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

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
Hello John,

How is it going?

I'm preparing to measure the Frequency Response and Distortion of my amplifiers (4x150w RMS and 4x80W RMS, these amplifiers are car audio products, and the impedance of the speaker is 4 Ohm). To prevent from destroying my sound card,I made some research (here)(this article was not written by me, I just use it for reference), and I will prepare to make a circuit like this below.

And My sound card is Beringer UMC202HD, which supports 24bit 192KHz.

Should I choose 1/4W version for the resistors (2.2K Ohm and 22K Ohm)? Does it matter?

Actu


48364


Updated 2021.12.18

This circuit is the simplest one, and it comes from this article "Howto - Distortion Measurements with REW"

There is no votage divider in this simplest circuit.


48483


There is a fix divider in the circuit to protect the soundcard.

48484


UMC202HD Specifications: Line Max. input level +20dBu Mic in Impedance 3000Ohm

48492
 

Attachments

  • FULL CIRCUIT2.jpg
    FULL CIRCUIT2.jpg
    46.2 KB · Views: 12
Last edited:
Joined
Feb 8, 2018
Messages
279
Location
Italy
In principle it might be fine but there are many things that can be improved in practice.

1 if you want to measure the harmony distortion at maximum power the 50W load resistor is sufficient you need at least a 200W load.

2 the divider made up of 2.2K 22K resistors does not attenuate the level of the input signal so as to be compatible with the dynamics of the converter. you have to increase the resistance from 22K to 82K, as for the power 1 / 4W is enough.

3 The sound card used does not have sufficient bandwidth to measure that of an amplifier and the distortion also does not seem to be very low, so the measurements you want to make will be inaccurate.

Greetings Antonio
 
Joined
Feb 8, 2018
Messages
279
Location
Italy
I recommend that you also see this document that I am attaching to you.

Greetings Antonio
 

Attachments

  • AN1-MeasuringBox-Ver2-Rev3Eng(1).pdf
    826.9 KB · Views: 66

dotnet

Member
Joined
Mar 30, 2020
Messages
44
Given those resistor values (2.2k and 22k), the Behringer input impedance (27k) and its max. input level (2dBV, i.e. 1.26Vrms), you can put just under 28W through that 8Ω resistor before you overload the Behringer input. That‘s already more than the resistor will take without a largish heatsink. Unless you want to measure at full amplifier power your setup should work. The power dissipation over the 2.2k and 22k resistors is minuscule, 1/8W parts will be fine.
 

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
In principle it might be fine but there are many things that can be improved in practice.

Thank you for you help,Antonio.:T

I don't understand much about the first circuit, I just copy it from that article I mentioned. That article was not written by me. :)

About the principle of Voltage Divider, I still don't fully understand it.

My mother language is no English, so it will take me some time to understand these informations

Acturally, my sound card is Beringer UMC 202HD, support 24bit 192khz,is it O.K.?
 

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
I recommend that you also see this document that I am attaching to you.

Greetings Antonio

Thank you so much, Antonio,

This Box is professional, and flexible, but I too complex for me, I don't have enought knowledge to understand the circuit.:dizzy: But I will try to get some useful information from it.

I just want to make a simple circuit to measure the High level output of my amplifiers and Low level out put of my DSPs, I think maybe two simple and different circuits are good for me.:)
 

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
Given those resistor values (2.2k and 22k), the Behringer input impedance (27k) and its max. input level (2dBV, i.e. 1.26Vrms), you can put just under 28W through that 8Ω resistor before you overload the Behringer input. That‘s already more than the resistor will take without a largish heatsink. Unless you want to measure at full amplifier power your setup should work. The power dissipation over the 2.2k and 22k resistors is minuscule, 1/8W parts will be fine.

Thank you, Dotnet.

I don't understand much about the first circuit, I just copy it from that article I mentioned. That article was not written by me. I'm sorry for not expressing clearly.

Acturally, my sound card is Beringer UMC 202HD, support 24bit 192khz 。
 

dotnet

Member
Joined
Mar 30, 2020
Messages
44
Acturally, my sound card is Beringer UMC 202HD, support 24bit 192khz 。

The UMC 202HD has a max. input level of +20dBu at the line input (i.e. 7.75Vrms), so that won’t be overloaded by your circuit under any circumstances. Just don’t fry that 8Ω resistor, they need good heat-sinking from a few watts upward.
 

jschwender

Member
Joined
Mar 16, 2021
Messages
140
Location
GERMANY
More  
Front Speakers
Nubert digital pro
Other Speakers or Equipment
Philips dss940
The Behringer sound card has itself considerable distortion, you will not be able to distinguish between the distortion of the amplifier and the sound card. For this purpose you certainly need a high quality sound card with explicitly low distortion itself. Something like MOTU M2 or Audient ID4.
 

trobbins

Member
VIP Supporter
Joined
Nov 26, 2017
Messages
121
Location
Australia
jschwender, the OP can do single frequency tone testing of distortion and use REW's harmonic adjustment panel to null the soundcard distortion to very low levels. And if you looked at the article linked in the first post then a default UMC 202 had a loop distortion of 0.0045%, so not too shabby for starters.
 

jschwender

Member
Joined
Mar 16, 2021
Messages
140
Location
GERMANY
More  
Front Speakers
Nubert digital pro
Other Speakers or Equipment
Philips dss940
I have tested that Behringer device, and i can not confirm this data, at least if you have higher input level, the distortion can exceed 1%, no kidding! So yes, I do consider this device shabby.
I also tried to play around with the harmonic compensation, but that is very limited. Mainly because the angle resolution is not high enough, and so it prevents setting the required compensation. At the end i did not get the distortion to very low levels.
 

trobbins

Member
VIP Supporter
Joined
Nov 26, 2017
Messages
121
Location
Australia
For sure the distortion goes up at or near clipping, but it is just a tool, and if that means operating it at an appropriate input range to get low distortion then that's what one does. Wrt harmonic cancellation then maybe you weren't using an ASIO interface or there was some other aspect of the setup that caused your performance issue. Flatly dismissing such a tool without a little more credible assessment is not helpful to the OP imho.
 

jschwender

Member
Joined
Mar 16, 2021
Messages
140
Location
GERMANY
More  
Front Speakers
Nubert digital pro
Other Speakers or Equipment
Philips dss940
-42dBFS is for sure not that close to clipping. And this is the best value, the worst is 3% without clipping. Distortion in that devices clearly comes from the preamps, this is an inferior amplifier design. And it is not only the distortion, also the noise of that device is pretty bad compared to other devices. You may consider this acceptable – i am not. For me it is rather the Behringer specification that is not credible. This is just not truth in advertizing.
48386


And REW allows setting the compensating harmonics only to full degrees, and that is just not enough to fully compensate harmonics at very low levels. In worst case the remaining misalignment is 0.5 degrees, that means the harmonic can be compensated not to zero, but to sin(0.5) which is approximately 0.9%. May not be enough.
 

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
The Behringer sound card has itself considerable distortion, you will not be able to distinguish between the distortion of the amplifier and the sound card. For this purpose you certainly need a high quality sound card with explicitly low distortion itself. Something like MOTU M2 or Audient ID4.

Thank you jschwender.

Do you know the distortion level of beringer sound card? Is there any data for judgement?
 

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
-42dBFS is for sure not that close to clipping. And this is the best value, the worst is 3% without clipping. Distortion in that devices clearly comes from the preamps, this is an inferior amplifier design. And it is not only the distortion, also the noise of that device is pretty bad compared to other devices. You may consider this acceptable – i am not. For me it is rather the Behringer specification that is not credible. This is just not truth in advertizing. View attachment 48386

And REW allows setting the compensating harmonics only to full degrees, and that is just not enough to fully compensate harmonics at very low levels. In worst case the remaining misalignment is 0.5 degrees, that means the harmonic can be compensated not to zero, but to sin(0.5) which is approximately 0.9%. May not be enough.
Hi,jschwender

What type of Beringer sound card are you using ?

Could you tell me the circuit you are using to measure the distortion level of your Beringer sound card.

I'm interested in testing mine-UMC 202 HD.
 

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
The UMC 202HD has a max. input level of +20dBu at the line input (i.e. 7.75Vrms), so that won’t be overloaded by your circuit under any circumstances. Just don’t fry that 8Ω resistor, they need good heat-sinking from a few watts upward.
Cool, So I will buy some parts to make this circuit.
 

jschwender

Member
Joined
Mar 16, 2021
Messages
140
Location
GERMANY
More  
Front Speakers
Nubert digital pro
Other Speakers or Equipment
Philips dss940
Behringer specs give no data about distortion (guess why?), but they refer to Midas, and there you find a value of 0,01%. The distortion actually depends on gain setting and leveI, so generally making a statement about distortion is often more than just one number. I had a UMC202HD, and after my tests i immediately returned it, as it is not usable for my purpose. For testing i have a ASUS U7, which has other problems, but is has a very low distortion of 0.002% at max level and down to 0.0005% minimum. This is quite good for such a low cost sound card. But beware, this one has serious problems with channel swapping that makes it unusable in REW mesurement other than RTA. I have a hardware digital signal generator for test signals too, with distortion level of 0.001%. I also have a Steinberg UR22. I have tested many others too, on a bottom line, the truth is, the more expensive ones generally perform better. The Steinberg or even the Behringer is absolutely fine for acoustic measurements, as you rarely find speakers with distortion below 0.1%. But especially measuring amplifiers takes a tool that is better than the amplifier, and even cheap amplifiers can have something like 0.02%. This is not possible to measure with the Behringer as it's distortion is higher in any setting. If you want to measure a sound card, you need to have another, better sound card. Otherwise you cannot tell if the measurement result is from the sound card input or output. BTW the Steinberg for example has very robust inputs: even hooking it directly to a 250W amp does not harm it. The ASUS has no input protection and also has a weak level limit of 1.00Veff.
 

jschwender

Member
Joined
Mar 16, 2021
Messages
140
Location
GERMANY
More  
Front Speakers
Nubert digital pro
Other Speakers or Equipment
Philips dss940
One more thing: If the amplifier is a classD i would recomend to utilize a resistor with Ayrton-Perry-winding. The cheap ones have high induction values which cause trouble on class d amps.
 

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
Behringer specs give no data about distortion (guess why?), but they refer to Midas, and there you find a value of 0,01%. The distortion actually depends on gain setting and leveI, so generally making a statement about distortion is often more than just one number. I had a UMC202HD, and after my tests i immediately returned it, as it is not usable for my purpose. For testing i have a ASUS U7, which has other problems, but is has a very low distortion of 0.002% at max level and down to 0.0005% minimum. This is quite good for such a low cost sound card. But beware, this one has serious problems with channel swapping that makes it unusable in REW mesurement other than RTA. I have a hardware digital signal generator for test signals too, with distortion level of 0.001%. I also have a Steinberg UR22. I have tested many others too, on a bottom line, the truth is, the more expensive ones generally perform better. The Steinberg or even the Behringer is absolutely fine for acoustic measurements, as you rarely find speakers with distortion below 0.1%. But especially measuring amplifiers takes a tool that is better than the amplifier, and even cheap amplifiers can have something like 0.02%. This is not possible to measure with the Behringer as it's distortion is higher in any setting. If you want to measure a sound card, you need to have another, better sound card. Otherwise you cannot tell if the measurement result is from the sound card input or output. BTW the Steinberg for example has very robust inputs: even hooking it directly to a 250W amp does not harm it. The ASUS has no input protection and also has a weak level limit of 1.00Veff.
Hi jschwender,

Thank you for your experience.

Originally , my UMC202HD is for speaker impedance measurement with REW.

A few days ago, I got a DSP with amplifier (4 x 80W RMS) from a factory for testing purpose, to make sure that DSP is good for me. But after listening to it, I decided to measurement the Frequency Response of the DSP, and I got to know that using a rig (circuit) and REW I can measure the Frequency response and distortion, so I want to make a circuit to measure it.

I don't want to make things too complex, if my UMC202HD is not good enough for distortion measurement, maybe it's still good for frequency reponse measurement. and I think the circuit should be the same,right?

BTW, is this diagram below correct for sound card distortion compensation?

48408
 

icbcodc

New Member
Thread Starter
Joined
Apr 29, 2020
Messages
187
One more thing: If the amplifier is a classD i would recomend to utilize a resistor with Ayrton-Perry-winding. The cheap ones have high induction values which cause trouble on class d amps.
Actually, the amplifier is class A/B. I also have another class A/B amplifier (4 x 150W rms).
 
Joined
Feb 8, 2018
Messages
279
Location
Italy
Given those resistor values (2.2k and 22k), the Behringer input impedance (27k) and its max. input level (2dBV, i.e. 1.26Vrms), you can put just under 28W through that 8Ω resistor before you overload the Behringer input. That‘s already more than the resistor will take without a largish heatsink. Unless you want to measure at full amplifier power your setup should work. The power dissipation over the 2.2k and 22k resistors is minuscule, 1/8W parts will be fine.
For the calculation of the divider I considered the measurement of the distortion at the level of maximum output power. In our case the maximum power was 150W which on 8 ohm corresponds to a voltage of about 35Veff while the maximum level accepted by a sound card has been set at 1Veff therefore the 2.2Kohm and 82Kohm partition guarantees that at the maximum power the input voltage the sound card is less than the expected maximum level. with 35 Veff the input level to the sound card is 0.91Veff .
 
Joined
Feb 8, 2018
Messages
279
Location
Italy
Attention
the scheme proposed by our friend icbcodc works only with amplifiers that have the output referred to ground. use this circuit under penalty of breaking the power amp.

Greetings Antonio
 

trobbins

Member
VIP Supporter
Joined
Nov 26, 2017
Messages
121
Location
Australia
jschwender, there are two notable links to tests on UMC20x devices that indicate that range can provide better distortion performance.
https://www.audiosciencereview.com/...ringer-umc204-hd-audio-interface-review.9856/
https://www.audiosciencereview.com/...et-dac-review-behringer-umc204hd.1658/page-24

Maybe there are version or revision differences, but it seems strange for such a device to measure such large distortions without it being a setup issue.

Wrt harmonic cancellation, your results may highlight two aspects - the first being that very large levels of harmonics are not easy to suppress, and two that it may be related to selecting the 'Lock frequency to rta fft' which you may not have done as the frequency is shown as 1000.00kHz rather than being say 1000.49Hz for a 128kHz FFT.
 

jschwender

Member
Joined
Mar 16, 2021
Messages
140
Location
GERMANY
More  
Front Speakers
Nubert digital pro
Other Speakers or Equipment
Philips dss940
Thanks for the hint. I just can't confirm that. Also this report states: »This is with 0.54 volt input. Anything beyond that hammers the input stage causing distortion.« I also don't know version differences, i never tested this specific version. I am pretty sure there is no setup issue: i can compare that to many other devices that i tested so far. This Behringer soundcard was the second worst i ever had in my hands. That is my provable insight.
And yes i did use freq. locking in the test. The issue here is really that the quantization by level in full dB and angle in full degrees allows only a fairly large raster of discrete values for compensation. Maybe @John Mulcahy can add a digit on both, that would improve it, but he is probably not happy with the downside that operating in such small steps with the up and down buttons is less comfortable?
 

trobbins

Member
VIP Supporter
Joined
Nov 26, 2017
Messages
121
Location
Australia
I have been able to trim harmonics down to very low noise floor levels which is why I don't appreciate how your testing didn't, but I started with 'nominal' harmonic levels from the soundcard, but I also did the same technique when adding in a buffer amp, so that the whole loop including buffer was trimmed to very low harmonic levels.

Wrt to practical input level for UMC20x, even at 0.5V input level, the linked report indicates there is still a very large dynamic range available for noise and harmonic testing.
 
Top Bottom