the paper (measurements with better cartridge and board shielding to prevent mains interference peaks) just uploaded:
It is visually nicer, but there is no change in conclusions.
It is visually nicer, but there is no change in conclusions.
- Thread Starter
- #45
Thanks for your question. I made a similar set of measurements back in 2008, LT1028 was probably the best performing bipolar input opamp in MM phono pre those days. It has very low voltage noise and low current noise (for a bipolar opamp)
In = 1pA/rt(Hz), which is 2.4x better than the ADA4898 in my test and 1.7x better than OPA1612 in Fosi X5. You might expect similar result (above 2kHz) as with LM6171 here. This would be a considerable improvement from ADA4898 or OPA1612 and a good candidate even for the MM preamp. The part is very expensive and unfortunately I have no more of them in stock. In 2008, I used it for a commercial MC phono preamp design and also for a balanced input MC preamp design.
- Joined
- Nov 19, 2020
- Messages
- 736
- Likes
- 1,627
- Thread Starter
- #47
Thanks for the info. I may add it to my next purchase at Audiophonics. Back in 2008 I bought those parts at about $50 per piece.
Interestingly enough, the price is 19.73 EUR (w/o VAT) at Mouser:
So, still expensive. I prefer reliable suppliers to prevent fake opamps delivery.
- Thread Starter
- #49
I have just purchased 2 pieces at Mouser (DIP8 package). With FedEx shipment, duties and VAT, two pieces cost is 71.58 EUR. Very close to what I paid in 2008. So it goes.
Note: I cannot risk fakes, I have quite deep experience with Chinese and Taiwanese fakes of opamps and JFET transistors and Bipolar power transistors, namely 2SA1302/2SC3281. However, it never happened with official certified distributors like Mouser and RS components
Note: I cannot risk fakes, I have quite deep experience with Chinese and Taiwanese fakes of opamps and JFET transistors and Bipolar power transistors, namely 2SA1302/2SC3281. However, it never happened with official certified distributors like Mouser and RS components
Last edited:
- Joined
- Nov 19, 2020
- Messages
- 736
- Likes
- 1,627
@Audiophonics is a member here, so I hope they chime in to solve the riddle....
Ragnarsson
Member
- Joined
- Jun 26, 2021
- Messages
- 15
- Likes
- 20
For the source impedance model and related phono premap noise woulnd'nt be interessing coming back to existing old work, especially from Hallgren:
"On the Noise Performance of a Magnetic Phonograph Pickup
The thermal noise of a magnetic phonograph pickup has been studied in detail. The analysis has been made using computer-aided design techniques employing a new empirical model of the electrical impedance. Two types of pickups are considered. The frequency spectra and the total rms value of the thermal noise have been studied, and the factors which are taken into account are the value of the load capacitance, the load resistance, the RIAA equalization, and subjective weighting according to the IEC 268-1 standard A curve. The noise performance of the pickup when used in a phonograph amplifier system has also been studied. It was found that with a preamplifier using modern semiconductor devices it is possible that the pickup can contribute as much as twice the noise voltage compared with the semiconductor devices alone. Attention should thus be placed on the noise performance of the pickup when considering the overall noise performance of a phonograph system."
Electrical model proposed by Hallgren is more complete than just a resistor and inductance in serie. My experience in measuring impedance of cartridges, using my DATS, shows that impedance above the usual 1kHz differs from the expected impedance due to resistor and inductance.
For the final noise performance, cartridge and phono preamp cannot be dissociated, but for the evaluation of the performance of phono preamp alone, in order to be able to get comparison between them, how to proceed? To use always the same source impedance? And in this case what shall it be?
A basic example, consider your cartridge is a classic Shure M75 (630 ohm, 720mH as per https://pubs.shure.com/view/guide/M75E/en-US.pdf), you are interested in the result having this as a source impedance. This would help making the best choice for your phono preamp.
My cartridge has impedance that differs from M75, this is a Rega Exact with 200 ohm and 90mH (I measured this as there is no manufacturer info on it). And then I might be interested by another phono preamp model.
Phono preamp cannot be measured with all of the possible source impedances, the only thing than can be done is to define two or three extreme cases, and measure noise for them.
Or having measurements that can be compared with identical conditions, independant from the cartridge choice.
"On the Noise Performance of a Magnetic Phonograph Pickup
The thermal noise of a magnetic phonograph pickup has been studied in detail. The analysis has been made using computer-aided design techniques employing a new empirical model of the electrical impedance. Two types of pickups are considered. The frequency spectra and the total rms value of the thermal noise have been studied, and the factors which are taken into account are the value of the load capacitance, the load resistance, the RIAA equalization, and subjective weighting according to the IEC 268-1 standard A curve. The noise performance of the pickup when used in a phonograph amplifier system has also been studied. It was found that with a preamplifier using modern semiconductor devices it is possible that the pickup can contribute as much as twice the noise voltage compared with the semiconductor devices alone. Attention should thus be placed on the noise performance of the pickup when considering the overall noise performance of a phonograph system."
Electrical model proposed by Hallgren is more complete than just a resistor and inductance in serie. My experience in measuring impedance of cartridges, using my DATS, shows that impedance above the usual 1kHz differs from the expected impedance due to resistor and inductance.
For the final noise performance, cartridge and phono preamp cannot be dissociated, but for the evaluation of the performance of phono preamp alone, in order to be able to get comparison between them, how to proceed? To use always the same source impedance? And in this case what shall it be?
A basic example, consider your cartridge is a classic Shure M75 (630 ohm, 720mH as per https://pubs.shure.com/view/guide/M75E/en-US.pdf), you are interested in the result having this as a source impedance. This would help making the best choice for your phono preamp.
My cartridge has impedance that differs from M75, this is a Rega Exact with 200 ohm and 90mH (I measured this as there is no manufacturer info on it). And then I might be interested by another phono preamp model.
Phono preamp cannot be measured with all of the possible source impedances, the only thing than can be done is to define two or three extreme cases, and measure noise for them.
Or having measurements that can be compared with identical conditions, independant from the cartridge choice.
- Thread Starter
- #52
That is fine, but I have to repeat that it is not only about the cartridge and 47k thermal noise, but more importantly about frequency dependent voltage drop across cartridge impedance (that is in parallel to 47k loading resistor) due to amplifier input current noise In. Again, please read post #1 of this thread. As per Ohm law, the noise voltage drop component due to amplifier input current noise is Vn(load) = In * Z(load), Where Z(load) is cartridge // 47kohm. This is extremely important noise source for bipolar input opamps with input noise current of pA/rt(Hz) order. And not important for JFET input opamps. I will repeat this over and over, as I can see that this basic information is not generally well understood. However, we must follow the electrical circuit rules, as this is supposed to be a "scientific" oriented forum. Input noise voltage has 3 components: thermal noise of the input impedance, amplifier input voltage noise and voltage drop across input impedance due to amplifier input noise current.
Last edited:
- Thread Starter
- #53
There is a quite nice and clean result with LM6171 at the position of the input opamp, and I am happily surprised, as it is a bipolar input opamp. Noise degradation with the M35X cartridge is quite low in comparison with 50 ohm terminated generator SINAD. Though the SINAD number with 50 ohm generator is not stellar, the real world result with the cartridge is very good. Looking forward the Mouser delivery of LT1028.
Setup
Setup
Last edited:
- Thread Starter
- #54
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.
- Thread Starter
- #56
I will maybe find 5534 in my stock. It will be quite good. I need to simulate loop gain stability with 5534 first. I had to add FB compensation for AD797 and LT1028.
OPA604 will be a bit worse than OPA134, no need to test, and I do not have it.
The list will be updated. And I can only test single opamps in DIP8 packages, or with a SMD/DIP adapter.
BTW, this:
is extremely suspicious. They sell 2 x LT1028 in SMD/DIP adapter for half price of the single piece. I do not believe they are from genuine source.
Thanks a lot @pma!
In regard to the panel I would suggest, again, to add at least the general purpose NE5534, or NE5532 respectively. And reiterated, the self-noise of the pickup assembly may be worth a consideration:
N = sqrt( Vn^2 + (In * Z)^2 + self-noise^2 )
Not so much for a more 'fair' comparison, but in order to get a reasonable feel for 'diminishing returns'. As far as my recollection holds, the famous Douglas Self stated, that the thermal self-noise (no pun intnded) of the pickup, no Vn or In of the amp included, would limit the S/N ratio to something about 83dB anyway. Considering the self-noise would allow for stating a "noise figure". By how much the amplification adds onto the self-noise of the signal generator. If the self-noise would be at -80dB and after due amplification the noise is at -78dB, the noise figure would be 2dB (alas, it is always positive).
So once a noise figure of 3dB is achieved, and the lower noise is more a convenience feature (as with phono)* only, people could decide if the extra effort for 2dB is worth it. An NF of 2dB is clearly better than 3dB, but for 10 times the cost?
In defense of SINAD, it makes so much sense in the digital era, especially with lossy codecs. Think of spurious glitches from incomplete reconstruction analog - digital - lossy codec - digital - analog. They won't be easily described as harmonic distortion, but could be everything. So better say just "noise" and ask for SINAD to be lower than x, as low as is possible within reason.
Funny: are the losses of mpeg-audio ... noise?
*
- in the past turntables connected the pickup's signal to ground as long as the needle didn't ride the groove, resulting in at least neither generator nor current noise
- as the vinyl record cannot be quieter than -60..70dB anyway, a phono pre could use a common 'noise gate' to shut down the output as long as the needle doesn't ride the groove, no noise at all
- once the needle hits the groove, all noise from the pre is drown in the surface noise of the record anyway
Last edited:
morillon
Major Contributor
- Joined
- Apr 19, 2022
- Messages
- 1,902
- Likes
- 443
- Thread Starter
- #60
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).
Similar threads
