Hi everyone,
I am new to the forum and new to the world of audio. I decided to undertake my first audio project and the objective is to measure the THD+N of an audio amplifier. The set up goes like this:
A signal will be input from a laptop and pass through a DAC. From there the signal goes to a A/V receiver where it will be amplified and sent to a speaker. From the back of a speaker a cable tap will be taken which will lead to an ADC. From the ADC a USB cable will link back to the laptop to measure the THD+N.
I have chosen ARTA software but I am open to other free softwares. The problem is that I really do not know how to use ARTA software for this set up. I have read the user manual and it seems to focus on using a microphone which I do not want to use. After running a few test the THD+N measurements are extremely high (40-95%) which is obviously wrong. I am inputing a standard 1 kHz sine wave. I realize this set up will not provide the most accurate THD+N measurements but I would like to get to reasonable numbers if possible. Could anybody explain how to correct this problem or why these high THD+N measurements are being recorded.
I appreciate an insight or help provided,
Rob
I am new to the forum and new to the world of audio. I decided to undertake my first audio project and the objective is to measure the THD+N of an audio amplifier. The set up goes like this:
A signal will be input from a laptop and pass through a DAC. From there the signal goes to a A/V receiver where it will be amplified and sent to a speaker. From the back of a speaker a cable tap will be taken which will lead to an ADC. From the ADC a USB cable will link back to the laptop to measure the THD+N.
I have chosen ARTA software but I am open to other free softwares. The problem is that I really do not know how to use ARTA software for this set up. I have read the user manual and it seems to focus on using a microphone which I do not want to use. After running a few test the THD+N measurements are extremely high (40-95%) which is obviously wrong. I am inputing a standard 1 kHz sine wave. I realize this set up will not provide the most accurate THD+N measurements but I would like to get to reasonable numbers if possible. Could anybody explain how to correct this problem or why these high THD+N measurements are being recorded.
I appreciate an insight or help provided,
Rob
Warning: This is only a guess! You are probably overloading the input to the A2D. Is there any way to display the samples that the A2D is seeing?
djoffe, thanks for the advice. To view the samples the A2D is seeing I assume I would need to acquire the software CD that came with it. I am using the E-MU 0404 USB Audio/MIDI Interface borrowed from a buddy. If I am overloading the A2D, what type of adjustments would you recommend i make?
There are many good and free software's available that you can use here are some links that might help you.
Some of these as well as others have scope functions as well to help determine if you are clipping your input.
I had used them just recently while measuring one of my PPA10 Pyle 10" drivers and found that I was clipping my mic input and after some gain adjustments everything was fine.
Your input to your sound card will only handle a certain amount of voltage (I don't know what yours is rated for) so it is advisable to use a dummy load resistor on the output of the amp (optional) and a voltage divider on the input of your sound card in order to protect it and to keep from clipping the input stage to the A/D from the amplifiers voltage swings over say 15 volts peak or so,or less.
The Wallin jig is a great example of such a device it is basically dummy load and a voltage divider and takes just a few resistors is all.
I use Wave spectra alot and is a great program and I just recently have been using Virtual Multi instrument as well and I can do nearly every thing in those programs inculding realtime THD measurements,FFT and scope display once I had figured out how to use them.
Sometimes I will use Winlsd's signal generator or some other software like audacity or an external signal to generate the test signal as it does make things a little easier to test.
Just make sure that you set them up for the highest sample rate and 24bit (if applicable) in order to get your most accurate measurments and I usually use 32bit floating point for the processing as well.
HolmImpluse as a real super program and is what I have been mainly using as well.
It does THD but it is displayed in db's so you can use this online calculator to do the conversion,
THD to dB - convert percent to decibels dB and dB decibels to percentage percent voltage % vs per cent converter THD+N distortions calculation signal distortion factor attenuation in dB to distortion factor k in percent decibel damping - sengpielaudi
Usually you will have to do a loopback test inorder to calibrate the software to your hardware after that you just have to make sure that your signal is not being clipped (I am sure that you already know this).
Getting every thing calibrated is a pain but not hard and after lots of practice it will become second nature when you get used to your setup.
If you have a good mixer available to you this will make every thing a whole lot easier as far as setup goes.
I have everything running in and out of my Mackie 32-8 mixing board and I have a few channels setup just for measuring and rarely do I have to do any recalibration's.
But I am limited to the specs of the board in which isn't a problem unless I want to start doing measurement below like .001%THD or so or what ever the board happens to be performing at.
But it is not a big deal for me at this time.
Although the best results will be straight in and out of the sound card of course.
I also use the setup to do THD measurements of my speakers as well and this is very cool as the software has not cost me a single dime !! he,he
These should help you to get started,
http://www.diyaudio.com/forums/soft...response-entire-sound-system.html#post2916540
http://www.diyaudio.com/forums/soft...lti-instrument-v3-2-software.html#post2915649
http://www.diyaudio.com/forums/software-tools/190658-best-free-audio-software.html#post2601840
http://www.diyaudio.com/forums/soft...-recommendations-newbie-here.html#post2578937
some more info,
http://www.diyaudio.com/forums/soft...-up-correctly-amp-measurment.html#post2648317
A whole bunch more can be found here,
http://www.dxzone.com/catalog/Software/Spectrum_analyzers/
Good luck !!!!
jer 🙂
Some of these as well as others have scope functions as well to help determine if you are clipping your input.
I had used them just recently while measuring one of my PPA10 Pyle 10" drivers and found that I was clipping my mic input and after some gain adjustments everything was fine.
Your input to your sound card will only handle a certain amount of voltage (I don't know what yours is rated for) so it is advisable to use a dummy load resistor on the output of the amp (optional) and a voltage divider on the input of your sound card in order to protect it and to keep from clipping the input stage to the A/D from the amplifiers voltage swings over say 15 volts peak or so,or less.
The Wallin jig is a great example of such a device it is basically dummy load and a voltage divider and takes just a few resistors is all.
I use Wave spectra alot and is a great program and I just recently have been using Virtual Multi instrument as well and I can do nearly every thing in those programs inculding realtime THD measurements,FFT and scope display once I had figured out how to use them.
Sometimes I will use Winlsd's signal generator or some other software like audacity or an external signal to generate the test signal as it does make things a little easier to test.
Just make sure that you set them up for the highest sample rate and 24bit (if applicable) in order to get your most accurate measurments and I usually use 32bit floating point for the processing as well.
HolmImpluse as a real super program and is what I have been mainly using as well.
It does THD but it is displayed in db's so you can use this online calculator to do the conversion,
THD to dB - convert percent to decibels dB and dB decibels to percentage percent voltage % vs per cent converter THD+N distortions calculation signal distortion factor attenuation in dB to distortion factor k in percent decibel damping - sengpielaudi
Usually you will have to do a loopback test inorder to calibrate the software to your hardware after that you just have to make sure that your signal is not being clipped (I am sure that you already know this).
Getting every thing calibrated is a pain but not hard and after lots of practice it will become second nature when you get used to your setup.
If you have a good mixer available to you this will make every thing a whole lot easier as far as setup goes.
I have everything running in and out of my Mackie 32-8 mixing board and I have a few channels setup just for measuring and rarely do I have to do any recalibration's.
But I am limited to the specs of the board in which isn't a problem unless I want to start doing measurement below like .001%THD or so or what ever the board happens to be performing at.
But it is not a big deal for me at this time.
Although the best results will be straight in and out of the sound card of course.
I also use the setup to do THD measurements of my speakers as well and this is very cool as the software has not cost me a single dime !! he,he
These should help you to get started,
http://www.diyaudio.com/forums/soft...response-entire-sound-system.html#post2916540
http://www.diyaudio.com/forums/soft...lti-instrument-v3-2-software.html#post2915649
http://www.diyaudio.com/forums/software-tools/190658-best-free-audio-software.html#post2601840
http://www.diyaudio.com/forums/soft...-recommendations-newbie-here.html#post2578937
some more info,
http://www.diyaudio.com/forums/soft...-up-correctly-amp-measurment.html#post2648317
A whole bunch more can be found here,
http://www.dxzone.com/catalog/Software/Spectrum_analyzers/
Good luck !!!!
jer 🙂
Last edited:
The EMU 0404 works very well with ARTA (I use this combination at work). It also works well for distortion measurements with the free 'Visual Analyser' from Sillanum Soft.
You need to take care with the input attenuator to your sound card (whatever one you use) to ensure that you get the appropriate voltage level at the sound card input for the signal level expected at the amplifier output.
You need to take care with the input attenuator to your sound card (whatever one you use) to ensure that you get the appropriate voltage level at the sound card input for the signal level expected at the amplifier output.
Ok I am currenly running tests with ARTA software and the THD+N measurements are decent (around 0.1 %) compared to my previous results. My question is for anybody who may have experience with ARTA software or this type of project , what signal generator settings will usually produce the lowest THD+N measurements for an audio amplifier? What settings usually highlight the true performance/quality of an audio amplifier? Currently I am inputting a 1kHz sine wave with a dB level of -3 and Dither level of 20-bit. My sampling frequency is 96kHz and the FFT is 16,384. I have the Amplifier volume set at about -65 dB.
Please refer to my original post for details of my setup.
Thanks to all who have posted so far, and thanks in advance for anyone's input.
Rob
Please refer to my original post for details of my setup.
Thanks to all who have posted so far, and thanks in advance for anyone's input.
Rob
- Status
- Not open for further replies.