measured voltage

[dmorris]

I'm measuring the output of a piezoelectric sensor.  It's just an analog voltage generator with a two-pin output.  The nominal output gets up near a volt peak-to-peak, and when I feed these two pins into a scope, I see output in that neighborhood when I stimulate (tap on) the device, at least getting up above 400mV peak-to-peak.

Now I'd like to read that same sensor via a USB-6009.  I had no luck in single-ended mode (almost all noise), but the signal looks okay in differential mode.  So I'm connecting just one sensor to the AI0+ and AI0- pins on the 6009.  I'm doing all of my testing in the "Test Panel" in MAX right now.  I set the appropriate channel, typically set the max/min inputs to 1 and -1, respectively, set that input configuration to "differential", set the mode to "continuous", and the rate to 1000Hz and "samples to read" to 1000 (so the screen is updated once a second).

Here are the two issues I'm having:

* The overall output is reading very low compared to what I'm used toseeing on the scope.  Instead of hundreds of millivolts, I'm seeing more like 60mV peak-to-peak.  Signal-to-noise is not terrible, but I feel like it could be a lot better, and it is better when I read the same signal on other devices (a scope, for example).  Any tips on why I'd see such low-voltage output when wiring into the differential analog inputs on the 6009?  Here's a shot of my capture window:

* Second issue: almost every time I stop and re-start the device, I get the following error:

I have to un-plug the device and plug it back in to get it to work again.  There's nothing particularly CPU-intensive happening on my machine, I'm not anywhere near the bandwidth limit of the device or the bus, and it doesn't feel like this is really a load issue anyway since it _always_ works the first time I press "start".  Soft-resetting the device doesn't help when this occurs, I need to unplug the device and plug it back in.  Any ideas what could be happening here?

Thanks!

-Dan

[Henrik_Volkers · Accepted solution]

A piezoelectric sensor generates a charge and has a very high impedance. The correct way to adapt it to your 6009 is a charge amplifier. It is possible to measure the output as a voltage, however you need at least an amplifier with a very high input impedance AND short, low cap cables!  

So your results are very reasonable 😛   take a look at the input impedances of both devices.

Depending on your soldering skills a simple (maybe not so linear) charge amplifier can be build up with a high impedance, low input bias current OPamp  and a (poly) capacitor  and a >10M resistor ......

Take a look at this OP LT1113 

or the LMH6611.  

[dmorris]

This is super-helpful, thanks... I assume the reason I'm able to read higher voltages on a scope from the same sensor is just that a scope typically has a much higher impedance?  The input impedance of the 6009 is listed as 144kohm, and the output impedance of the device ranges from 1Mohm to 650Mohm across the relevant frequency range, so that would explain most of what I'm seeing.  Sounds like I'll have to either use a pre-amp as you suggested, or work with what I'm seeing.

Thanks!

[Marking as a solution but still hoping someone has thoughts on the other issue...]

-Dan

 

[Ali M]

The USB-6008/6009 has a relatively small onboard FIFO, with a Windows OS when you perform operations such as opening other programs, minimizing/maximizing windows which are CPU engaging, the driver cannot access data from the device, the FIFO overflows/get the -200361 error. The best solution is to minimize all other operations on the computer. I've inlcuded a link below regarding error -20061

1. USB-6008/6009 Error -200361: Buffer Overflow Error:

http://digital.ni.com/public.nsf/allkb/611475F9BE62881E86256FDC0062B1BB?OpenDocument

Regards,

Ali M

Applications Engineer

National Instruments