Unusual signal acquired from piezoelectric sensor

[Msc_stu]

I'm acquiring piezoelectric signal for traffic system using NI 9215 and NI USB-9162 modules. The sensor is designed for vehicle detection (Roadtrax BL). However there is instability in the signal. After the vehicle passes the voltage start increasing gradually to high values, sometimes even higher than the vehicle peak it self. Then it takes relatively long time to settle. Is there any signal conditioning, or any other idea, that I can apply to solve this problem?

Also unusually the DAQ is giving an output voltage of 2 v and 60 Hz when nothing is connected to it, a very strange reading knowing that the module takes its power through a USB cable. Any ideas? Thanks.   

[AnalogKid2DigitalMan]

Almost sounds like an impedance mismatch between the sensor and DAQ. What type of signal conditioning does the sensor have? If you can post attachments for any data that would be helpful (Roadtrax site requires registration).

As to the DAQ detecting 2V @ 60Hz with nothing connected, that is proabably normal due to it's high input impedance picking up ambient AC powerline fields. Short the input together, it should yield 0V

-AK2DM

[Msc_stu]

Thanks for the reply. As for the the 2v, 60Hz signal, I tried to short the input before and the 60Hz dropped to millivolts.

As for the instability in piezoelectric signal, I thought that the -75 CMRR of the NI 9215 might be the problem. However since I'm using differential mode, then common-mode signal should not be a problem.

Attached is the datasheet for Roadtrax BL. If the sensor have high impedance, would the impedance mismatch create a problem?

regards,

Samer

[AnalogKid2DigitalMan]

Samer:

What I think is going on is that both the sensor and the DAQ have high resistances. The sensor is primarily acting as capacitor/voltage generator yielding a long RC time constant. Is the risetime of the signal fairly fast?

In differential mode, use a 1Mohm resistor tied between the AI- and COM of you 9215.

-AK2DM

[Msc_stu]

AK2DM

Thanks again for your help. Rise time of the signal after the vehicle passes is relatively larger than that of the vehicle it self. I'm going to test tomorrow evening using the 1MOhm resistor and post a feedback.

Regards,

Samer