I recently harvested a used I/O Tech Daqbook 200 data acquisition system from a prototype that we were scrapping out at work.
The system is a parallel port connected 16 bit data acquisition system. We were using it with windows to control a prototype medical device with pumps, tubing clamps, lights, etc.
I have the Daqbook and an expansion chassis with a DBK2 and DBK20 card in it. The DBK2 is a 4 channel D/A voltage output card and the DBK20 is a 48 channel general purpose I/O card.
So my question is-Can I use this to do FFT distortion anaylsis on my tube amps and other gear? Or does it make more sense to pick up an M-audio (or similar) sound card and use some freeware like rightmark audio analyser? Is there anything to be gained by attempting to get this up and running or is a modern sound card and rightmark (or equivalent) software a better solution?
The system is a parallel port connected 16 bit data acquisition system. We were using it with windows to control a prototype medical device with pumps, tubing clamps, lights, etc.
I have the Daqbook and an expansion chassis with a DBK2 and DBK20 card in it. The DBK2 is a 4 channel D/A voltage output card and the DBK20 is a 48 channel general purpose I/O card.
So my question is-Can I use this to do FFT distortion anaylsis on my tube amps and other gear? Or does it make more sense to pick up an M-audio (or similar) sound card and use some freeware like rightmark audio analyser? Is there anything to be gained by attempting to get this up and running or is a modern sound card and rightmark (or equivalent) software a better solution?
You must take into account:
1) The noise floor of ADC/DAC converters
2) The frequency response which is the half of the sampling rate
3) The bit depth which in your case can gives a: 16bit X 6 = 96dB dynamic range.
Sound cards of 24b/192Ks have:
1) 192Ks : 2 = 96KHz frequency response.
2) 24bit X 6 = 144dB dynamic range.
3) Very small noise floor, which is rejected usually, from any FFT software.
The main drawback of a sound card is the delicate inputs of its ADC converters. You can fry the input buffers or the ADC chip very easy, if the amplitude of the measurement signal exceeds the 2Vrms.
Regs
Fotios
1) The noise floor of ADC/DAC converters
2) The frequency response which is the half of the sampling rate
3) The bit depth which in your case can gives a: 16bit X 6 = 96dB dynamic range.
Sound cards of 24b/192Ks have:
1) 192Ks : 2 = 96KHz frequency response.
2) 24bit X 6 = 144dB dynamic range.
3) Very small noise floor, which is rejected usually, from any FFT software.
The main drawback of a sound card is the delicate inputs of its ADC converters. You can fry the input buffers or the ADC chip very easy, if the amplitude of the measurement signal exceeds the 2Vrms.
Regs
Fotios
Last edited:
I have to tell you some useful things.
I have enough experience with sound cards and FFT softwares, so it is better to explain you some "dark" points.
a) About sound cards; i have a S.B. X-Fi Platinum (24b/96K), an EMU-0404 usb and an ASUS Xonar Essence STX pci (both 24b/192K). From the 3 cards, the ASUS has the better performance but is the most delicate from the 3. I have fried one ASUS during its hardwired loopback tests. From the ASUS forum, i have read the same event from other users. Fortunately, i got a new free of cost for replacing the destroyed. The main problem with essence stx, it is that its inputs are not protected against overloading, thus its super ADC CS5361 it is exposed at any time in hazard. The EMU 0404 usb (as well the SB X-Fi) it seems to be very rugged against overloadings.
All cards supported from ASIO driver.
b) I have almost any free FFT analyzer installed in my PC. But i use the Virtins MI Pro 3.1 which is not free of cost. Why? Because its unsurpassed signal generator, its trustable FFT analyzer, its unique Device Test Plan function and its helpful user manual of 250 pages. This software supports the ASIO driver of most sound cards which means that its signal generator it is not dependent on the Windows mixer. The calibration procedure of the sound card with the MI Pro 3.1 is very simple and can be executed in 3 simple steps. There is no need for an external generator or scope for this. You can use the output of the signal generator of MI Pro 3.1 to get a stimulus for calibration. The only external instrument which is needed it is a true RMS voltmeter. You can place the sliders of the patch mix of your sound card at the maximum (only the output master level and the record level). After this, you have to measure the level of signal from the output of sound card with the voltmeter. Now you connect the signal in the input of sound card (external loopback) and you read the RMS value which shows the oscilloscope of MI Pro 3.1. In the calibration window, you write the two values in the corresponding boxes, and then you press "calculate". That is all. After this, you can adjust the level of signal generator from its own sliders in steps of 0,1dBFs.
I bought the ASUS Xonar Essence STX purposively, because it includes a booklet with its performance obtained from an Audio Precision analyzer in a lot of plots. You can find these plots in the site of ASUS as well. My scope was to i make an estimation of the Virtins MI Pro 3.1 as well of the other FFT analyzers which are installed in my PC, in comparison with the Audio Precision analyzer. Indeed, the MI Pro 3.1 after the right setup, gives the closest results with Audio Precision. There are many secrets and tricks (explained in detail in the user manual of Virtins) about the test frequency, the FFT size, the record length and the smoothing windows to you get trustable results from the measurement. For example, if i want to make the usual bunch of tests (THD, THD+N, SNR, NL) in a device, i use as stimulus a sinus of 1007.8125 Hz instead the standard 1KHz, a sampling rate of only 48KHz, a small FFT size of 2048 bins and a double record length of 4096 points. I experimented enough time to find the correct values to obtain similar plots with those of Audio Precision.
In conclusion, there are free or inexpensive FFT softwares which in conjunction with a good sound card can do the job very well, almost like an Audio Precision. Moreover, i believe that the new APx520 series of Audio Precision analyzers (under 10000 USD) indicates the effort of this company to offer inexpensive products, before a giant like Creative or ASUS decides to drop in the market a sound card addressed exclusively for measurements for 200 - 300 USD. And this is not so difficult. The only that is needed is the removing of the useless things from the main processor of sound card like dolby, equalizer etc. and the I/O signal conditioning by changing the input - output buffers, and by protecting the inputs of ADC so as can accept signals up to 200V.
This is the main drawback of sound cards. These can accept directly in their input signals of few Volts. The ASUS concretely, can be fried if the signal which is injected in its input exceeds the 2Vrms. Thus, to make measurements of signals of big amplitude (i.e. from power amplifiers) we are under the compulsion to ad in the input of sound card voltage drop networks which can affect the Zin of card, consequently the precision of measurement.
There is as well another one case: the ready Audio Codecs like PCM2902 of 16bit/48Ks with usb interface. Maybe Texas Instruments has ready a similar new chip of 24bit/192KHz to drop it in the market in 2010! What a miracle! With a such chip on the hand, most of us we can build our external sound card or measurement interface. And then... bye-bye Audio Precision, Sound Technology, Creative, ASUS etc.
Dreams! But who knows?
I quote bellow the plot obtained from Virtins MI Pro3.1 for the ASUS xonar essence stx. You can download the plots of Audio Precision from the ASUS site, to make a comparison.
Regs
Fotios
I have enough experience with sound cards and FFT softwares, so it is better to explain you some "dark" points.
a) About sound cards; i have a S.B. X-Fi Platinum (24b/96K), an EMU-0404 usb and an ASUS Xonar Essence STX pci (both 24b/192K). From the 3 cards, the ASUS has the better performance but is the most delicate from the 3. I have fried one ASUS during its hardwired loopback tests. From the ASUS forum, i have read the same event from other users. Fortunately, i got a new free of cost for replacing the destroyed. The main problem with essence stx, it is that its inputs are not protected against overloading, thus its super ADC CS5361 it is exposed at any time in hazard. The EMU 0404 usb (as well the SB X-Fi) it seems to be very rugged against overloadings.
All cards supported from ASIO driver.
b) I have almost any free FFT analyzer installed in my PC. But i use the Virtins MI Pro 3.1 which is not free of cost. Why? Because its unsurpassed signal generator, its trustable FFT analyzer, its unique Device Test Plan function and its helpful user manual of 250 pages. This software supports the ASIO driver of most sound cards which means that its signal generator it is not dependent on the Windows mixer. The calibration procedure of the sound card with the MI Pro 3.1 is very simple and can be executed in 3 simple steps. There is no need for an external generator or scope for this. You can use the output of the signal generator of MI Pro 3.1 to get a stimulus for calibration. The only external instrument which is needed it is a true RMS voltmeter. You can place the sliders of the patch mix of your sound card at the maximum (only the output master level and the record level). After this, you have to measure the level of signal from the output of sound card with the voltmeter. Now you connect the signal in the input of sound card (external loopback) and you read the RMS value which shows the oscilloscope of MI Pro 3.1. In the calibration window, you write the two values in the corresponding boxes, and then you press "calculate". That is all. After this, you can adjust the level of signal generator from its own sliders in steps of 0,1dBFs.
I bought the ASUS Xonar Essence STX purposively, because it includes a booklet with its performance obtained from an Audio Precision analyzer in a lot of plots. You can find these plots in the site of ASUS as well. My scope was to i make an estimation of the Virtins MI Pro 3.1 as well of the other FFT analyzers which are installed in my PC, in comparison with the Audio Precision analyzer. Indeed, the MI Pro 3.1 after the right setup, gives the closest results with Audio Precision. There are many secrets and tricks (explained in detail in the user manual of Virtins) about the test frequency, the FFT size, the record length and the smoothing windows to you get trustable results from the measurement. For example, if i want to make the usual bunch of tests (THD, THD+N, SNR, NL) in a device, i use as stimulus a sinus of 1007.8125 Hz instead the standard 1KHz, a sampling rate of only 48KHz, a small FFT size of 2048 bins and a double record length of 4096 points. I experimented enough time to find the correct values to obtain similar plots with those of Audio Precision.
In conclusion, there are free or inexpensive FFT softwares which in conjunction with a good sound card can do the job very well, almost like an Audio Precision. Moreover, i believe that the new APx520 series of Audio Precision analyzers (under 10000 USD) indicates the effort of this company to offer inexpensive products, before a giant like Creative or ASUS decides to drop in the market a sound card addressed exclusively for measurements for 200 - 300 USD. And this is not so difficult. The only that is needed is the removing of the useless things from the main processor of sound card like dolby, equalizer etc. and the I/O signal conditioning by changing the input - output buffers, and by protecting the inputs of ADC so as can accept signals up to 200V.
This is the main drawback of sound cards. These can accept directly in their input signals of few Volts. The ASUS concretely, can be fried if the signal which is injected in its input exceeds the 2Vrms. Thus, to make measurements of signals of big amplitude (i.e. from power amplifiers) we are under the compulsion to ad in the input of sound card voltage drop networks which can affect the Zin of card, consequently the precision of measurement.
There is as well another one case: the ready Audio Codecs like PCM2902 of 16bit/48Ks with usb interface. Maybe Texas Instruments has ready a similar new chip of 24bit/192KHz to drop it in the market in 2010! What a miracle! With a such chip on the hand, most of us we can build our external sound card or measurement interface. And then... bye-bye Audio Precision, Sound Technology, Creative, ASUS etc.
Dreams! But who knows?
I quote bellow the plot obtained from Virtins MI Pro3.1 for the ASUS xonar essence stx. You can download the plots of Audio Precision from the ASUS site, to make a comparison.
An externally hosted image should be here but it was not working when we last tested it.
Regs
Fotios
Last edited:
- Status
- Not open for further replies.