Is NI-9230 a good choice for PCB 393B12 accelerometer?

[ShahabJHU]

Hello Community,

I want to measure floor vibration and the maximum expected frequency is 1000 Hz. 

Is NI-9230 a good choice for PCB 393B12 accelerometer? Do I need PCB signal conditioner or I can just connect them?

NI 9230 BNC, 3-Ch, +/-30 V, 12.8 kS/s/ch, 24-Bit, IEPE AI Module

Is the sampling rate of 12.8kS/s/ch seems enough for maximum 1k Hz data?

Thanks

[santo_13 · Accepted solution]

In general, 9230 is a good choice for vibration measurements (due to the 24-bit Delta Sigma ADC and excellent timebase)

Next, general rule of thumb, 10x the max signal frequency gives a very good sampled signal. But given this is vibration measurement, 10x sampling rate is plenty enough. So, yes, 12.8kS/s rate is great to capture 1kHz signal.

If you're looking to measure any kind of harmonics, the max harmonic is also limited by your sampling rate, in this case, a max 6.4kHz (Nyquist).

Refer to this article to see if your sensor signal is compatible with 9230's IEPE and signal range - https://knowledge.ni.com/KnowledgeArticleDetails?id=kA00Z000000P8uwSAC&l=en-US

[Henrik_Volkers · Accepted solution]

We use 353B31 accelerometers (similar sensitivity) for noise measurements on optical tables (or other vibration sensistive stuff) and  add a Kistler signal conditioner if we really want to digg in the noise and can't place big & heavy seismometers.

But by looking at the specks and if you don't want to go down or below 1 Hz* without chain calibration this combination tooks fine IMHO.

Edit: *)To measure mechanical transferfunctions in magnitude and phase with low uncertainty 😉

[ShahabJHU]

Thanks for the great help!

Is NI-9231 a better option for me? Because we need more than 4 channel. Could you please compare it vs 2 x NI-9234?

I noticed that my sensor output is±0.5g x 10000 mv/g=±5V. If I use NI-9230 with ±30 V input range, do I get lower resolution rather than using NI-9231 with ±5V input range? It seems NI-9231 is even faster 51.2 kS/s per channel update rate and 24-bit resolution. Indeed more expensive but it seems worth it!

Thank you,

[Henrik_Volkers]

---

@ShahabJHU wrote:

Thanks for the great help!

Is NI-9231 a better option for me? Because we need more than 4 channel. Could you please compare it vs 2 x NI-9234?

I noticed that my sensor output is±0.5g x 10000 mv/g=±5V. If I use NI-9230 with ±30 V input range, do I get lower resolution rather than using NI-9231 with ±5V input range? It seems NI-9231 is even faster 51.2 kS/s per channel update rate and 24-bit resolution. Indeed more expensive but it seems worth it!

Thank you,

---

I would go with the 8 CH , but if you need to recable a lot, the extra space migth be more comfortable...

Concerning sync timing all in one looks better.

the resolution isn't your problem, the noise is of more concern.  the 9231 seems better by spec.