Electronics > Projects, Designs, and Technical Stuff

Audio Measurement Pre-Amplifier

<< < (4/4)

udok:

--- Quote from: toli on August 17, 2019, 07:02:35 pm ---udok, lets not forget what this pre-amplifier is meant to do. It wasn't designed to operate with a 20dBu signal right at the input of the op-amp, its only meant to measure the distortion of that signal.

--- End quote ---

If it never sees 20dBu at the input, this is not a problem.  Anyway it is simple to change the supply to +-15 Volt in the future if needed.


--- Quote ---Using a voltage divider at the input to bring it down to a voltage level that is more comfortable for the opamp is completely ok from my point of view. Especially when you consider the fact that the output signal must be 2Vrms or lower to be within the linear range of my sound-card. BTW, the 12V supply isn't limiting this here, the input protection circuit will kick in before this happens, as it wasn't designed to have more than a 2Vrms signal at the opamps input. 20dBu out of range? No problem, switch to a higher range, all the way up to 200Vrms  ;)

--- End quote ---
Using a divider at the input is bad for SNR and you get more THD and noise from the Opamps.   Better use a divider in front of
the ADC as you could use low value resistors there.
If you use a JFet Opamp you could even increase the input impedance to 1 MOhm as 100k is often on the low side.


--- Quote ---Absolute reading accuracy is also not too critical here, I'm totally ok with 1-2% measurement error as the readout is meant mostly for measuring output power of amplifiers.

--- End quote ---

That circuit is a lot of complexity for 2% measurements.  For power amps a simple voltage divider is all you need.
But if the circuit work for your needs, it is ok.  There are always compromises to be taken.

udok:

--- Quote from: magic on August 17, 2019, 07:49:57 pm ---
--- Quote from: udok on August 17, 2019, 05:29:33 pm ---Better use a LM4562 or AD795.  They have lower noise

--- End quote ---
No, they are horrible. Their current noise of a few pA/rtHz will annihilate whatever voltage noise advantage they have, in this circuit. There are better bipolar opamps for 1.5k source impedance.

--- End quote ---

Horrible is somwhat exaggerated....
1.6 pA * 1kOhm of source impedance is 1.6 nV , the 1 kOhm alone has 4 nV of Gaussian noise.


--- Quote ---
--- Quote from: udok on August 17, 2019, 05:29:33 pm ---Maybe it would be better to use an Audio transformer for the balanced to single ended conversion?

--- End quote ---
Haven't those been ditched because of distortion and frequency response inaccuracy?

--- End quote ---
No, but because the good ones are expensive and not small.
The top line of Audio Precisions Analyzers still use them in the generator output stage.

toli:

--- Quote from: udok on August 17, 2019, 08:47:07 pm ---Using a divider at the input is bad for SNR and you get more THD and noise from the Opamps.   Better use a divider in front of
the ADC as you could use low value resistors there.
If you use a JFet Opamp you could even increase the input impedance to 1 MOhm as 100k is often on the low side.

--- End quote ---
There's the issue of high voltage at the input which means the input must have a divider, simply because I want to measure output of amplifiers directly with this.
Regarding 1Mohm, I doubt it will be practical. The output impedance of the divider will result in significantly increased distortion, as magic mentioned earlier in this thread.


--- Quote from: udok on August 17, 2019, 08:47:07 pm ---That circuit is a lot of complexity for 2% measurements.  For power amps a simple voltage divider is all you need.
But if the circuit work for your needs, it is ok.  There are always compromises to be taken.

--- End quote ---
What I was really after is the THD(+N) in the audio band, and getting low enough noise density on the sensitive ranges. The absolute amplitude accuracy of the readout isn't of significant importance to me and 1% is completely acceptable. If I ever need a more accurate absolute reading, I can always connect the 34401A in parallel. So far I didn't need better than 1% accuracy when measuring output power of amplifiers. No one really cares if its 64W or 65W :)


--- Quote from: udok on August 17, 2019, 09:00:10 pm ---No, but because the good ones are expensive and not small.
The top line of Audio Precisions Analyzers still use them in the generator output stage.

--- End quote ---
Interesting, I didn't know they still use these, I was also under the impression these weren't good enough for these applications. Do you know what part # they use there or something else that is of similar spec? It'd be interesting to see what sort of spec they have (and how much they cost :scared:)

udok:

--- Quote from: toli on August 17, 2019, 09:02:04 pm ---
--- Quote from: udok on August 17, 2019, 08:47:07 pm ---Using a divider at the input is bad for SNR and you get more THD and noise from the Opamps.   Better use a divider in front of
the ADC as you could use low value resistors there.
If you use a JFet Opamp you could even increase the input impedance to 1 MOhm as 100k is often on the low side.

--- End quote ---
There's the issue of high voltage at the input which means the input must have a divider, simply because I want to measure output of amplifiers directly with this.
Regarding 1Mohm, I doubt it will be practical. The output impedance of the divider will result in significantly increased distortion, as magic mentioned earlier in this thread.

--- End quote ---

Maybe, but as suggested you could use 1 MOhm in the 200 Volt range.  Useful, if you plan to use the Diff amp for in circuit measurements.


--- Quote ---
--- Quote from: udok on August 17, 2019, 08:47:07 pm ---That circuit is a lot of complexity for 2% measurements.  For power amps a simple voltage divider is all you need.
But if the circuit work for your needs, it is ok.  There are always compromises to be taken.

--- End quote ---
What I was really after is the THD(+N) in the audio band, and getting low enough noise density on the sensitive ranges. The absolute amplitude accuracy of the readout isn't of significant importance to me and 1% is completely acceptable. If I ever need a more accurate absolute reading, I can always connect the 34401A in parallel. So far I didn't need better than 1% accuracy when measuring output power of amplifiers. No one really cares if its 64W or 65W :)

--- End quote ---

The same argument is valid for 60 dB or 100 dB THD.  Nobody cares in real life ???


--- Quote ---
--- Quote from: udok on August 17, 2019, 09:00:10 pm ---No, but because the good ones are expensive and not small.
The top line of Audio Precisions Analyzers still use them in the generator output stage.

--- End quote ---
Interesting, I didn't know they still use these, I was also under the impression these weren't good enough for these applications. Do you know what part # they use there or something else that is of similar spec? It'd be interesting to see what sort of spec they have (and how much they cost :scared:)

--- End quote ---

They are custom made Jensen Transformers.  In addition they use an opamp feedback trick to lower distortion at low frequencies. 
The circuit is described in a patent by Bruce Hofer.

magic:

--- Quote from: udok on August 17, 2019, 09:00:10 pm ---Horrible is somwhat exaggerated....
1.6 pA * 1kOhm of source impedance is 1.6 nV , the 1 kOhm alone has 4 nV of Gaussian noise.

--- End quote ---
OK, exaggerated. I thought that they are slightly worse, actually.
Nevertheless, source impedance is 1k5 due to that series protection resistor so 2.4nV/rtHz. Combined with voltage noise, 3.3nV/rtHz. That's only marginally better than the 40+ year old NE5534 :P
Meanwhile OPA2210 gets 2.3nV/rtHz. Combined with Johnson noise, it's only 5.5 vs 6, though.

Somebody have the specs of LT1028 at hand?

BTW, JRC has an interesting part, the NJM2122. Quite cheap, rated 1.5nV/rtHz and 3.6µA bias, so only 1pA/rtHz of shot noise. Not sure how close they get to theoretical limits, though. The datasheet is as helpful as you can expect from JRC :-+


--- Quote from: udok on August 17, 2019, 09:33:43 pm ---Maybe, but as suggested you could use 1 MOhm in the 200 Volt range.  Useful, if you plan to use the Diff amp for in circuit measurements.
--- End quote ---
Suggested by me? Not really, to make things clear, I suggested reducing input impedance at -20dB to 10k for lower noise and THD.

But maybe there is a solution to this madness. Make it LEGO: one box for preamp, several boxes with various dividers connecting straight onto its BNC inputs. No relays, only music in the signal path :D


--- Quote from: udok on August 17, 2019, 08:47:07 pm ---The same argument is valid for 60 dB or 100 dB THD.  Nobody cares in real life ???
--- End quote ---
Try to sell a 60dB amplifier without some good "audiophile" story behind it :P
Audio, ahem Science Review will review it, measure the lousy THD and then perform a listening test definitely concluding you are worse than the Benchmark AHB2.
 :-DD
I'm not even kidding, check them out.

Navigation

[0] Message Index

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod