In the process of getting setup to do communications receiver testing, I became interested in being able to measure IMD at levels better than .01% in traditional amplifiers. I'm primarily looking at the performance of external DSP units, preamps, powered communications speakers, etc.
I have four RF spectrum analyzers, one of which goes down to 100hz, an HP 8566B. I have a couple of HP synthesized signal generators both for audio and RF, one of which is at least good to -75db. I have a couple of sound cards and my primary is a Steinberg UR242 external USB type. I have who knows how many oscilloscopes, one of which is a Digital Signal Analyzer that does a decent job of spectrum analysis and will calculate THD to about -70db, maybe better.
I've been looking at older HP equipment like the HP 339a and the HP 8903B. The 339a was of interest as it is often used for receive MDS and RMDR testing but in those cases, the dB calibrated RMS meter is what is really used and I have both HP 3400a and Ballantine 322-7 AC RMS meters. An analog meter is easier to visually average compared to a digital meter as in the 8903B but since I have two A/C RMS meters, the 8903B is now a better choice if I elect to go that route.
So given that background and having the UR242, should I go the PC software route, get an 8903B, look at something else? Get a better USB sound adapter and software? I would like to keep the investment sub $500 but I'm flexible. I use Audacity, matlab, soundforge, SpectrumLab and other software codes today. I've also written a lot of Visual basic code to interrogate my spectrum analyzers over GPIB to make more accurate THD measurements on RF though I've tested it on audio and it maps well against other methods. One of the problems I continue to have is both attenuation and impedance matching. I've found on the UR242 I get widely varying results in recording in both hi and low impedance settings even paying careful attention to the input levels. I forgot to mention that I have a number of both SHARC and BLACKFIN DSP units I can also use to generate pretty clean signals.
Having written all this, I've come to the conclusion that I guess what I need is a better recording setup and analysis software. Thoughts?
Thanks,
Jerry
I have four RF spectrum analyzers, one of which goes down to 100hz, an HP 8566B. I have a couple of HP synthesized signal generators both for audio and RF, one of which is at least good to -75db. I have a couple of sound cards and my primary is a Steinberg UR242 external USB type. I have who knows how many oscilloscopes, one of which is a Digital Signal Analyzer that does a decent job of spectrum analysis and will calculate THD to about -70db, maybe better.
I've been looking at older HP equipment like the HP 339a and the HP 8903B. The 339a was of interest as it is often used for receive MDS and RMDR testing but in those cases, the dB calibrated RMS meter is what is really used and I have both HP 3400a and Ballantine 322-7 AC RMS meters. An analog meter is easier to visually average compared to a digital meter as in the 8903B but since I have two A/C RMS meters, the 8903B is now a better choice if I elect to go that route.
So given that background and having the UR242, should I go the PC software route, get an 8903B, look at something else? Get a better USB sound adapter and software? I would like to keep the investment sub $500 but I'm flexible. I use Audacity, matlab, soundforge, SpectrumLab and other software codes today. I've also written a lot of Visual basic code to interrogate my spectrum analyzers over GPIB to make more accurate THD measurements on RF though I've tested it on audio and it maps well against other methods. One of the problems I continue to have is both attenuation and impedance matching. I've found on the UR242 I get widely varying results in recording in both hi and low impedance settings even paying careful attention to the input levels. I forgot to mention that I have a number of both SHARC and BLACKFIN DSP units I can also use to generate pretty clean signals.
Having written all this, I've come to the conclusion that I guess what I need is a better recording setup and analysis software. Thoughts?
Thanks,
Jerry
Not to beat my own drum, but something like his would work: The L|A Autoranger | Linear Audio NL
And from your equipment list, the easiest and highest resolution would be the soundcard and software like ARTA. None of the other options come close, except of course an AP at $10k+. (A vintage AP S1 is inferior to your soundcard+ARTA. I know, I tried both).
Jan
So if I wanted to test an amp, a KSA-50S specifically, I would need something to drive the amp into that is switchable from like 8ohms down to 2ohms and then attenuate the signal enough for a soundcard. At least that is how I am thinking. I can also just use a 4ohm load of some kind and just look at the signal with one of my scopes or a Tek Digital Signal Analyzer which is good to around -70db of THD. But in thinking about all this I figured I would get setup for both receiver testing as well as amp/preamp. Preamps are easy, of course.
This is clearly a different issue than receiver testing as I started, but was thinking today I would like to see what is going on with that 50S amp. The amp is used in a bi-amp setup with a KSA-80B. It is driving the mid/highs of some BMW 802s. A while back my son was playing some rap song with ridiculous base and mid and the amp cut out. I then found when opening it that one leg of the high voltage supply came unsoldered. There must have been a high resistance (couple of ohms) on that leg connection that overheated and unsoldered. It looks like the bias supply was then overloaded and I replaced the obviously failed components. I would have to go back and look at the schematic to figure out why the bias supply tried to provide all that current but it made sense at the time, almost as if the bias supply was trying to provide all the current once the other leg opened. The issue is that every now and then, that channel gets a little ratty sounding after playing for a while. It's a great amp and setup overall, especially coupled with the 80B, an active crossover, Krell preamp, krell phono stage and a TNT-MK IV turntable.
So to summarize, I can use the sound card with various suggested codes for high level testing, it's a better option than any of my spectrum analyzers, but I still need some type of load for amps as well as an attenuator for the sound card. Unless I can use the sound card with some type of high-impedance probe?
thanks for all the input.
Jerry
This is clearly a different issue than receiver testing as I started, but was thinking today I would like to see what is going on with that 50S amp. The amp is used in a bi-amp setup with a KSA-80B. It is driving the mid/highs of some BMW 802s. A while back my son was playing some rap song with ridiculous base and mid and the amp cut out. I then found when opening it that one leg of the high voltage supply came unsoldered. There must have been a high resistance (couple of ohms) on that leg connection that overheated and unsoldered. It looks like the bias supply was then overloaded and I replaced the obviously failed components. I would have to go back and look at the schematic to figure out why the bias supply tried to provide all that current but it made sense at the time, almost as if the bias supply was trying to provide all the current once the other leg opened. The issue is that every now and then, that channel gets a little ratty sounding after playing for a while. It's a great amp and setup overall, especially coupled with the 80B, an active crossover, Krell preamp, krell phono stage and a TNT-MK IV turntable.
So to summarize, I can use the sound card with various suggested codes for high level testing, it's a better option than any of my spectrum analyzers, but I still need some type of load for amps as well as an attenuator for the sound card. Unless I can use the sound card with some type of high-impedance probe?
thanks for all the input.
Jerry
Do not run the amp without load, some do not like it (esp. the digital ones https://www.diyaudio.com/forums/class-d/237086-tpa3116d2-amp.html#post3528020). Just place a power resistor (inexpensive on ebay) on some old heatsink.
Circuits of the divider are plentiful here and elsewhere on internet. This section has many examples too. If you do not mind I would link my take on the issue Balanced Amp Outputs Measurement Adapter
Circuits of the divider are plentiful here and elsewhere on internet. This section has many examples too. If you do not mind I would link my take on the issue Balanced Amp Outputs Measurement Adapter
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.