I have two units and will measure again tonight or tomorrow to reconfirm the numbers. The QA401 settings are pretty straightforward and integrated noise is computed in the standard RMS manner - a point I confirmed with Matt
I have two units and will measure again tonight or tomorrow to reconfirm the numbers. The QA401 settings are pretty straightforward and integrated noise is computed in the standard RMS manner - a point I confirmed with Matt.
My calculation is 30uV integrated noise divided by 631 (56 dB) = 47.5nV input referred noise.
Take a rough guide for total equivalent noise density, divide the above figure by rt bandwidth and I get 330 pV/rt Hz.
My calculation is 30uV integrated noise divided by 631 (56 dB) = 47.5nV input referred noise.
Take a rough guide for total equivalent noise density, divide the above figure by rt bandwidth and I get 330 pV/rt Hz.
Here are the measurements just completed
Re-measurement of X-Altra Final Unit 13 August 2020
MM Input shorted, Gain 36 dB: 11uV RMS
1kHz spot noise: -135 dBV
MC Input with 47 Ohm 1206 TF resistor ‘47R0’: 27 uV RMS
1 kHz Spot noise: -120 dBV
(MC Amp gain setting load 494 Ohms)
MC Amplifier with inputs OPEN Circuit: 17uV
1 kHz Spot noise: -124 dBV
Inputs to MC Amplifier SHORTED (Gain approx. 152x or 43.6 dB): 87uV RMS
1 kHz Spot noise: -108 dBV
(Noise floor with QA401 connected but preamp powered OFF is -140 dBV at 1kHz)
These measurements are not easy to make since there is a lot of LF thermal noise and very high gain that confounds the readings. I will measure the other unit later tonight (I've changed some gain settings on the MC side of that one, but we can back calculate the equivalent input noise in any event)
Re-measurement of X-Altra Final Unit 13 August 2020
MM Input shorted, Gain 36 dB: 11uV RMS
1kHz spot noise: -135 dBV
MC Input with 47 Ohm 1206 TF resistor ‘47R0’: 27 uV RMS
1 kHz Spot noise: -120 dBV
(MC Amp gain setting load 494 Ohms)
MC Amplifier with inputs OPEN Circuit: 17uV
1 kHz Spot noise: -124 dBV
Inputs to MC Amplifier SHORTED (Gain approx. 152x or 43.6 dB): 87uV RMS
1 kHz Spot noise: -108 dBV
(Noise floor with QA401 connected but preamp powered OFF is -140 dBV at 1kHz)
These measurements are not easy to make since there is a lot of LF thermal noise and very high gain that confounds the readings. I will measure the other unit later tonight (I've changed some gain settings on the MC side of that one, but we can back calculate the equivalent input noise in any event)
Last edited:
Here are the measurements of the 2nd unit (this was the first proto I built)
MM SHORTED 36 dB Gain: 7.6uV RMS
1 kHz Spot noise: -134 dBV
MC with 47 Ohm Resistor: 27.5uV and -118 dBV (Total Gain @ 1kHz 56 dB as per unit I posted earlier measurements of)
MM SHORTED 36 dB Gain: 7.6uV RMS
1 kHz Spot noise: -134 dBV
MC with 47 Ohm Resistor: 27.5uV and -118 dBV (Total Gain @ 1kHz 56 dB as per unit I posted earlier measurements of)
Last edited:
Scott, Yes - the input Z I calculate is 3.25 Ohms. If you short the inputs, the gain is then 494/(3.25 + any parasitic R) - assume R is zero and the gain is c. 152. Might be a bit lower if the connection resistances are 0.5 Ohms at a rough guess.
Last edited:
Hi Andrew,
With this revised -135dBV instead of the previous -124dBV, things are getting a lot better now.
To get this 11uV with shorted input, I now need an amp with 1.4nV/rtHz while using a feedback resistor of 50R, see image below.
When using the 0.22pV/rtHz MC amp with a 494R feedback resistor, I get exactly the 17uV that you specify but ... with a 470R resistor.
Using a 47R Resistor, output voltage becomes 60.55uV, see image below.
So it looks as if your input resistor is not 47R but 470R.
I have added the LTSpice model for your convenience.
Hans
Attachments
Hans, I've checked it with a meter - definitely 47 ohm. The resistor marking is 47R0 - 4.7 Ohm would be 4R7
The Zin is 3.25 Ohms ideally so I'm wondering if the 47 Ohm and the Zin form a divider. There is some additional input resistance obviously maybe 0.5Ohm as I mentioned above.
The Zin is 3.25 Ohms ideally so I'm wondering if the 47 Ohm and the Zin form a divider. There is some additional input resistance obviously maybe 0.5Ohm as I mentioned above.
Last edited:
When the input resistor is indeed 47R than the only remaining possibility is that the 494R feedback resistor is in fact 49R4.
LTspice is very accurate in calculating noise, so there must be an obvious reason.
Did you check my LTspice file ?
Hans
LTspice is very accurate in calculating noise, so there must be an obvious reason.
Did you check my LTspice file ?
Hans
I'll run your model now
Update- I'm getting the same results as you.
Thing is you are getting c. 60uV RMS and I am getting 56uV RMS but the measurement says 27uV for the total signal chain.
There is a 6 dB error somewhere.
I will have to do some more work on it in the morning. 🙂
Update- I'm getting the same results as you.
Thing is you are getting c. 60uV RMS and I am getting 56uV RMS but the measurement says 27uV for the total signal chain.
There is a 6 dB error somewhere.
I will have to do some more work on it in the morning. 🙂
Last edited:
We were always there, the noise is what it is you can't look at it in a way that makes it something else.
Just an idea: You could put a 1mV sinewave from a signal generator with a 50R source at the MC input and check whether the overall gain is indeed 630.
Hans
Hans
My 0603 49R9 0.5% thin film resistors from Susumu
read "68R". The R stands for the decade.
499R is "68A"
I've remeasured the gain of the MC stage. I input 7.33 VRMS through a 110k resistor to the input to give an input current of 66.6uA. 66.6uA x 494 = 32.92mV output. The scope is reading 32.5mV.
Note also, when I measured the MC, I output -66dB from the DAC through the DAC's 47 Ohm resistor. This made the input current a shade under 10uA (47 Ohms + 3.25 Ohm Zin) which gave 5mV out.
For the MM, I input 7.33 VRMS through an inverse RIAA filter at 1kHz that gives me 47 mV at the output. The preamp output is 2.96V. The gain is 63x which is 36 dB - i.e. correct.
Re the 60uV vs 30uV discrepancy we are seeing between the sim and the measurement. I'm using both the DAC and the A-D in single ended mode so there should be no gain accounting issues to worry about. Nevertheless, I've emailed Matt at QA to ask if there is some other gain accounting factor to be included in the calculation.
All measurements above are at 1 kHz.
(Source for the 7.33 V RMS signal is a TTi function generator, voltage readings are from TEK MDO3024 O'scope)
Note also, when I measured the MC, I output -66dB from the DAC through the DAC's 47 Ohm resistor. This made the input current a shade under 10uA (47 Ohms + 3.25 Ohm Zin) which gave 5mV out.
For the MM, I input 7.33 VRMS through an inverse RIAA filter at 1kHz that gives me 47 mV at the output. The preamp output is 2.96V. The gain is 63x which is 36 dB - i.e. correct.
Re the 60uV vs 30uV discrepancy we are seeing between the sim and the measurement. I'm using both the DAC and the A-D in single ended mode so there should be no gain accounting issues to worry about. Nevertheless, I've emailed Matt at QA to ask if there is some other gain accounting factor to be included in the calculation.
All measurements above are at 1 kHz.
(Source for the 7.33 V RMS signal is a TTi function generator, voltage readings are from TEK MDO3024 O'scope)