data acquisition system based on multiplexer conditioning modules and DAQ devices

[akbar]

I want to configure a high multichanel data acquisition system (strain=80, thermocouple= 32, accelerometers=16)) for experimental stress analysis and vibration analysis. What are the limitations/disadvantages/problems of multiplexing based signal conditioning modules and DAQ devices as compared to the simultaneous sampling( having indivisual signal conditioning and digital to analog convertor for each channel)

[markwysong]

There are some basic differences between SCXI-based measuring and simultaneous sampling. First, by using SCXI modules, for many applications you can have individual signal conditioning for each channel (filters, gains, etc). However, multiplexing these signals creates three situations: 1) there is a small delay between channel samples, although this is usually very small, 2) You are usually limited to a smaller scan rate, as the SCXI can multiplex at 333 kS/sec, but your actual max scan rate is 333 kS/sec divided by the number of channes you are scanning (providing your DAQ card can scan at 333 kS/sec or higher), and 3) you sometimes have to deal with settling times of the A/D converter, because you are multiplexing signals to it. If your successive channels are, say, 15 volts different, it may take time for the A/D to settle and sample these channels accurately.

SCXI does have some advantages. The SCXI cards are cheaper, and you can have a higher channel count for less money. But for vibration, you may need higher sampling rates, so you will have to determine if you can get the scan rate you need with the amount of channels you wish to get data.

With simultaneous sampling, you don't have to worry as much about settling times, and you can use the max rate of the cards no matter how many channels you have. These cards tend to be more expensive. It is much more expensive to make a data system of this type.

There's no reason you can't mix these two to make a system which can work for you. For example, thermocouples usually don't change very quickly, so they can be SCXI modules. Strain gauges as well usually don't need very high speed. Accelerometers, however, usually do need some excitation and rather quick scan rates. Therefore, maybe these can be their own simultaneous sampling DAQ board, with the rest as SCXI.

It's best to determine what scan rates you need for your application, and develop a balance between these two methods to meet your needs.

Hope that helped,

Mark

[akbar]

The explanation in the answer is very helpfull. One thing is still to be addressed and that is noise. Whether the multiplexer creats the noise in the system while switching through the controller from chanel to chanel at high scan rate.

[markwysong]

Akbar,

I have used the SCXI multiplexer at its highest rate, and I have not noticed any significant noise. The A/D seems to have a settling rate fast enough that most of this noise is effectively gone, although at the max rate, I have seen a small amount, but only +/- 1 bit.

Mark

[akbar]

Thank you very much Mr. Markwysong for such a valuable comments.That helped a lot.If you dont mind Can I get your Email address for future contact. My Email Address is akbaralipak@yahoo.com

Thanks

akbar ali

[Filipe A.]

The main constraint on a multiplexed signal conditioning system is the aquisition speed that has its maximum at 333KS/s. With that many channels, you will have a scan rate of 333k/128. To do vibration analysis, usualy a high sample rate is needed, so a good frequency analysis can be conducted.

If you could get at least the signals that will be used on the vibration analysis, going to the DAQ board in parallel mode, that initial constraint could be overcomed.

Hope this helps.
Filipe