morillon
Major Contributor
- Joined
- Apr 19, 2022
- Messages
- 2,124
- Likes
- 530
This noise is added to the noise of the LP... it's all the more unfortunate...
-
Welcome to ASR. 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
StevenSGT98
Active Member
- Joined
- Jan 19, 2025
- Messages
- 102
- Likes
- 103
StevenSGT98
Active Member
- Joined
- Jan 19, 2025
- Messages
- 102
- Likes
- 103
Please explain. I would assume the surface noise would be above the thermal noise. If the maximum dynamic range is - say 70 db - won’t that sit on top of the thermal noise?
MaxwellsEq
Major Contributor
- Joined
- Aug 18, 2020
- Messages
- 3,420
- Likes
- 5,966
Noise in this case will be incoherent, so one doesn't act like a concrete floor for the other. In other words, we can detect things below a noise floor.
narshorn
Member
- Joined
- Feb 8, 2025
- Messages
- 11
- Likes
- 10
I was myself a little off-topic ... ; ) It is the built-in Wavelab FFT function.
Regards,
n.
StevenSGT98
Active Member
- Joined
- Jan 19, 2025
- Messages
- 102
- Likes
- 103
Perhaps this has been answered somewhere else, but I haven’t run across it. How far below a noise floor can a person hear?
It begs another question.
Is what is beneath a noise floor worth hearing? The maximum dynamic range I’ve personally experienced from vinyl is 53 db on a Mofi pressing of the Cincinnati Orchestra. I’ve heard rumors of 70 db, though have never heard if that was a commercial pressing. If someone knows what recording this is, I would be interested to know.
My turntable has a published spec of 78db snr. So, let’s say if the noise on a preamp matches this, then theoretically rumble from the table and noise from the preamp will match. Ok. Again this is a published spec.
(Denon Dp-45f) Denons best tables spec’d out at 82.According to Recording Technology Inc.
On SVS Forum.
“Although the blog gives vinyl dynamic range as 70 dB, that's optimistic. A more realistic value for a very good pressing would probably be more like 60 dB, and then only at mid-band. High-frequency, and especially low-frequency, dynamic range will be substantially less. …. And, indeed, most recorded music has a dynamic range of 40 dB or less--often much less in pop music.”
I tend to agree.
So, a valid question is, if a preamp does generate slightly more noise, but the design leads to other desirable qualities, such as higher overhead, is the tradeoff ok?
Skeeter
Member
No. You can have low noise AND high overload. They are not mutually exclusive and if one is traded off for another, it’s a flawed design period.Perhaps this has been answered somewhere else, but I haven’t run across it. How far below a noise floor can a person hear?
It begs another question.
Is what is beneath a noise floor worth hearing? The maximum dynamic range I’ve personally experienced from vinyl is 53 db on a Mofi pressing of the Cincinnati Orchestra. I’ve heard rumors of 70 db, though have never heard if that was a commercial pressing. If someone knows what recording this is, I would be interested to know.
My turntable has a published spec of 78db snr. So, let’s say if the noise on a preamp matches this, then theoretically rumble from the table and noise from the preamp will match. Ok. Again this is a published spec.
According to Recording Technology Inc.
View attachment 436969
On SVS Forum.
“Although the blog gives vinyl dynamic range as 70 dB, that's optimistic. A more realistic value for a very good pressing would probably be more like 60 dB, and then only at mid-band. High-frequency, and especially low-frequency, dynamic range will be substantially less. …. And, indeed, most recorded music has a dynamic range of 40 dB or less--often much less in pop music.”
I tend to agree.
So, a valid question is, if a preamp does generate slightly more noise, but the design leads to other desirable qualities, such as higher overhead, is the tradeoff ok?
On a separate note, any phono preamplifier that has hum components above the thermal noise floor with a real-world cartridge attached would be an instant fail in my book. On measurement setups, you may get a 50/60 Hz spike from external mag fields coupling into the input wiring but nothing more than that.
Skeeter
Member
We can ask exactly the same question about distortion. Below what point is absolute distortion level of no consequence? Research says a highly trained ear may be able to detect 0.05% if the harmonic content is high order odds (ie > 3rd). So if your amp produces no more than 0.01% at all power levels and frequencies across the audio band and it’s low order harmonic distortion, you’re good to go.
For a phono preamp, you want at least 6 dB better than the surface noise of a record during play. But, such a preamp will have clearly audible hiss at nominal listening levels in the normal seating position when the needle is up. For this reason, you really want to be above about 72dB ref 5mV. Anything above this is a bonus and reflects good engineering.
Nick_Sukhov
Member
- Joined
- Nov 27, 2023
- Messages
- 21
- Likes
- 10
Nick_Sukhov
Member
- Joined
- Nov 27, 2023
- Messages
- 21
- Likes
- 10
Wrong. Mandated is >= 47kOhm.
Right in case if your special equalizer is used.
You're in a very luxurious position if you can afford to go all the way up to 150k, probably with a preamp (or at least a buffer) inside the turntable. Higher-end AT cartridges used to be notorious for their treble peaking with typical cable lengths and input capacitances, and people would go down to 36-39k to combat this on a fairly regular basis. But yes, from a noise and bandwidth POV a higher input resistance would absolutely be preferred.
I have thought about what it would take to integrate buffer circuitry pretty much directly at the cartridge or headshell interface (today's small-fry SMD electronics weigh next to nothing after all), but it's not entirely trivial, not to mention that you'd have to choose between MM or MC.
I have thought about what it would take to integrate buffer circuitry pretty much directly at the cartridge or headshell interface (today's small-fry SMD electronics weigh next to nothing after all), but it's not entirely trivial, not to mention that you'd have to choose between MM or MC.
Last edited:
Nick_Sukhov
Member
- Joined
- Nov 27, 2023
- Messages
- 21
- Likes
- 10
see my version attached or more detailed here https://www.patreon.com/posts/otsenite-tekst-90783025
Attachments
Nick_Sukhov
Member
- Joined
- Nov 27, 2023
- Messages
- 21
- Likes
- 10
Shorting input @ SNR measurements leads to a measurement error of 9 to 14 dBA, and failure to take into account the mandatory time constant of 7950 μs leads to an incorrect measurement of the frequency response with an error of +3 dB at a frequency of 20 Hz. ASR must bring their measurement methods into compliance with the requirements of the current IEC 60098:2020 standard [ https://www.diyaudio.com/community/...hono-preamp-review.422514/page-2#post-8092101 ]
Yep, I posted some rubbish before. Based on this:
a spead sheed was published (not by me):
SYclotron Audio | RIAA Noise Calculator
www.syclotron.com
But it has an error in the (hidden) line 27, that becomes relevant if the |Z| value for the cartridge is actually needed. It is if the contribution of current noise is to be evaluated. I corrected it.
With standard cartridges like 1kOhm/750mH at 300pF, and using the OPA1612 (Fosi X5) the noise figure would be 8,8dB, the feedback path not considered. The pickup's self noise--it is a real thing, would limit the SINAD to -76,4dB.
Under the same conditions the NE5532 would show a noise figure of just 2,5dB, while naturally the the limit in SINAD remains the same.
That's quite a penalty for using an opamp, that would suit moving coil duties also. The NE5532 is close to optimum for MM use, a FET would need to be quieter than 8nV/sqrtHz to beat it. The OPA1612 is simply 6dB worse.
In contrast a pickup of moving iron type aka MI, like Grado Green would be happy to collaborate with the OPA1612, leaving the NE5532 behind, but not by much. According to my measurements it is as good as it gets.
Last edited:
That, maybe, answers the question raised in the headline of this thread. How to reasonably measure the sensitivity of a phono pre for MMs in regard the input impedance? In case of an FET based design, a (virtually) short on the input is appropriate. If bipolar, it's complicated. I personally would stick with the simulation based on a known noise current. A dummy is too bothersome in many aspects. How to measure noise current? It's a thing of the past anyway. Nobody expects the former 'perfect' from vinyl anymore, me thinks.
Which would be? Something has to be screwy given that |Z| dips below Re(Z) in the treble, which is obviously an impossibility (since |Z|² = Re(Z)² + Im(Z)² >= Re(Z)² for any complex number Z), but Excel formulas rarely tend to be particularly pretty or transparent. Side note, the Xc line ought to be labeled -Xc. I don't really get what they're doing, I should probably have a look at the '70s article describing the approach again.
Side note, I remember the trusty Tietze-Schenk mentioning a favorable interaction between inductive sources and JFET input capacitance in noise terms but would have to look up the specifics. I don't know how relevant this would be if C is dominated by a capacitor and cable, mind you.
Last edited:
In line 27 for |Z| the error is, that line 21, Rp namely, is referenced, but it should be line 23, Rp//Rin. At least according to the article. And once changed, the spread sheet will reproduce the article's example calculation for |Z|.
That |Z| is not used for the self-noise in focus, though. Hence it doesn't affect the "max S/N" numbers. Only when trying to calc/ the current noise's contribution as an add-on it came to me, that there's something fishy.
Crosspost (partially) since this is the best place for this data:
---
Out of curiosity, I decided to do a little Google-fu to see just how many other cheapo "good measuring" phono preamps were "guilty" of using OPA1612 or some similar device with a needlessly large amount of current noise in order to make bogus "low noise" claims. And it's .... a whole lot of op-amp based devices on the top of the SINAD chart. ART Precision is probably using OPA1612t (see review thread this was cross-posted from). Schitt Mani 2 (OPA1612)? Yep. Fosi Box X5 (OPA1612)? Knew that one. Emotiva XPS-1 (LT1115CSW)? Yup. U-Turn Pluto 2? A little better with an LM837, but still not as good as an old 5534A. Large numbers of the "top of the chart" are basically Dieselgate style frauds (one could argue), assuming they are unlikely to be used with MC cartridges. Sutherland KC Vibe, to its credit, probably does fine with its expensive opamp that's actually probably on par with an old NE5534A. But it's not cheap.
----
If anyone cares, the Marantz AV8805 uses an NJM8068G, which also has fairly low current noise. I grabbed a schematic for the old Denon AVR-X4800H, and it was basically the same circuit with an NJM8080G. I have no idea what that part is, but probably some variant of an NJM4580? Which is also suitable for phono. Point being, Denon and Marantz are building phono preamps into their receivers with arguably better parts selection than you get from standalone boxes. Sad. If these devices had been tested properly with a cartridge on the input, an AV receiver would probably be ahead of most of them. FWIW, Parasound is using NJM5534D on the Zphono (on sale for $99), so also not a scam there. "Legacy" companies apparently know a thing or two.
It's just shocking that these clowns would all sell sub-$200 boxes with "low noise" opamps that fail to perform in actual use with the cartridges they are likely to be used with. The "low noise" claims by the manufacturers are basically just scams in many cases (intentional or not), and the test results are completely bogus since the product immediately ceases functioning as tested immediately upon being placed into a real world application. Cambridge is discreet, so there is hope there also. Fortunately, I have one on the way, so we shall soon see. Sadly, @pma was far more right than he knew. It's not just Fosi that is incompetent. It's almost everyone selling cheap boxes that wants to toot their horn about "low noise" and measure good.
We could actually probably start re-ranking these things. Anyone care to take a stab at it? (Not that it really matters, but it would be a fun little exercise to see just how far some would fall...
---
Out of curiosity, I decided to do a little Google-fu to see just how many other cheapo "good measuring" phono preamps were "guilty" of using OPA1612 or some similar device with a needlessly large amount of current noise in order to make bogus "low noise" claims. And it's .... a whole lot of op-amp based devices on the top of the SINAD chart. ART Precision is probably using OPA1612t (see review thread this was cross-posted from). Schitt Mani 2 (OPA1612)? Yep. Fosi Box X5 (OPA1612)? Knew that one. Emotiva XPS-1 (LT1115CSW)? Yup. U-Turn Pluto 2? A little better with an LM837, but still not as good as an old 5534A. Large numbers of the "top of the chart" are basically Dieselgate style frauds (one could argue), assuming they are unlikely to be used with MC cartridges. Sutherland KC Vibe, to its credit, probably does fine with its expensive opamp that's actually probably on par with an old NE5534A. But it's not cheap.
----
If anyone cares, the Marantz AV8805 uses an NJM8068G, which also has fairly low current noise. I grabbed a schematic for the old Denon AVR-X4800H, and it was basically the same circuit with an NJM8080G. I have no idea what that part is, but probably some variant of an NJM4580? Which is also suitable for phono. Point being, Denon and Marantz are building phono preamps into their receivers with arguably better parts selection than you get from standalone boxes. Sad. If these devices had been tested properly with a cartridge on the input, an AV receiver would probably be ahead of most of them. FWIW, Parasound is using NJM5534D on the Zphono (on sale for $99), so also not a scam there. "Legacy" companies apparently know a thing or two.
It's just shocking that these clowns would all sell sub-$200 boxes with "low noise" opamps that fail to perform in actual use with the cartridges they are likely to be used with. The "low noise" claims by the manufacturers are basically just scams in many cases (intentional or not), and the test results are completely bogus since the product immediately ceases functioning as tested immediately upon being placed into a real world application. Cambridge is discreet, so there is hope there also. Fortunately, I have one on the way, so we shall soon see. Sadly, @pma was far more right than he knew. It's not just Fosi that is incompetent. It's almost everyone selling cheap boxes that wants to toot their horn about "low noise" and measure good.
We could actually probably start re-ranking these things. Anyone care to take a stab at it? (Not that it really matters, but it would be a fun little exercise to see just how far some would fall...
StevenSGT98
Active Member
- Joined
- Jan 19, 2025
- Messages
- 102
- Likes
- 103
FWIW the WiiM Ultra is using SA5532A. One on each channel.
Similar threads
