Measuremnt hardware conundrum.

[Pano]

OK, maybe some of you can help. I want to make some valid measurements of the DCX2496, both in its stock form and with some modifications. But it seems that hardware limitations will get in the way.

Since I do not own or have access to high quality measurement gear such as AP, HP or Agilent, measurements will have to be done with more modest devices. Software does not worry me, but hardware does. The best equipment I have to do these measurements with is an M-Audio USB Duo soundcard. So far it has been great for loudspeaker measurements, which is what I bought it for, but for measuring a line level device like a DAC/crossover, it's going to be the limiting factor.

Here is why:
I do not know the performance of the ADC in the DUO card. And how can it be determined? With a simple loop back from the analog output the measurement will contain both DAC and ADC performance. How could the ADC only be determined? I don't have a known source with which to calibrate the ADC. With a known source, the ADC distortions could be nulled.

Noise is not hard; just measure a non connected or terminated input. But spectrum? No reference.

Although of course the difference between stock and modified can be measured, it would still be nice to have a baseline – to know the influence of the ADC. I can look at the waveforms on the O'scope, but that does not help with FFT, which is going to be important.

Anyone have any suggestions of a way around this conundrum? How to determine the influence of the ADC on the measurements?

[MJL21193]

Hi Pano,
I'm not sure what you are looking for - it's a soundcard like any other - you can test it's performance with RMAA, just loop line in to line out adjust the levels and away you go.

[AR2]

So far I had the best results with Right Mark Audio.
Latest News. Audio Rightmark

Nice thing is that is free. I have a paid version so it could utilize ASIO drivers for my Fireface 800, but it works great with any other card with the free version. I do not think you will be able to isolate ADC. The only way to do it is if you are not doing ADC at all, but digital in. That way you could send digital out of your card to Behringer and than measure the loop. It will give you a lots of useful data as well as you could save the file. I would suggest to work in 24/96 mode because your Fq. response will be extended to 48KHz.

[zener_diode]

Originally Posted by panomaniac

OK, maybe some of you can help. I want to make some valid measurements of the DCX2496, both in its stock form and with some modifications. But it seems that hardware limitations will get in the way.

Since I do not own or have access to high quality measurement gear such as AP, HP or Agilent, measurements will have to be done with more modest devices. Software does not worry me, but hardware does. The best equipment I have to do these measurements with is an M-Audio USB Duo soundcard. So far it has been great for loudspeaker measurements, which is what I bought it for, but for measuring a line level device like a DAC/crossover, it's going to be the limiting factor.

Here is why:
I do not know the performance of the ADC in the DUO card. And how can it be determined? With a simple loop back from the analog output the measurement will contain both DAC and ADC performance. How could the ADC only be determined? I don't have a known source with which to calibrate the ADC. With a known source, the ADC distortions could be nulled.

Noise is not hard; just measure a non connected or terminated input. But spectrum? No reference.

Although of course the difference between stock and modified can be measured, it would still be nice to have a baseline – to know the influence of the ADC. I can look at the waveforms on the O'scope, but that does not help with FFT, which is going to be important.

Anyone have any suggestions of a way around this conundrum? How to determine the influence of the ADC on the measurements?

You could always check to see what A/D is on the card and get the data
sheet for it. Then get the data sheet for the opamps on the card. The
distortion numbers will add up, but probably the A/D will be the determining
factor.

Or you could pick up a DAL CarDeluxe. I have one and from output to input
(direct connected) I get .00055% THD+N at full scale. Also, the IMD is
.0005%. Pretty good for a computer audio interface.

Digital Audio Labs: Digital Audio Labs Home Page

See attached plot of the input to output spectrum.

[Pano]

Thanks guys.
John, I see that I forgot to mention that I'll be going digital into the DCX, so the analog out of the DUO card will not be used. If I were going Analog-Digital-Analog it would be less of a problem, for sure.

But I don't use the analog in of the DCX and would like to measure just Digital>Analog.

Sorry for the confusion!

I've been wondering if there might be a way to measure another DAC or 2 and do some math to figure out what parts of the measurements belong to the DACs and what to the ADC. Possible?

[MJL21193]

Oh, ok. But...
If the mods you've made are limited to the output stage, you can use the analog input to run the tests and show if there is a measurable difference.


BTW, just noticed that you are now a mucky-muck...I'll need to watch my P's and Q's
I think you will do a great job.

[Pano]

Thanks John! =)

Yes, the analog path might be the way to go. I'll have to give it a try. Was really hoping to use the digital input.