The phono stage Env(tot) is 1.6 uV and is swamped by the MM cart noise. There is no noise current contribution on the MM input. The MC Env(tot) is C. 36 nV - there is essentially no current noise component either since the input Z of the source is a few ohms. The SNR of the MC stage ref 500 uV is 83 dB.
The unweighted MM noise floor is c. -124 dBV at 1 kHz as measured on the output with the input shorted. With a cart connected (500 mH and 1.35 k ohms), the SNR ref 5 mV is 75dB. The cart noise is 3.1 uV - this is the combo of the cart//47 k load resistor.
All BW’s are 20-20 kHz.
I can’t comment more at this stage you will have to wait until it’s published a bit later this year. I used a QA401 for the measurements, the spread sheet on my website for noise calcs (based on SY’s SS from Natsemi’s AN104) And LTspice sims.
The unweighted MM noise floor is c. -124 dBV at 1 kHz as measured on the output with the input shorted. With a cart connected (500 mH and 1.35 k ohms), the SNR ref 5 mV is 75dB. The cart noise is 3.1 uV - this is the combo of the cart//47 k load resistor.
All BW’s are 20-20 kHz.
I can’t comment more at this stage you will have to wait until it’s published a bit later this year. I used a QA401 for the measurements, the spread sheet on my website for noise calcs (based on SY’s SS from Natsemi’s AN104) And LTspice sims.
Bonsai,
Sorry, but this time its me, not because of nit-picking but because IMO you are making a bit of a jungle with figures, where we try to come to clear definitions.
You mention a S/N of 75dB for the 500mH/1.35K MM Cart ref 5mV@1Khz, terminated with 200pF//47K.
So far so good, that's 100% clear, no further discussion possible.
But then you mention a S/N of 83dB for an MC of a few Ohms, which is exactly the outcome of 0.5mV and 36nV equivalent input noise @1Khz.
This head-amp however with a gain of 20dB, producing 5mV@1Khz output for a 0.5mV@1Khz Cart according to your specs, has to be used in combination with the MM amp.
The much larger MM's input noise (ca 800nV) will be added to the (10x36nV=360nV) output noise of the MC head-amp, to be further processed by the Riaa curve.
So important is again the S/N of this whole chain referred to the input @ 1Khz.
A quick check based on this 20dB gain for the head amp plus MM amp, brings the S/N then to ca. 77dB after Riaa processing.
With a 47R source resistance I even get a S/N of ca 73dB, still unweighted, quite a bit away from your >82dB ??
So it seems that you are ruining the excellent head-amp noise figures by the much noisier MM amp.
When the head-amps gain would be higher than 20dB, these figures could be reduced dramatically, but that's not in line with what your spec's are telling.
Here is a link to what SNR means: signal power divided by noise power,
where S/N means: signal voltage divided by noise voltage, that's what we are doing.
What is Signal-To-Noise Ratio (SNR)?
Hans
Sorry, but this time its me, not because of nit-picking but because IMO you are making a bit of a jungle with figures, where we try to come to clear definitions.
You mention a S/N of 75dB for the 500mH/1.35K MM Cart ref 5mV@1Khz, terminated with 200pF//47K.
So far so good, that's 100% clear, no further discussion possible.
But then you mention a S/N of 83dB for an MC of a few Ohms, which is exactly the outcome of 0.5mV and 36nV equivalent input noise @1Khz.
This head-amp however with a gain of 20dB, producing 5mV@1Khz output for a 0.5mV@1Khz Cart according to your specs, has to be used in combination with the MM amp.
The much larger MM's input noise (ca 800nV) will be added to the (10x36nV=360nV) output noise of the MC head-amp, to be further processed by the Riaa curve.
So important is again the S/N of this whole chain referred to the input @ 1Khz.
A quick check based on this 20dB gain for the head amp plus MM amp, brings the S/N then to ca. 77dB after Riaa processing.
With a 47R source resistance I even get a S/N of ca 73dB, still unweighted, quite a bit away from your >82dB ??
So it seems that you are ruining the excellent head-amp noise figures by the much noisier MM amp.
When the head-amps gain would be higher than 20dB, these figures could be reduced dramatically, but that's not in line with what your spec's are telling.
Here is a link to what SNR means: signal power divided by noise power,
where S/N means: signal voltage divided by noise voltage, that's what we are doing.
What is Signal-To-Noise Ratio (SNR)?
Hans
The noise in a signal chain is always added rss from stage to stage so the equivalent heating power is preserved. Where did you get that number for the MM input noise, Bonsai didn't state the gain or where the weighting was applied? I would assume the head amp noise was stated as flat at 20dB and the MM stage as weighted at at least 40dB of gain.
Last edited:
The 83 dB SNR on the MC input is indeed jut for the MC stage. I should have made that clearer. Apologies for that.
It was not that difficult to get the input noise for the MM stage.
With 5mV Cart, specified S/N is 75dB, giving 889nV equivalent EIN@1Khz.
With input shorted, EIN@1Khz was reported to be -124dB or 630nV.
To get this in a Sim, I had to use a Fet amp with En=5.6nV/rtHz and Enk=30Hz.
See the two images at the left for resp EIN@1Khz with Cart and with short circuited input giving resp. 889nV and 611nV, quite close to the figures above.
The third image gives the straight input noise before Riaa, ca 800nV.
This is how I got my figure.
I also have the MC head amp with 220pV/rtHz, giving a perfect 36nV for 1R source.
Connecting this to the MM amp, I got the resp 77dB and 73dB S/N ref 0.5mV@1Khz.
Hans
With 5mV Cart, specified S/N is 75dB, giving 889nV equivalent EIN@1Khz.
With input shorted, EIN@1Khz was reported to be -124dB or 630nV.
To get this in a Sim, I had to use a Fet amp with En=5.6nV/rtHz and Enk=30Hz.
See the two images at the left for resp EIN@1Khz with Cart and with short circuited input giving resp. 889nV and 611nV, quite close to the figures above.
The third image gives the straight input noise before Riaa, ca 800nV.
This is how I got my figure.
I also have the MC head amp with 220pV/rtHz, giving a perfect 36nV for 1R source.
Connecting this to the MM amp, I got the resp 77dB and 73dB S/N ref 0.5mV@1Khz.
Hans
Attachments
Bonsai,
If you could increase the Head-amp's gain to 32dB, the additional noise contribution from the MM amp would be reduced to ca 1dB, but I don't know if this is possible.
But 26dB also gives already a significant gain.
Hans
If you could increase the Head-amp's gain to 32dB, the additional noise contribution from the MM amp would be reduced to ca 1dB, but I don't know if this is possible.
But 26dB also gives already a significant gain.
Hans
Last edited:
Here are the numbers straight from the QA 401 measurements. The BW is 20 kHz and no weighting.
I get a shade over 80 dB signal to noise for the MC Ref 500 uV.
Because of the current injection input stage, if you just short the input, the gain is very high - so I soldered a 47 ohm SMD resistor across the inputs.
(I am using a Denon DL103 which gives 390 uV and 30 Ohm Rgen so OP current c. 13 uA - this is considered a low output cart in the context of a current injection stage)
I get a shade over 80 dB signal to noise for the MC Ref 500 uV.
Because of the current injection input stage, if you just short the input, the gain is very high - so I soldered a 47 ohm SMD resistor across the inputs.
(I am using a Denon DL103 which gives 390 uV and 30 Ohm Rgen so OP current c. 13 uA - this is considered a low output cart in the context of a current injection stage)
Attachments
Last edited:
~Whe have said this before, a S/N of 75dB(A) is more than adequate, more does not make sense.
Since this Head amp has extremely low noise of 220pV/rtHz, it could be used for low output Carts like 0.25mV 33R, not possible with an MC amp with higher noise.
See image below.
Hans
Attachments
Bonsai's picture of his signal chain makes sense, I don't know where the wide disagreement comes from. Both RIAA and A weighting are normalized to 0 at 1kHz, since with MC all noise sources are flat with frequency why not simply talk about spot noise and gain at 1kHz. MM has potentially a current noise component that is weighted by the cartridge equivalent circuit so it has to be factored in separately.
BTW he is using a current mode MC stage so the gain depends on the cartridge, I'll bet that Sonic Labs cartridge might cause some overload issues.
Last edited:
I is not my intention to turn this into a who's right and who's wrong, so I will back off after this reply.
In your drawing you show a 20dB head amp having 220pV/rtHz with a 47R resistor at its input.
A 47R alone already has 125nV over a 20Khz BW.
In combination with your Head amp this increases to 129nV input noise.
However you show 47.5nV input noise in the drawing,
When I take 4.7R instead we are on exactly the same line, so this is what you must have used.
With 20dB gain and this 47.5nV, output noise from the head amp is 475nV.
To get the 30uV at the output of your 36dB@1Khz Riaa MM amp, its input noise should be 423nV or 3nV/rtHz.
This would give 77dB S/N for the given Cart and -130dBV noise when shorted, not exactly in line with 75dB S/N and -124dBV specs.
So there are a number of inconsistencies in your figures.
Hans
Scott, the MC gain can be switched down using load resistors on the output to 83 ohms min. The Rin of the MC stage is 3.25 Ohms so the total input current is 117uA with Salas’s cart. . The best the stage can handle as it is about 70uA in for 5 mV out vs his 117uA. Either the overload will be reduced to 24 dB or the MC gain setting load will have to be reduced to 42 ohms.
I see 7uV with input shorted in the second half of the figure. The 30uV in the first half of the figure can only be referred back to the input at 56dB of gain.
A cassette head puts out about the same ( 0.3 - 0.5) mv as a MC cartridge. coil resistance varies from 150 to 350 ohms. the best non dolby s/n I have seen there is 62 db. the hiss generated here is also the high frequency stuff that is most audible.
the quietest MC preamp should be somewhere around 75db. (RIAA vs NAB factored in)
the quietest MC preamp should be somewhere around 75db. (RIAA vs NAB factored in)
Scott,
7uv out with shorted MM input means -139dBV EIN for 36dB gain @ 1Khz.
Specification is EIN of -124dBV @ 1 Khz, meaning 39uV at the output instead of 7uV.
Just try to do the math and you will find the same inconsistencies in the figures, like the mentioned 47R that in fact is 4.7R meaning a S/N of 80db with a 4.7R Cart.
There must be a reason for all this, maybe the setting of the QA400 ?
It’s probably just a matter of going again through all figures again and line them up in a consistence way.
I would be more than willing to assist.
Hans
7uv out with shorted MM input means -139dBV EIN for 36dB gain @ 1Khz.
Specification is EIN of -124dBV @ 1 Khz, meaning 39uV at the output instead of 7uV.
Just try to do the math and you will find the same inconsistencies in the figures, like the mentioned 47R that in fact is 4.7R meaning a S/N of 80db with a 4.7R Cart.
There must be a reason for all this, maybe the setting of the QA400 ?
It’s probably just a matter of going again through all figures again and line them up in a consistence way.
I would be more than willing to assist.
Hans