Hi all,
I have been looking at building Pete Millett's soundcard interface board, for testing and analyzing amplifier output.
Then I stumbled across a low(ish) cost NI USB-based DAQ module:
NI USB-6009 - 14-Bit, 48 kS/s Low-Cost Multifunction DAQ - National Instruments
It is only 14-bit, up to 48 KHz, and there are probably some software hurdles (acquisition of LabView), but it can handle +/- 10V.
Has anyone successfully used a DAQ for this type of application?
Thanks.
I have been looking at building Pete Millett's soundcard interface board, for testing and analyzing amplifier output.
Then I stumbled across a low(ish) cost NI USB-based DAQ module:
NI USB-6009 - 14-Bit, 48 kS/s Low-Cost Multifunction DAQ - National Instruments
It is only 14-bit, up to 48 KHz, and there are probably some software hurdles (acquisition of LabView), but it can handle +/- 10V.
Has anyone successfully used a DAQ for this type of application?
Thanks.
Pete's board is basically an instrumentation amp (Burr-Brown) with some other nice features.
The ELVIS board is really cool... wish we had that when I was in school! But I don't have access to one, nor do I have LabView.
If you don't mind the calibration hassles and low bandwidth, I would second the use of a sound card. A good USB sound card will have a quite low noise floor, but even a built-in sound card can be acceptable. The main drawback of a sound card is that the full scale voltage is often not specified -- and rather high in many cases (10 V for many external sound cards). If you are willing to calibrate the system or if you just need ratios between, say a fundamental frequency and harmonic content, a sound card is fine. It gets even better if you add Pete Millett's interface.
However, if you want to measure very low-noise circuits, there's no substitute for a real instrument. Like an HP 3562A (-140 dBV noise floor) -- or a modern equivalent. But those are more expensive, of course.
I would shy away from LabView and NI at all cost. NI's products tend to only work with VISA drivers and LabView so you paint yourself into a pretty tight box from the start. I hate LabView with a passion. Granted, the end result (virtual instrument) can be really pretty and sometimes even useful, but the process to get there is dreadful. The idea of doing "programming" with variables represented by wires (different color for different data types) is not compatible with my way of thinking. Maybe it's because I was brought up with real programming languages (Basic, Comal-80, Pascal, C/C++, assembly language, etc) that the idea of doing even simple arithmetic using colorful icons drawing wires between them just does not represent a smooth workflow. Try doing something more complex, like loading a shift register makes you want to shoot yourself. Something like "A >> 4" in C takes a screenful of icons and wires to accomplish in LabView. Anyway... Like I said. I hate it. And it's been a while (2-3 years) since last I was forced to use it. It may have gotten better since then. And you are of course free to form your own opinion. I'll get off my soap box now...
~Tom
However, if you want to measure very low-noise circuits, there's no substitute for a real instrument. Like an HP 3562A (-140 dBV noise floor) -- or a modern equivalent. But those are more expensive, of course.
I would shy away from LabView and NI at all cost. NI's products tend to only work with VISA drivers and LabView so you paint yourself into a pretty tight box from the start. I hate LabView with a passion. Granted, the end result (virtual instrument) can be really pretty and sometimes even useful, but the process to get there is dreadful. The idea of doing "programming" with variables represented by wires (different color for different data types) is not compatible with my way of thinking. Maybe it's because I was brought up with real programming languages (Basic, Comal-80, Pascal, C/C++, assembly language, etc) that the idea of doing even simple arithmetic using colorful icons drawing wires between them just does not represent a smooth workflow. Try doing something more complex, like loading a shift register makes you want to shoot yourself. Something like "A >> 4" in C takes a screenful of icons and wires to accomplish in LabView. Anyway... Like I said. I hate it. And it's been a while (2-3 years) since last I was forced to use it. It may have gotten better since then. And you are of course free to form your own opinion. I'll get off my soap box now...
~Tom
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.