Data acquisition at maximum sampling rate

[Rama2141]

Why NI-9223 module does not work directly for ICP/IEPE based accelerometers (eg. PCB 356a16 )? How can we make this work?

While NI 9234 is a sound and vibration module where we can use ICP/IEPE based accelerometers directly and take measurements, NI 9223 seems to not work with such accelerometers. In NI-9234. the excitation voltage is provided to the accelerometers from the module itself. But the maximum sampling rate available is 51kHz per channel only. NI-9223 is a voltage measurement type module which can give sampling rate of 1 MS/s/channel but it is not designed for IEPE based accelerometers where excitation voltage is needed for accelerometers to get data. Can we connect some charge amplifier or signal conditioner etc to make this work? Please explain and suggest some examples. Thank you.

[santo_13]

9223 provides only a 16-bit dynamic range, whereas a typical sound and vibration module provides a 24-bit dynamic range.

In short, it is not the same for sound and vibration measurement applications! Depending on your use case, you may be able to live with the constraints.

[mcduff · Accepted solution]

---

@Rama2141 wrote:

Can we connect some charge amplifier or signal conditioner etc to make this work? 

---

You can use a signal conditioner with the 9223, something like Model 482C16 | PCB Piezotronics

However, you will need to find a signal conditioner that has the bandwidth you need; the one mentioned above only goes to 100kHz.

[Henrik_Volkers]

Well, for the PCB-356A16 this wouldn't make sense, since the specified frequency range is 5 kHz and might be useable up to 20 kHz if a dedicated calibration in that frequency range is available, followed by a adequate correction in the frequency domain (since the complex transferfunction is not constant! ). So  51 kSPS is fine for this type (or class) of sensor. (and it will get hard to find a (certified) calibration lab  for accelerometers that goes beyond 20 kHz 😉 )

For shock measurements with short pulse durations you will have some energy beyond 20 kHz and a 1 MSPS aquisition can make sense, however this involves other sensors and signal conditioners .. and it get tricky to measure (vs detect)  beyond 100 kHz with accelerometers  BTDT (and I prefer Laservibrometer 😄 )

So what is your actual use case?