Signal Conditioning

cancel
Showing results for 
Search instead for 
Did you mean: 

strain amplification using cDAQ

Solved!
Go to solution

Hello,

I am a newbie in using DAQ's and hope some one can help me out.

I recently bought a cDAQ with NI 9237 to measure strains. The strains that I am looking to measure are very small 1 micro strain. A little bit of research showed me that I need to amplify the signal to measure such small strains (correct me if I am wrong). However I am not able to find any amplifiers for the cDAQ.So, the questions I have are -

 

1. Is it possible to measure such small strains (1 micro strain) using the cDAQ ? If so, do I need an amplifier ?

2. If I need an amplifier, are there specific ones for the cDAQ  with the 9237 ? If not, where/how can I get one ?

 

Thanks,

Sid

0 Kudos
Message 1 of 16
(6,764 Views)

Sid,

 

Although it is possible, I would try to avoid amplifying if possible unless completely necessary.  It will only serve to add extra noise considerations to your readings.

 

The 9237 is a 24bit resolution module and should be able to measure precisely enough to meet your requirements if paired with the correct strain gauge.   There are many different types of strain gauges and your ability to measure this precisely will depend on the output from the strain gauge.  The 9237 module has an input range of +/- 25mV/V.  There should be plenty of gauges (you will have to look at the specs) to offer the precision you want in that range.    Keep in mind, you need to choose to measure very precisely over a small range or less precisely over a larger range, you can't have both!

 

 

Regards,

 

 

 

 

Peter C.
Applications Engineer
National Instruments
0 Kudos
Message 2 of 16
(6,744 Views)

Hello Peter,

 

Thanks for your reply. However, I am not sure I entirely understood you there. What do you mean by “precision in that range (+/- 25 mV/V)”? I looked up a few Strain gage data sheets (http://www.setpoint.gr/en/site/technicalibrary/Products/7%20Vishay/VISHAY%20Precision%20Strain%20Gag... ) and all I see there is a Strain range as +/- 3% (or) +/- 1.5 % depending on the gage we select. Isn’t this strain range the maximum strain the gage can measure? Could you clarify and correct me if I missed something.

 

Thanks,

Sid

0 Kudos
Message 3 of 16
(6,733 Views)

Hi Sid,

 

I want to apologize, after looking further into this it looks like I was mistaken.   The absolute accuracy of the 9237 device is outline in this document here:  http://zone.ni.com/devzone/cda/tut/p/id/8909

 

As you can see, (assuming Gauge Factor = 2 with full bridge configuration) the absolute accuracy of the 9237 is 12.5 mm/m +/- .725mm/m.    I have discussed this with some colleagues and, unfortunately, we do not believe there is a way to amplify the readings to allow the 9237 to measure with the precision you desire.

 

What is it you are trying to do? 

 

 

Peter C.
Applications Engineer
National Instruments
0 Kudos
Message 4 of 16
(6,706 Views)

Hi Peter,

 

 

This is the test I am doing –

I am taking a thin cylinder and subjecting it to some internal pressure. My test involves in finding out what the internal pressure is and if it changes over time. My plan is to keep a strain gage on the cylinders outer surface to measure the hoop strain and back calculate the internal pressure. I am expecting strains of about 500 micro strain (based on my cylinder dimensions, material properties and rough pressure estimates). However, the changes of pressure in time could cause strain changes in small micro strain increments to reach a final value of about 550 micro strain or less. I would like to measure these small strain increments with a high degree of accuracy.

 

The smaller strain I can measure the better data I will have. If you think that measuring a 1 micro strain increment is very difficult, I could live with measuring 5 ~ 10 micro strain increments. Either way the document that you provided tells that the smallest strain the 9237 can measure is 12.5 mm/m +/- 0.725 mm/m (or) 12,500 +/- 750 micro strain which is surprisingly huge!! Are there any other NI DAQ’s that can measure strains in the range (10 micro strain) I want?

 

Sid

0 Kudos
Message 5 of 16
(6,701 Views)
Solution
Accepted by topic author nssid

Sid,

 

Those results were very shocking to me as well, so I looked into it further.  It appears the 9237 can do better.   Assuming you are within 5 degrees of the calibration temp (25C), the 9237 has .05% gain error and .05% offset error.  With full bridge configuration and 25mV/V range you can calculate the absolute accuracy of the 9237 with the following formula.  (Noise for full bridge = 0.9 microV/V)

 

 

absolute accuracy = (gain error * reading) + (offset error * range) + (noise)

 

Doing the math we get

 

25mV/V  +/- 25.9 microV/V

 

 

In full bridge configuration:  V/Vex = -(Gauge Factor)* Strain


 

With a typical GF of ~2.0 we can say that the strain accuracy will be

 

12.5mm/m +/- 12.95 microm/m     

 

This is a bit more in the range you were hoping for.  Sorry about the confusion.  

 

If this will not work for you, we do sell PXI solutions that will be able to measure with more accurately.  Using similar math and full bridge configuration type with PXIe 4330 specs we get:

 

12.5mm/m +/- 3.48 microm/m

 

I believe this is the most precise and accurate bridge-based data aqcuisition card we sell.  It does require a PXI chasis/controller.

 

 

Please let me know if you have any more questions!

 

 

Sincerely,

 

 

 

Peter C.
Applications Engineer
National Instruments
0 Kudos
Message 6 of 16
(6,689 Views)

Peter,

 

Sorry for intrusion in this topis but I have just completed our setup for table shaking test for seismic qualification of a power equipment. For this we just bought additional 3x cDAQ-9188 and 20x NI-9237 for 80 SG channels (to our 72x PXI-4472). And we were sure we are getting at least 10 to 20 [uS] resolution!

 

Were we wrong? Can you translate those 12.5 mm/m in [uS] (micro strains). And furthermore, what do I read with "DAQmx Strain Channel" - Strain or MicroStrain?

 

BTW, we are going to use 120 Ohm gauges with GF=2, and internal excitation of 2.5V. Bridge is Quarter-2 type but with a 2nd passive SG for temperature compensation - so physically it is a Half bridge with RemoteSensing applied.

 

Thanks in advance,

0 Kudos
Message 7 of 16
(6,676 Views)

golubovski,

 

It is entirely possible to get [uS] resolution with the 9237 in full bridge configuration.  I accidentally reported the accuracy in half-bridge configuration at first (sorry about the confusion).   Strain is technically unitless, but is often reported as mm/mm or m/m.  So you can look at those accuracy specs as:

 

12.5mS +/- 12.95uS

 

 

Keep in mind, that accuracy is calculated based on both gain and offset error which are calculated as percentages of the range and read value.  I used the 25mV/V range for offset error and calculated the max gain error at the end of this range.  If you were to measure a point closer to the middle of the range, you would have less error.   Offset error and noise would remain the same, but the gain error would be smaller.

Peter C.
Applications Engineer
National Instruments
0 Kudos
Message 8 of 16
(6,671 Views)

Peter,

 

Just to clarify, if I use a half bridge will the accuracy be 

 

25mV/V  +/- 25.9 microV/V  ?

 

 

{  V/Vex = -(Gauge Factor)* (Strain) /2 

 

which means for a Gauge factor of 2 and using the same values of gain factor and off set error as you used;  accuracy =25mV/V  +/- 25.9 microV/V }

 

I ask this as in this page http://zone.ni.com/devzone/cda/tut/p/id/8909 it is reported as 25 mV/V ± 1.45 mV/V.  A half bridge tolerence is added in the calculations. Can you tell me what this is ?  }

 

Also, for a quarter bridge can I know what the noise level will be ? 

 

Thanks,

 

Sid

0 Kudos
Message 9 of 16
(6,667 Views)

Peter,

 

I use a HALF bridge (active and passive SG) reported to the NI-9237 as QUARTER-2 because the passive is off the model and just for temperature compensation.

 

1.

Do I read correctly that my precision is 12.500 [uS] or is it just the max reading for the 25mV/V span limits?!? I need the resolution step that can be sensed by the NI-9237. What would be the "center" resolution, can we see a 10 to 20 [uS] step?

 

2.

And what is the dimension  of  the output unit "strain" of the "DAQmx Strain Channel" - strain [S] or micro-strain [uS]? - Do I need to multiply the readings with 1.000.000 to get [uS]?

 

Thanks in advance,

0 Kudos
Message 10 of 16
(6,662 Views)