Measuring Frequency Response and Distortion of amplifiers (4x150w RMS and 4x80W RMS)

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
My suggestion would be to re-read post #48 and see if you can do a suite of automated test measurements that show how the 2nd harmonic % varies with signal level and other important settings. Imho that should help characterise what the device can do (compared to just two spot measurements). An example plot of step level changes is in post #3 in https://www.audiosciencereview.com/...s/tascam-us-2x2hr-loopback-measurments.29212/ but note that imho only H2 is initially of interest to benchmark as that is relatively high and needs to be confirmed as to whether it is always that high (in %) or whether it has a region where it is relatively low (ie. a sweet spot). And even if it is relatively high, then the next query is whether it can be nulled to a much lower level using 'add harmonic distortion' feature, or it can't be as jschwender has noted.

Hi trobbins

I tried several times.

48765


48766


48767


48768


48769


48770


48771


48772


48773


48774


I will try different step level later.
 

trobbins

Member
VIP Supporter
Joined
Nov 26, 2017
Messages
121
Location
Australia
You were using the 'step frequency' option in the 'stepped sine' feature. There is a 'Step level' function in RTA to see how the 2nd Harmonic level changes with change in amplitude of test signal.
 

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
You were using the 'step frequency' option in the 'stepped sine' feature. There is a 'Step level' function in RTA to see how the 2nd Harmonic level changes with change in amplitude of test signal.

Hi trobbins

I got it.

48820


48821


And I also attached the mdat file for details.
 

Attachments

  • STEP LEVEL.mdat
    2.3 MB · Views: 6

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
I'd suggest -20dBfs, based on the level plot of H2.

Hi trobbins

I tried again and got new results, should I focus on THD only? or a balance between THD and THD+N?

LEVEL: -20dBFS THD: 0.017% THD+N: 3.17%


48827


LEVEL: -15dBFS THD: 0.026% THD+N: 1.77%

48829


BTW, is this soundcard good for distortion measurement of car amplifiers ?
 

trobbins

Member
VIP Supporter
Joined
Nov 26, 2017
Messages
121
Location
Australia
I'd suggest just trying to appreciate how low you can get 2H%. THD and Noise level (as in THD +N) is only a secondary issue imho. Perhaps vary the Gain1 setting and see if - 20dBfs is still the sweet spot for minimum %2H (eg. do a step level scan from say - 30 to -10 at 2dB steps.
 

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
I'd suggest just trying to appreciate how low you can get 2H%. THD and Noise level (as in THD +N) is only a secondary issue imho. Perhaps vary the Gain1 setting and see if - 20dBfs is still the sweet spot for minimum %2H (eg. do a step level scan from say - 30 to -10 at 2dB steps.

I found -19dBFS is the best sweet spot.

48836


Signal Gen and RTA:

48835
 

trobbins

Member
VIP Supporter
Joined
Nov 26, 2017
Messages
121
Location
Australia
You have two output options (line and phones) so its worthwhile checking what is the low 2H% sweetspot for each of them, and also measuring with a DMM what that output voltage is (Vrms).

The % distortion comes from the DAC and the ADC combined, so a similar effort is needed to check (using the -19dBFS level) how the 2H% changes with GAIN1 level and whether there is a sweet spot for the GAIN1 level setting.

If it is difficult to detect the 2H level due to noise floor then its worth lowering the noise floor level using averages =2, or 4 etc, and using a higher FFT rate.
 

trobbins

Member
VIP Supporter
Joined
Nov 26, 2017
Messages
121
Location
Australia
If it is difficult to detect the 2H level due to noise floor then its worth lowering the noise floor level using averages =2, or 4 etc, and using a higher FFT rate.
I forgot to add that ticking the 'coherent averaging' option in the RTA options section for 'Distortion Settings' is also a benefit (and avoids the use of large FFT lengths).
 

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
You have two output options (line and phones) so its worthwhile checking what is the low 2H% sweetspot for each of them, and also measuring with a DMM what that output voltage is (Vrms).

For LEFT line output, signal 1KHZ SINE

The output voltage is 0.005V @-19dBFS, it's too low. :rolleyesno:

I increase the level in REW , the output voltage is 0.251V @ -6dBFS.
 

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
You have two output options (line and phones) so its worthwhile checking what is the low 2H% sweetspot for each of them, and also measuring with a DMM what that output voltage is (Vrms).

I change the output and use PHONES as the output like this:

48848


-15dBFS is a good starting point.

48849


2nd Harmonic Distortion: 0.0090%, but the output voltage of PHONES is only 0.004V.:(

48850
 

trobbins

Member
VIP Supporter
Joined
Nov 26, 2017
Messages
121
Location
Australia
That gives more insight in to what your device can do. Note that REW works better with smaller FFT length (ie. 64k) and a lot more averages (eg. 32) when using coherent averaging - it may be worth seeing if that gives you more headroom between the 2HD level and the noise floor. If you do get more headroom, then you could turn down the GAIN1 setting by say 10dB and see if the HD% changes or not.

Note that when you test an amp, the amp's output level is a lot higher than the line or phones output level.

Are the above results the same if you modify the UMC Control Panel settings for Volume for inputs and/or outputs?
 

jschwender

Member
Joined
Mar 16, 2021
Messages
140
Location
GERMANY
More  
Front Speakers
Nubert digital pro
Other Speakers or Equipment
Philips dss940
These audio devices have not so sophisticated amplifiers built in, as the intended use case is microphone. And when using the symmetric input, the H1 and H3 cancel each other out, so they are significantly lower, compared to an asymmetric signal use case.
 

trobbins

Member
VIP Supporter
Joined
Nov 26, 2017
Messages
121
Location
Australia
Agree about the likely better performance of a balanced input configuration. I reckon it is still worthwhile trying to characterise the DAC and ADC parts and to gauge the performance of each part, as to confirm what is the best way to use the device for the OP. It may well be that the OP only has amps that contribute substantially more distortion than the UMC, which would be a good outcome as that removes the UMC from measurement integrity.
 

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
That gives more insight in to what your device can do. Note that REW works better with smaller FFT length (ie. 64k) and a lot more averages (eg. 32) when using coherent averaging - it may be worth seeing if that gives you more headroom between the 2HD level and the noise floor. If you do get more headroom, then you could turn down the GAIN1 setting by say 10dB and see if the HD% changes or not.

Hello trobbins,

Happy New Year!

I changed settings in REW and decrease the GAIN1 to the level near Min, and improved the measurement.

48910


Phones is the output, GAIN 1 to min (near)

48911


Because levels below -12dBFS are too low for output, so I choose -6dBFS

48912


Looks not too bad. 2nd 0.0068%, THD: 0.10%


48914


This soundcard is not easy to find out the sweet spot . :-)
 
Last edited:

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
These audio devices have not so sophisticated amplifiers built in, as the intended use case is microphone. And when using the symmetric input, the H1 and H3 cancel each other out, so they are significantly lower, compared to an asymmetric signal use case.
Hello jschwender,

Happy New Year!

This UMC 202HD has too high 2nd distortion, I cann't increase the output level much, cause distortion will increase quickly, but if the output level is too low, the amp which will be measured will have a low output level.

Maybe this is not good for frequency response and distortion measurement.
 

trobbins

Member
VIP Supporter
Joined
Nov 26, 2017
Messages
121
Location
Australia
Now that the HD levels are better understood, you could try using it with an amp and see if the amps you have in mind have much more distortion even if your output signal is made a lot higher. That may uncover other issues or ways to use your soundcard to get some measurement tool value out of it.

Another option is to make up a buffer amp to boost the phones or line output signal - even a simple 5534 type opamp with 20dB gain would likely keep HD low. I've had to make up buffer amps as some testing needs a higher signal level than many soundcards can provide. I don't see that anyone has reverse engineered the UMC202HD, but if you were keen you could identify the parts/circuitry for the line or phones output. Quite a bit of reverse engineering has been done on some other soundcards to try and improve their performance, but the incentive to do that for your soundcard may be quite low.

Btw, THD measurement level can be ambiguous when the focus is on HD levels, as THD calculation sucks in other higher levels - you can try and adjust what THD is reported by checking the 'enable high pass' and low pass options in the distortion settings popup.
 

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
Now that the HD levels are better understood, you could try using it with an amp and see if the amps you have in mind have much more distortion even if your output signal is made a lot higher. That may uncover other issues or ways to use your soundcard to get some measurement tool value out of it.

Another option is to make up a buffer amp to boost the phones or line output signal - even a simple 5534 type opamp with 20dB gain would likely keep HD low. I've had to make up buffer amps as some testing needs a higher signal level than many soundcards can provide. I don't see that anyone has reverse engineered the UMC202HD, but if you were keen you could identify the parts/circuitry for the line or phones output. Quite a bit of reverse engineering has been done on some other soundcards to try and improve their performance, but the incentive to do that for your soundcard may be quite low.

5534 opamp is a bit complicated for me.

Acturally, I can try to connect the output of UMC202HD to a CD player with a high quality preamp such as Pioneer P99 which can provide 4-5V RCA output to the amplifier I will measurement.

Btw, THD measurement level can be ambiguous when the focus is on HD levels, as THD calculation sucks in other higher levels - you can try and adjust what THD is reported by checking the 'enable high pass' and low pass options in the distortion settings popup.

48918


With High pass and Low Pass,the result looks better with low distortion. -5dBFS maybe a new sweet spot ,and the output of UMC202HD is around 9 o'clock.
 

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
I tried to measured the amp.

Connection

49089


Amp output level, still too low.

49090


SoundCard settings:

49091



Frequency Response:

49092


Distortion:

49088
 
Top Bottom