0
\$\begingroup\$

I have two strain gauges on a 1 m length wind turbine blade which will rotate at 800 RPM. The amplifier and DAQ system will rotate with the blades. I'll have relatively noise environment and the turbine generator will be close. I am using strain gauge for first time and was wondering if any body has any suggestion to reduce noise in my measurements.

For example, What is the best solution to connect the strain gauge shield wires to PCB (amplifier) in order to reduce noise? I was wondering to use screw terminal, but probably there is a better way.

\$\endgroup\$
4
  • \$\begingroup\$ You need STP cable for starters, and differential op-amps to scale the readings. Are you going to broadcast or record the data? \$\endgroup\$
    – user105652
    Mar 6 '18 at 1:33
  • \$\begingroup\$ What is the bandwidth of the signal you want to measure and what are the noise sources? \$\endgroup\$
    – Andy aka
    Mar 6 '18 at 12:01
  • \$\begingroup\$ @Sparky256 Yes I am using AD8221 amplifier. I'll record data in a USB flash drive \$\endgroup\$
    – Arash
    Mar 6 '18 at 16:41
  • \$\begingroup\$ @Andyaka The strain gauge output is less than 5 mV, then I'll amplify it 500 times to measure with an analog input channel which can measure 0-5 V (I have 2.5V level shift in signal and 2.5 V shows the zero strain voltage). The noise can be from rotating blades, generator, screw terminals on amplifier PCB, and so on \$\endgroup\$
    – Arash
    Mar 6 '18 at 16:47
1
\$\begingroup\$

One approach (I saw this used in a nuclear-testing setup) is to drive the strain gauge bridge with 20,000Hz sin (something much faster than the generator risetimes, so the correlation is very small), then use narrow-band amplifiers to boost the 20,000Hz, then perform a synchronous-demodulation.

===================

The nuclear testing channel used 5,000,000 Hz drive to the bridge, so the response could be proportionally faster. From what I've read over the decades, such a test generates the energy in "2 shakes of a lamb's tail" or 20 nanoseconds. After that, its all about the energy going wherever the energy wants to go.

\$\endgroup\$
1
\$\begingroup\$

I have been looking at something similar. I don't have a very noisy environment but I do have the sensor 'far' away from the MCU (3 meters). I plan to put an ADC with a digital interface next to the sensor. I found one which has differential inputs (of course) but also has a differential reference output. That should eliminate noise on the wires.

enter image description here

But in your case I don't know how much noise the generator injects direct into the sensor itself. I am sure @analogsystemsrf solution is better but it is rather a lot more difficult to implement.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.