LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Erroneous input values from accelerometer

Hi! I hope this is the correct topic for my question. I struggle at making my acquisition system acquiring correct measurement values. I use a NI USB-4432 device with LabView 2014 SP1 on windows 7. My setup involves an impact hammer test. In particular, I use a PCB ICP 086E80 impact hammer and a PCB ICP 352A24 accelerometer (-50G to + 50G, with 100 mV/G as sensitivity, with excitation current of 2,1 mA). The hammer enters the ai0 channel, while the accelerometer is put on the ai1 channel. Finally, the hammer triggers the acquisition of both the hammer impact and acceleration within the device. The trigger is an analog edge with rise slope. The DAQ assistant virtual channels settings are attached as images. Basically the hammer impact range is from -5 to 15 N, while the accelerometer is used at its full range (-50G to 50G). My problem is the following: when I perform the impact, the accelerometer registers an acceleration which initially ranges from -120 to 120 G, then correctly sets on values below +-50 G. I would expect my accelerometer to saturate, nevertheless the acquired waveform seems not having troubles with such out of range values. I tried to convert the virtual channel from acceleration to voltage and measure the input. In this configuration, the sensor seems not to saturate at +- 5 Volts, which should correspond to +- 50 G for the acceleration. I have attached some screenshots to better explain the problem. 

I hope someone could help me, as I tried everything but still cannot get my head around it.

Thank you in advance to everyone would help me!

Download All
0 Kudos
Message 1 of 6
(3,235 Views)

Hi! I am not sure this is the correct topic for such a question, because I struggle at identifying whether my problem is at a HW or SW level. I struggle at making my acquisition system acquiring correct measurement values. I use a NI USB-4432 device with LabView 2014 SP1 on windows 7. My setup involves an impact hammer test. In particular, I use a PCB ICP 086E80 impact hammer and a PCB ICP 352A24 accelerometer (-50G to + 50G, with 100 mV/G as sensitivity, with excitation current of 2,1 mA). The hammer enters the ai0 channel, while the accelerometer is put on the ai1 channel. Finally, the hammer triggers the acquisition of both the hammer impact and acceleration within the device. The trigger is an analog edge with rise slope. The DAQ assistant virtual channels settings are attached as images. Basically the hammer impact range is from -5 to 15 N, while the accelerometer is used at its full range (-50G to 50G). My problem is the following: when I perform the impact, the accelerometer registers an acceleration which initially ranges from -120 to 120 G, then correctly sets on values below +-50 G. I would expect my accelerometer to saturate, nevertheless the acquired waveform seems not having troubles with such out of range values. I tried to convert the virtual channel from acceleration to voltage and measure the input. In this configuration, the sensor seems not to saturate at +- 5 Volts, which should correspond to +- 50 G for the acceleration. I have attached some screenshots to better explain the problem. 

I hope someone could help me, as I tried everything but still cannot get my head around it.

Thank you in advance to everyone would help me!

Download All
0 Kudos
Message 2 of 6
(3,267 Views)

Hi Riccardo,

 

do you really think you need to attach 3 images of nearly 20MB in sum two times in the forum?

Why do you need to attach such large images? (Can't you use the default Windows snapshot tool?)

 

My problem is the following: when I perform the impact, the accelerometer registers an acceleration which initially ranges from -120 to 120 G, then correctly sets on values below +-50 G. I would expect my accelerometer to saturate, nevertheless the acquired waveform seems not having troubles with such out of range values. I tried to convert the virtual channel from acceleration to voltage and measure the input. In this configuration, the sensor seems not to saturate at +- 5 Volts, which should correspond to +- 50 G for the acceleration.

So neither the displayed acceleration nor the underlying voltage signal saturate. What's wrong with it?

 

Best regards,
GerdW


using LV2016/2019/2021 on Win10/11+cRIO, TestStand2016/2019
0 Kudos
Message 3 of 6
(3,225 Views)

Hi GerdW,

sorry for irritating you with such a huge amount of wasted space in 2018 on the NI forum, actually I have so much things to do that my very last problem was to save space with the attached images.

 

Anyway, probably you did not read my question, besides noticing the image size. My setup should be correct for acceleration measures, and my accelerometer seems to measure way beyond its acceleration measurement range. What's wrong with it? Well, I think this should not be normal behavior, and since I need exact physical values from my equipment I probably would not trust such a measurement output. I would be glad if you could help me debugging the problem first, than complaining about the images.

 

All the best,

Riccardo

0 Kudos
Message 4 of 6
(3,220 Views)

Hi Riccardo,

 

even with a measurement range of ±50g your accelerometer is able to detect even higher accelerations - and can output them.

As there is no in-built (real) signal conditioning you should not expect some saturation…

 

 sorry for irritating you with such a huge amount of wasted space

It's not "irritation", it's annoying to have to download 20MB over (potentially) slow internet connections when YOU can (and should) provide information with much less bandwidth demand.

See the attachment…

Best regards,
GerdW


using LV2016/2019/2021 on Win10/11+cRIO, TestStand2016/2019
0 Kudos
Message 5 of 6
(3,199 Views)

Hi GerdW,

thank you for your answer. So this behaviour should be quite acceptable. But there is still a problem, because while the input voltage is always within the voltage range of the system (which should be +-5 V), once the virtual channel is converted to accelerometer from voltage measure, there is no more correlation between the previous average measured voltage and the actual average acceleration. 

 

As for the image size, you are right. I did not take into account downloading problems with such big images. That is my fault and I apologise for that, I will pay attention to this aspect next time 🙂

0 Kudos
Message 6 of 6
(3,180 Views)