Yes, I played a bit with various cable configurations but could not get it better than the best configuration/measurement yesterday. The TS - RCA converter is a bit "flimsy" so could be nice to use a balanced configuration from amp to 2i2. And also nice with a TS - RCA cable for the output so I don't need the converter.
I measured the other mono block and they are both quite similar. I am so far happy with the measurement from yesterday and from this evening where I experimented a bit with cables. When I compare distortion measurement presented in amp forums I think my looks quite good.
I will print out the amp to balanced schematic presented earlier in this thread.
I measured the other mono block and they are both quite similar. I am so far happy with the measurement from yesterday and from this evening where I experimented a bit with cables. When I compare distortion measurement presented in amp forums I think my looks quite good.
I will print out the amp to balanced schematic presented earlier in this thread.
Attachments
Well the issue is not so much with the sound card as with the cables 
But I blew up my 2i2 a while ago, and was looking for a new card, and decided on a Motu M4. This is more expensive than a 3rd Gen scarlett 2i2.
However there was no stock and Focusrite support sent me a new DAC chip for my 2i2 (for free, 7 years after my purchase!) So I replaced the codec and it is now working fine again, and I'm not feeling a need to purchase the motu.
I suspect you would potentially have the same issues with a different card. To get the lowest noise you really need balanced connections, which is why I have revisited my little project to create some balanced to SE and SE to balanced converters
edit: link to someone who was having problems with a gen1 2i2 that implemented the balanced to se and was very happy Got Focusrite Scarlett 2i2 (2nd gen) for measurements. What's next? Note I'd recommend 7.5K or 10K resistors rather than the 2K's in that post. Decoupling caps should also be used, and resistors ideally 0.1%
Tony.
But I blew up my 2i2 a while ago, and was looking for a new card, and decided on a Motu M4. This is more expensive than a 3rd Gen scarlett 2i2. However there was no stock and Focusrite support sent me a new DAC chip for my 2i2 (for free, 7 years after my purchase!) So I replaced the codec and it is now working fine again, and I'm not feeling a need to purchase the motu.
I suspect you would potentially have the same issues with a different card. To get the lowest noise you really need balanced connections, which is why I have revisited my little project to create some balanced to SE and SE to balanced converters
edit: link to someone who was having problems with a gen1 2i2 that implemented the balanced to se and was very happy Got Focusrite Scarlett 2i2 (2nd gen) for measurements. What's next? Note I'd recommend 7.5K or 10K resistors rather than the 2K's in that post. Decoupling caps should also be used, and resistors ideally 0.1%
Tony.
Last edited:
I can double that your actual sound card is definetely not weakest part in your setup.
Noise level is quite low. You may get lower with averaging.
OK!
Yes, I also think the PSU now is satisfying. I would also like to spend some time now listening to music
But, is there an option for an external sinus generator that has low enough distortion/noise so it can be used for distortion/noise measurement using 2i2 and REW?
I tried a cheap function generator I had but that was quite bad.
What about older analog sinus generators? .....like vintage HP etc....are they good enough? ....any suggestions?
It could also be just a 1 kHz test tone kit.
Reason is that I can have a more simple setup and avoid the "closed loop" I have today. I think it will look nicer regarding the cable routing. Then I will go for a balanced input solution for 2i2.
Yes, I also think the PSU now is satisfying. I would also like to spend some time now listening to music
But, is there an option for an external sinus generator that has low enough distortion/noise so it can be used for distortion/noise measurement using 2i2 and REW?
I tried a cheap function generator I had but that was quite bad.
What about older analog sinus generators? .....like vintage HP etc....are they good enough? ....any suggestions?
It could also be just a 1 kHz test tone kit.
Reason is that I can have a more simple setup and avoid the "closed loop" I have today. I think it will look nicer regarding the cable routing. Then I will go for a balanced input solution for 2i2.
You can build your own. See here:
https://www.diyaudio.com/forums/equ...n-audio-range-oscillator-923.html#post6315598
Send a message to vicnic to have more information how to order the pcb.
https://www.diyaudio.com/forums/equ...n-audio-range-oscillator-923.html#post6315598
Send a message to vicnic to have more information how to order the pcb.
Victor's oscillator (THD -140 dB garantied). He doesn't sell it free on eBay now, but he sells it through private messages on email, etc.
Oh, I see he has already been mentioned.
You'll avoid group loop with an external generator, and things will be a bit easier, but you won't have a better spectrum than this (last yours):
If you want a lower noise shelf then use averaging (I don't know about REW, I use ARTA).
Attachments
The above spectrum was obtained by moving around with 2i2 until I got the lowest peaks on the monitor. Maybe it is easier to get an optimal setup using an external generator and there is also some fun in building it
I used averaging: Exponential 0.50 .....what that means I don't know.....there is also 0.75 etc. which probably performs a "harder" averaging.....
I am satisfied with the above spectrum.
I used averaging: Exponential 0.50 .....what that means I don't know.....there is also 0.75 etc. which probably performs a "harder" averaging.....
I am satisfied with the above spectrum.
Regarding external generator I took the easy part and ordered this kit:
1 kHz Oscillator
.....then I get everything and can build it in an evening and I don't have to think about drilling holes, mounting connectors, building PSU etc.
The distortion is low enough (< 0.0002%) for distortion measurement. Then I like that the circuit is simpler than most I have seen and it uses an old "HP-trick" by using a light bulb to control the feedback of the oscillator?
I ordered an extra light bulb......just in case.....but hope it will last a long time.
1 kHz Oscillator
.....then I get everything and can build it in an evening and I don't have to think about drilling holes, mounting connectors, building PSU etc.
The distortion is low enough (< 0.0002%) for distortion measurement. Then I like that the circuit is simpler than most I have seen and it uses an old "HP-trick" by using a light bulb to control the feedback of the oscillator?
I ordered an extra light bulb......just in case.....but hope it will last a long time.
Yes, and it seems to be a central part of the design and the reason why the circuit can be quite simple and with very low distortion. Schematic is on page 13:
https://www.akitika.com/documents/2PPMOscillatorAssemblyManualRev1p05.pdf
What is the performance of the more complex DIY circuits with more op-amps and external transistors?
https://www.akitika.com/documents/2PPMOscillatorAssemblyManualRev1p05.pdf
What is the performance of the more complex DIY circuits with more op-amps and external transistors?
Gone are the days of the RA53/4 Wien bridge audio oscillator ,luckily a built a couple with those fitted but if any of you lucky guys can get a REAL one -
Audio Test Oscillator
Also talks about the lamp type and others.
Audio Test Oscillator
Also talks about the lamp type and others.
MEPER, given one of your aims is to measure distortion at 1kHz, you also have the option of tweaking a 1kHz sinewave signal from REW to suppress the individual harmonic distortions of the output signal of the 2i2 to very low levels (much lower than the inherent THD from the 2i2 output signal, and getting down towards 0.0002%), and of using a 1kHz tone that is exactly frequency matched to the FFT algorithm to give better dynamic range of the amp's output measurement using the 2i2.
Last edited:
Ok, so I could also filter my Volcraft FG-250D function generator (set to 1 kHz) to get low noise and distortion?
The nice thing with the kit I get is that everything is included and easy to build. I could even go for "hand matched" components for the filters if I want and get even better performance.
I could also just buy an Audio Precision x555....or what is it called? .....but that will be more than 100 USD....or it could be a vintage HP distortion analyzer? ....with those I would also be able to check the phase of the 2nd harmonic relative to the fundamental?
But this is not so much Power Supply anymore......
Keeping it to a bit to Power Supply topic I think it is an advantage that 1 kHz kit has batteries as PSU to break any ground loop in my current setup with 2i2.
Also I think the conclusion of the amp measurement is that PSU has a lot to say for the performance of the amp. A small bit of imbalance in PSU is visible in measurement. The first immediate sign of imbalance is if -+ rails is not 100% "numerical" identical. With this re-built of the rectifiers that was one of the outcomes that I got rails 100% symmetrical (and stable). Another outcome was that I got higher rails and don't waste "switching power" in rectifiers.
The nice thing with the kit I get is that everything is included and easy to build. I could even go for "hand matched" components for the filters if I want and get even better performance.
I could also just buy an Audio Precision x555....or what is it called? .....but that will be more than 100 USD....or it could be a vintage HP distortion analyzer? ....with those I would also be able to check the phase of the 2nd harmonic relative to the fundamental?
But this is not so much Power Supply anymore......
Keeping it to a bit to Power Supply topic I think it is an advantage that 1 kHz kit has batteries as PSU to break any ground loop in my current setup with 2i2.
Also I think the conclusion of the amp measurement is that PSU has a lot to say for the performance of the amp. A small bit of imbalance in PSU is visible in measurement. The first immediate sign of imbalance is if -+ rails is not 100% "numerical" identical. With this re-built of the rectifiers that was one of the outcomes that I got rails 100% symmetrical (and stable). Another outcome was that I got higher rails and don't waste "switching power" in rectifiers.
Definitely breaking the loop between output and input on the 2i2 I think will help. I'm interested to see the results you get if you feed in the function generator to the 2i2 because I have never been sure whether the inherent distortion is on the output, then input or a combination of both, so will be very good to see the performance with a known low distortion source.
Tony.
Tony.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.