I have been trying in vain to find an audio signal generator that will generate the required 60 Hz and 7 kHz tones to measure intermodulation distortion on audio amps and preamps. I'm have a Potomac Instruments AA-51 Audio Analyzer that measures IMD among other things, but it does not have an audio generator built in to generate the required tones. Their generator is hard to find and can get expensive. As a last resort, I have seen free audio files online with the required tones. However, I would prefer to have an audio signal generator I can hook up.
Any suggestions and/or sources for an IMD signal generator?
Any suggestions and/or sources for an IMD signal generator?
I have always used a sine/square and triangle generator. As long as the sine is clean, that is all that matters.
I have one I no longer use, made from a kit in the 80s and it is still good.
There are plenty available from many sources.
FeelTech FY6800 60MHz Function Arbitrary Waveform Pulse DDS Signal Generator | eBay
is excellent but a tad over the top.
DDS Function Signal Generator Module Sine + Triangle + Square Wave + USB Cable 8256962105167 | eBay
wihttps://www.ebay.co.uk/itm/Digital-DDS-Function-Signal-Generator-Module-Sine-Triangle-Square-sawtooth-Wave/382878750434?hash=item59255a16e2&enc=AQADAAADEFjVrDbVsZ8oH%2F8PNHtt9VX4%2Fw7FZcmMuqsX8uaFEduVHurcFE3sf%2FhS4wfAkY2aDEg27d3rgU%2FJQrH17oRHtPXPQqPzp5s0jfzitP6sB3yKoipGkhhLONfp572D%2B%2Bti%2B%2BJO0yZsiBsWKfiBa2yyXOkbF9Gw%2FQIqOz%2B6HIn2GrHh7jriBURaJmQQg9TwkEKvXACMNE0pkk5oUZGDp3BUAc8FZv4BbLD79z2kNX2RUp3oFzcWYHEhWCPsIOcWF11sBfcEV68adkHMSVt39HwfEOVYHQXk%2BdniK%2BZ7ALYFJ%2BuxVva90W9qxgeMwVORFFCffeUEXx2TRVxbDTBXszCHDHggUGlUh0KS4Lq%2B2QbSYEXPK15BOKI68OT9ojaSyh0960pM1yLVs1SuW8lxgFG0PuV4IU1Qr%2BIYjqq6d5TxFlXwzHjNdt90MrVhogvB%2Fr3qdu83vhoKOWe1gYWfXSuLZ47F8F8OlW6xVsYYVQebell6nTPErj1z0apw5ppcZEYJAnDFgW96Cndp5w4ctuMsaztKQIaek0NpRLjIsPPqC%2BxmSQSAeDG6TXvfCT9y%2BXUdwQfDhIJ0epYT26HWMYPa%2F%2FpNOA7Jam%2FZa%2FK3ncw3%2FKQrqIzPucCH6FdFlIgInUk1cvERR0w9AL9Mil0kV327apZ9A2%2FMJY3fQ4O9zGhcpe78kZz3I81agWExTEvyWTMsslSC%2F6GjcXTYYCj8CAEmYrNDcTX4DsnYZuq%2FXz5pRlcOxwLiQ8%2FTniexZgixryzUk2u%2B31Tyu%2FeFqNpx2SRbS3wBS4ZBvPQmzMkdUbXrO1pmmiksk23vbpJAvwLI6AgJM3vz%2B7cptfYzEKsaqTL0M4ffvDEUZf5UvRJ%2FYn0yWbsoYMrL8flWds6yfORgUrKy2s60NOIkx8yGq4Xx9mFpZqTwSwdOO6lnNf1YH4PoFu72eWOJEnNbvJb6BBnO0%2BYhVKjsx9E5BUtbnoTT9XmYTRhjVdyDC5fFctyPsIXqwXHVdEiN%2F3KUtUta8csWjfveTA8V0zCsRQfUHDKJT1uFAgU%3D&checksum=3828787504344499fad8518848fbbdfbdcf1d5d971c4
ll do what is needed as will
I have one I no longer use, made from a kit in the 80s and it is still good.
There are plenty available from many sources.
FeelTech FY6800 60MHz Function Arbitrary Waveform Pulse DDS Signal Generator | eBay
is excellent but a tad over the top.
DDS Function Signal Generator Module Sine + Triangle + Square Wave + USB Cable 8256962105167 | eBay
wihttps://www.ebay.co.uk/itm/Digital-DDS-Function-Signal-Generator-Module-Sine-Triangle-Square-sawtooth-Wave/382878750434?hash=item59255a16e2&enc=AQADAAADEFjVrDbVsZ8oH%2F8PNHtt9VX4%2Fw7FZcmMuqsX8uaFEduVHurcFE3sf%2FhS4wfAkY2aDEg27d3rgU%2FJQrH17oRHtPXPQqPzp5s0jfzitP6sB3yKoipGkhhLONfp572D%2B%2Bti%2B%2BJO0yZsiBsWKfiBa2yyXOkbF9Gw%2FQIqOz%2B6HIn2GrHh7jriBURaJmQQg9TwkEKvXACMNE0pkk5oUZGDp3BUAc8FZv4BbLD79z2kNX2RUp3oFzcWYHEhWCPsIOcWF11sBfcEV68adkHMSVt39HwfEOVYHQXk%2BdniK%2BZ7ALYFJ%2BuxVva90W9qxgeMwVORFFCffeUEXx2TRVxbDTBXszCHDHggUGlUh0KS4Lq%2B2QbSYEXPK15BOKI68OT9ojaSyh0960pM1yLVs1SuW8lxgFG0PuV4IU1Qr%2BIYjqq6d5TxFlXwzHjNdt90MrVhogvB%2Fr3qdu83vhoKOWe1gYWfXSuLZ47F8F8OlW6xVsYYVQebell6nTPErj1z0apw5ppcZEYJAnDFgW96Cndp5w4ctuMsaztKQIaek0NpRLjIsPPqC%2BxmSQSAeDG6TXvfCT9y%2BXUdwQfDhIJ0epYT26HWMYPa%2F%2FpNOA7Jam%2FZa%2FK3ncw3%2FKQrqIzPucCH6FdFlIgInUk1cvERR0w9AL9Mil0kV327apZ9A2%2FMJY3fQ4O9zGhcpe78kZz3I81agWExTEvyWTMsslSC%2F6GjcXTYYCj8CAEmYrNDcTX4DsnYZuq%2FXz5pRlcOxwLiQ8%2FTniexZgixryzUk2u%2B31Tyu%2FeFqNpx2SRbS3wBS4ZBvPQmzMkdUbXrO1pmmiksk23vbpJAvwLI6AgJM3vz%2B7cptfYzEKsaqTL0M4ffvDEUZf5UvRJ%2FYn0yWbsoYMrL8flWds6yfORgUrKy2s60NOIkx8yGq4Xx9mFpZqTwSwdOO6lnNf1YH4PoFu72eWOJEnNbvJb6BBnO0%2BYhVKjsx9E5BUtbnoTT9XmYTRhjVdyDC5fFctyPsIXqwXHVdEiN%2F3KUtUta8csWjfveTA8V0zCsRQfUHDKJT1uFAgU%3D&checksum=3828787504344499fad8518848fbbdfbdcf1d5d971c4
ll do what is needed as will
I have frequency generators that can output a single frequency in all types of waveforms, but the SMPTE method to measure IMD requires two tones (60 Hz and 7 Khz) at the same time in a 4:1 ratio that fed into the preamp/amp and then into the IMD analyzer to determine IMD.
Looking at a wave form on a scope gives a pretty good idea of whether the preamp/amp is putting out a good, clean signal, but it doesn't give the precise measurements of THD or IMD. I work on a lot of high-end older tube gear and I'd like to be able to tell customers how well their equipment is performing as well as test my own stuff.
Looking at a wave form on a scope gives a pretty good idea of whether the preamp/amp is putting out a good, clean signal, but it doesn't give the precise measurements of THD or IMD. I work on a lot of high-end older tube gear and I'd like to be able to tell customers how well their equipment is performing as well as test my own stuff.
Hello,
You can simply use 2 signal generators and sum them with resistors networks.
I have done this in this way here.
You can also use an PC soundcard with audio editing software (like Audacity or others)
to create you own audio files with any tones combination. Then you play it with the soundcard
to get your test signal.
Regards.
Frex
You can simply use 2 signal generators and sum them with resistors networks.
I have done this in this way here.
You can also use an PC soundcard with audio editing software (like Audacity or others)
to create you own audio files with any tones combination. Then you play it with the soundcard
to get your test signal.
Regards.
Frex
That's a great option. However, I was hoping that someone might know of a company that made a not-too-expensive signal generator that was made specifically to test IMD using the SMPTE method by generating 60 Hz and 7 KHz at the same time in a 4:1 ratio.
My Heath IMD analyzer gets the 60 hz tone from the line frequency IIRC. You might take a look at the schematic for one of these. The high frequency tone is just an LC oscillator.
I have one of the FeelTech 60 MHz two channel DDS, but I'm not sure if it goes down to audio. I use it for RF, but if it goes low enough, it could be a good option. Make sure to check the amplitude with a scope or voltmeter.
Win W5JAG
I have one of the FeelTech 60 MHz two channel DDS, but I'm not sure if it goes down to audio. I use it for RF, but if it goes low enough, it could be a good option. Make sure to check the amplitude with a scope or voltmeter.
Win W5JAG
I'll check it out. I also found this on eBay: XR2206 Function Signal Generator DIY Kit Sine Output 1HZ-1MHZ + acrylic case USA | eBay
If I used two, I could set one to 60 Hz and the other to 7KHz and adjust the power ratio to the SMPTE spec and it might be an inexpensive way to generate the needed tones.
If I used two, I could set one to 60 Hz and the other to 7KHz and adjust the power ratio to the SMPTE spec and it might be an inexpensive way to generate the needed tones.
I am not sure a generator with claimed internal distortion of < 1% will allow you to measure distortions of modern amps.
IMO any internal 24bit soundcard (Intel HDA) will give much better results (THD <0.01%). If you do not feel like running the tone-generation software (it is trivial), you can have the tone pre-generated in a file and just play it with any audio player. IMO way easier and much better quality than this generator with sine shaper.
IMO any internal 24bit soundcard (Intel HDA) will give much better results (THD <0.01%). If you do not feel like running the tone-generation software (it is trivial), you can have the tone pre-generated in a file and just play it with any audio player. IMO way easier and much better quality than this generator with sine shaper.
REW's generator offers a range of IMD signals if a software solution is an option:
The signals can also be saved to WAV files.
The signals can also be saved to WAV files.
I have one of those XR2206 kit generators in my travel tool kit that I carry to my second house, and I think it would be suitable as a tone generator for a two tone IMD test. It would not be acceptable for THD testing of a hi fi amp.
Pros: cheap and very compact - can't homebrew an oscillator for the cost of the kit.
Cons: It takes some twiddling to get the frequency you want, and you still have to have a counter, scope or voltmeter to set the frequency and amplitude. The range is also limited.
For $30 or $35 dollars, I think you could get a two channel, hobby grade DDS generator, that would go out to 1 or 2 MHz, with all the cables, nice bench top case, a display for setting frequency, amplitude, etc., although I would still use a scope or voltmeter to set the amplitude initially.
My 60 MHz FeelTech, was only $85 or so.
Win W5JAG
Pros: cheap and very compact - can't homebrew an oscillator for the cost of the kit.
Cons: It takes some twiddling to get the frequency you want, and you still have to have a counter, scope or voltmeter to set the frequency and amplitude. The range is also limited.
For $30 or $35 dollars, I think you could get a two channel, hobby grade DDS generator, that would go out to 1 or 2 MHz, with all the cables, nice bench top case, a display for setting frequency, amplitude, etc., although I would still use a scope or voltmeter to set the amplitude initially.
My 60 MHz FeelTech, was only $85 or so.
Win W5JAG
The REW looks like it will do the trick. They even have a Mac version which is what I need. Thanks for the suggestion!
You'll need the beta version for the latest signal generator, the beta version is also needed if you are on Mojave and want to capture audio.
I can't understand how xr2206 oscillator (which has very bad distortion) can help with IMD if if can't help with THD?
Last edited:
My IMD analyzer works similar to an AM receiver, if I understand it correctly.
The high frequency tone is the "carrier" and the "passband" is the width of the analyzers high pass filter response. The "carrier" usually 6-7 Khz can be chosen anywhere in this "passband". The modulating frequency on the carrier, usually 50-60 Hz can be any frequency within the low pass filter response of the analyzer. The high frequency "carrier" is detected, and the low pass filter makes sure that the ac vtvm only responds to the created modulation products within the low pass filter response. I don't see that it matters much if the 60 Hz signal is dirty - it just makes another modulation product which is what you are trying to make and measure. Harmonics of the high frequency will be outside of the filter response.
My THD meter just notches the oscillator frequency, and detects what is left over, so a dirty oscillator that is notched, still leaves all of its harmonics to be detected and measured. You need to start with a clean oscillator.
This is how I understand it, anyway, and why I think the XR2206 would be OK for (edit: SMPTE ) two tone IMD, but not for THD. It will not hurt my feelings for anyone to correct me.
Win W5JAG
The high frequency tone is the "carrier" and the "passband" is the width of the analyzers high pass filter response. The "carrier" usually 6-7 Khz can be chosen anywhere in this "passband". The modulating frequency on the carrier, usually 50-60 Hz can be any frequency within the low pass filter response of the analyzer. The high frequency "carrier" is detected, and the low pass filter makes sure that the ac vtvm only responds to the created modulation products within the low pass filter response. I don't see that it matters much if the 60 Hz signal is dirty - it just makes another modulation product which is what you are trying to make and measure. Harmonics of the high frequency will be outside of the filter response.
My THD meter just notches the oscillator frequency, and detects what is left over, so a dirty oscillator that is notched, still leaves all of its harmonics to be detected and measured. You need to start with a clean oscillator.
This is how I understand it, anyway, and why I think the XR2206 would be OK for (edit: SMPTE ) two tone IMD, but not for THD. It will not hurt my feelings for anyone to correct me.
Win W5JAG
Last edited:
If xr2206 has not very long garmonics tail then you are right.
But, if you are watching ebay - there are better modules. Look at this:
1)
Assembled-Low-Distortion-Audio-Range-Oscillator-1KHz-Sine-Wave-Signal-Generator
- it is assembled, but it was and not assembled ones, just for $5 or something like that.
And one more (not so cheap):
2)
Low-distortion-0-01-sine-wave-generator-oscillator
I tried first one - it really has 0.006% THD. And works good on 50 Hz (several second start). And (as it is RC-type) - has very stable amplitude when you change frequency (but you don't need that).
3) Victors oscillator - is the best, but it has problem with low frequencies - at 50 Hz it has about 45 second start.
But, if you are watching ebay - there are better modules. Look at this:
1)
Assembled-Low-Distortion-Audio-Range-Oscillator-1KHz-Sine-Wave-Signal-Generator
- it is assembled, but it was and not assembled ones, just for $5 or something like that.
And one more (not so cheap):
2)
Low-distortion-0-01-sine-wave-generator-oscillator
I tried first one - it really has 0.006% THD. And works good on 50 Hz (several second start). And (as it is RC-type) - has very stable amplitude when you change frequency (but you don't need that).
3) Victors oscillator - is the best, but it has problem with low frequencies - at 50 Hz it has about 45 second start.
Last edited:
Erno Borbely did an IM meter construction article years ago and had this to say about the oscillators he included-
"producing 5V RMS output with approximately 0.01% THD. Because the test is independent of the oscillator THD, even more could be tolerated. However, low noise and very good frequency stability are of primary importance."
"producing 5V RMS output with approximately 0.01% THD. Because the test is independent of the oscillator THD, even more could be tolerated. However, low noise and very good frequency stability are of primary importance."
SMPTE-class IMD tests are not very sensitive to the quality of the 60 and 6k sources. That was part of the appeal, when this was started, about 80 years ago.
SMPTE IMD tests for a gross form of distortion that ANY modern audio amp will have very little of. It is particularly sensitive to undersize output transformers which may do either 60 or 6k "OK", but not both at the same time.
"Serious" amp testers moved decades ago to more stressful frequency pairs like 18k+19k(?).
SMPTE IMD tests for a gross form of distortion that ANY modern audio amp will have very little of. It is particularly sensitive to undersize output transformers which may do either 60 or 6k "OK", but not both at the same time.
"Serious" amp testers moved decades ago to more stressful frequency pairs like 18k+19k(?).
I don't understand how this is more stressful than SMPTE.
With 18 and 19 Khz tones:
first set of mix products: 1 and 37 KHz
3rd order IMD products: 17 and 20
5th order IMD products: 16 and 21
7th order IMD products: 15 and 22
9th order IMD products: 14 and 23
If my arithmetic is correct. So if I extend this out sooner or later there will be a tone every 1 KHz in the audio band, likely with severely decreasing amplitude as IMD order goes higher.
Obviously that would sound terrible. But there isn't much 18 or 19 KHz energy to start with in most real world situations, so how can it be stressful? Or even realistic? Or is there a bunch of energy at these frequencies?
Off the top of my head, this test would make an amplifier look better than it really is. My limited experience is that SMPTE is real hard for single ended tube amps, pretty easy for push pull.
If the purpose is to generate IMD, it seems like a clean oscillator is a liability.
Win W5JAG
- Status
- Not open for further replies.