Accelerometer/Vibration measurement. Difficulty: Total newb to vibration measurement.

[Ralph@NES]

Hello all.

I've got a shaft and some ball bearings spinning, and I need to monitor the vibration for potential failures.

I have a small army of MSI ACH-01 general purpose accelerometer ( http://www.meas-spec.cn/manage/sensortypes/ACH_01.pdf ) and the supplied IB-ACH-01 amplifier.

I may be reading this wrong, but what I THINK I should be getting as far as signal from this accelerometer is voltage, right?  I am using a SCXI-1122/1322 to obtain a voltage signal from this thing and there seems to be no 'baseline' so to speak; I get everything from a -x to a +x voltage depending on, apparently, nothing.  I've got my VI setup with a max and min value that would shut the system down if there was any positive or negative value outside a baseline, but I can't even use that since the 'baseline' is all over the place!

Has anyone worked with this sort of equipment before?  What direction should I be looking as far as successfully connecting and analyzing the signal from this unit?

Thanks very much in advance.

[Mark_E]

Hi Ralph,

Yes, you measure voltage from accelerometers.

You would need to use the SCXI-1122/1322 (typically temperature measurements) or other hardware for signal conditioning, but I see that your IB-ACH-01 is an amplifier with a gain of 10. That may be enough to get a reasonable signal.

You need to connect your SCXI to a computer somehow, either with a DAQ device or by other means. What device are you using to measure the output of your SCXI hardware? In the case where your box is doing the signal conditioning, you might want to just plug the box into whatever DAQ device is in your computer and measure it directly.

What type of noise are you seeing and how much? Accelerometers are sensitive devices, so your ‘baseline’ fluctuation could be normal.

There are specific tools available to help you with analyzing vibrations for potential failures. We have an order analysis toolkit that is bundled with our Sound & Vibration Measurement Suite. You can learn more about them in the following links:

Ni.com/soundandvibration
Performing Order Analysis in LabVIEW
NI Sound and Vibration Software Comparison

[Ralph@NES]

Hello Mark, thanks for the response.

I currently have a PCI-6229 DAQ and a chassis with the 1122/1322 and 1102/1302 modules.  I'm using the 1122 with this sensor.

What I gathered from talking to a person very familiar with this IB amplifier is that we're looking at AC voltage here, not DC.

They recommended I sample at at least 20khz to get a true full spectrum of what my 'baseline' is, and that if I want any kind of averaging that I would have to use Labview to average or get an RMS on the signal.

Here's where I start getting into murky water....  I've never been able to use Labview to filter without things getting confusing.

As far as the noise goes, I really don't know if I have noise or if the way I was sampling (1 sample, on demand) was only getting me odd points in the AC voltage spectrum that the item is putting out.

Any input on how to structure some code to get me an RMS reading on the voltage out (being AC) would be greatly appreciated, even though this isn't the right forum.

Thanks again for the assist!

Ralph

[Mark_E]

Hi Ralph,

You can test your signal very quickly by going in to Measurement & Automation Explorer (MAX) and creating an accelerometer task to see how things are working.

How to create a task:

1.       Right-click on Data Neighborhood, choose “Create New…” then “NI-DAQmx Task”

2.       Choose Acquire Signals » Analog Input » Acceleration

3.       Select your device and channel

4.       Name the task

5.       Set your acquisition mode (lower left) to continuous and adjust any other parameters you need

6.       Press “Run” at the top.

You should be seeing real data at this point. Hopefully that gets you started. Feel free to post a screenshot if your data isn’t what you expected.

[Ralph@NES]

As usual, here I am trying to make things more difficult than necessary.  🙂

Thanks for the input Mark, I'm going to try that now.

[Ralph@NES]

Mark, I just tried that and had no luck.  When it got to the screen for choosing device, it said no supported devices found.

So I guess that the 1122 isn't going to recieve whatever signal that the acceleration task wants.

Other ideas?

[preston_johnson]

the 1122 is not normally an accelerometer input module.  Try creating a voltage task with the sensor.  You can create a custom scale to change from voltage to G based on the sensitivity of the sensor.  For example, if the sensor is 100mv per g, then you would multiply the voltage value by 1/100mv or by 10 to get g. 

[AnalogKid2DigitalMan]

Hi Ralph:

Looking at the amplifier schematic, is has a capacitor in series with the output amp, so yes, it is AC coupled.

Since the bandwidth of the sensor is 20KHz, you may want to sample at least at a 40KHz rate in case there are frequencies in your vibration scenario that are greater than 10KHz you mentioned.

Hope you get this up and running!

-AK2DM

[Clecio]

Dear,

For a similar application using an accelerometer ACH-01 (Measurement Specialties), I would like to use with a USB-6009. Is it possible?

[EricS.]

The USB-6009  cannot connect to an SCXI chassis, so you cannot use the same SCXI modules the above users mentioned.  Also the 6009 cannot supply the IEPE excitation current which is required by most accelerometers.  Instead you would have to use a USB-9233 or USB-9234 for IEPE accelerometers.

However, the ACH-01 is not a IEPE accelerometer, but instead outputs a voltage, and requires voltage excitation instead of current.  So, in theory you could use it.  However, the accuracy of the 6009 is not very precise. The ACH-01 has a sensitivity of 9mV/g ( See Data Sheet) and the USB-6009 has an absolute accuracy of 14mV (6009 data sheet).  So, at any given time your reading could be up to 1.5g's off from what the acceleration actually is.

If this is accuracy is good enough for your application then you could definitely use it.