Shock DAQ using USB-9234

[gardel]

I need to record 1 to 5 Vpp at 10,000 Hz from a shock accelerometer.  I have some USB-9234 hardware.  What is the slew rate of the USB-9234?  If I filter the analog input signal at 23 kHz, do you think the USB-9234 has sufficient sample rate?

[Mr.F]

Hi Gardel,

Can you elaborate a bit on what a “Shock Accelerometer” is? The filtering information for the NI 9234 is available in the manual for the device.

http://www.ni.com/pdf/manuals/372307a.pdf

Based on this, the 9234 has a sufficient sampling rate to capture your 10,000 Hz signal. What does the signal coming from the accelerometer look like? The 9234 does have a low pass filter to remove high frequency components, so those components of the signal will be attenuated. Let me know if you have any additional questions.

[Gabo_79]

Hi, I would like to understand why using 9234 and IEPE accelerometer I got different results than using an specific shock sensor for drop tests. I'm configuring my tri-axial accelerometer to 25.6 Khz sampling rate. Raw data usually shows a higher shock level with 9234 module than my shock sensor. I'm processing data with Sound and Vibration SW. Do I need to apply an specific filter at software level to match the same wave shape as reported with the shock sensor?

[QueenSelene]

Hi Gabo_79

Just as a side note you may get more visibility on your question if you create a new post instead of continuing to post on this one as it's quite a few years old. 

That being said, I think it would help to have a bit more information about the specific two sensors you're using. It sounds like they are maybe just calibrated differently, I don't think it's how you have things configured in software. 

Is the accelerometer consistently the same amount higher than the shock sensor or does it scale according to the value measured? Are you able to perform other acceleration tests with both sensors (preferably with a known value) to compare them there, to see if this just shows up in drop tests or otherwise, and to see which is actually reading closer to the expected value. 

It may also be helpful if you can upload some pictures of your measurements so we can see the difference. 

Thanks!

[Henrik_Volkers]

---

@Gabo_79 wrote:

Hi, I would like to understand why using 9234 and IEPE accelerometer I got different results than using an specific shock sensor for drop tests. I'm configuring my tri-axial accelerometer to 25.6 Khz sampling rate. Raw data usually shows a higher shock level with 9234 module than my shock sensor. I'm processing data with Sound and Vibration SW. Do I need to apply an specific filter at software level to match the same wave shape as reported with the shock sensor?

---

It's all about frequency response 😄

Some shock sensors have a build in mechanical 10kHz filter ...  

A 'real' shock can have energy content at higher frequencies...  your transducers and DAQ have transfer functions ...

IF you have the transferfunctions of all your measuring chains you migth be able to apply a filter, IF you have to cut signal content! ...   It's hard to add content your transducer already cut away 😉

Here is a paper on that effect:Current Developments in the Field of Shock Calibration 

(from 2006, sorry my PHD isn't ready 😉 )

[Henrik_Volkers]

additional comment:

one experiment you can do:

Apply short shocks -> high frequency content

Measure with both sensors , swap positions , measure again,

do a FFT on the signals and look at the ratios (if you swaped the positions, multiply the complex ratios and take the square root)  , that's what your filter has to look like

However: Does it make sense to tweak a signal to match a sensorsignal that migth has a restrickted (10kHz limit?) bandwidth?  The real shock is still different 😄

If you can, post some data, I'm curious 😉

[Gabo_79]

I'm attaching the raw data got from the standalone sensor DTS and the other option was using NI9234 + PCB accelerometer. It was a drop test from 2 inches on concrete floor. Both sensors were installed in a CPE system and there were very close each other, recording during the same event. I was expecting to have a slightly different results, but the raw data shows a big difference in amplitude. From my prespective looks like the standalone sensor is applying a built-in filtering. Question is, Which one is the real impact on my product?

[Henrik_Volkers]

DO you have the calibration sheets of the sensor?

Exact types?