1
\$\begingroup\$

I am trying to make a low power device (based of RSL10 bluetooth SoC) that would sample a strain gauge @1kHz and transmit the measurements. The current drawn by the running SoC is about 1.5mA, while the current for the gauge excitation is about 3mA (1kOhm @ 3VDC) assembled in a 4-gauge Wheatstone bridge.

I wish to keep the gauge current in sub-mA current range, but they seem to not make stain gauges of higher resistance. I think the ones I have are 350Ohm each with some additional resistors to make it to a 1kOhm.

  1. I could add more series resistors outside the bridge, but I think this will make the assemble less sensitive to changes.
  2. Another option is using a buck converter to step up the excitation voltage, but this procedure would not help decreasing the power consumption.

Does anyone have an idea on how to implement the stain gauge assembly that would not require such (relatively) high current?

\$\endgroup\$
3
  • \$\begingroup\$ increase gain x3 and add 2xR in series. \$\endgroup\$ May 13 '19 at 21:13
  • \$\begingroup\$ How often do you need to sample the gauge? \$\endgroup\$ May 13 '19 at 21:52
  • \$\begingroup\$ @VladimirCravero 1kHz continuous \$\endgroup\$
    – Nazar
    May 14 '19 at 12:31
1
\$\begingroup\$

If you don't need to take measurements all the time you could put a P-FET on the voltage going to the bridge and turn it on only when you need it.

The FET just needs to have low threshold voltage (Vgsth), also watch out for the Rds on resistance of the FET that may be relatively high.

schematic

simulate this circuit – Schematic created using CircuitLab

For example wake up every second (or five, or whatever you need), turn on the FET, take a measurement, turn off FET. This should be able to happen in much less than a second, so your average current becomes low. Or, turn on the FET, take your readings for 10 (or however many) seconds, turn FET off. The specifics depend on your application.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ Given the top of the 4-R bridge will vary in voltage, because the FET's Ron will vary, you should use that TOP voltage as the VREF to the ADC. \$\endgroup\$ May 13 '19 at 23:59
  • \$\begingroup\$ I like the idea. However, I need continuous sampling at 1kHz - and no good way of syncing ADC sampling. Also, the datasheet does not explicitly say which voltage the ADC uses as the reference. Need to figure out. \$\endgroup\$
    – Nazar
    May 14 '19 at 12:35
0
\$\begingroup\$

You can supply the bridge with a lower voltage regulated supply and provide more amplification. Naturally, noise and drift will be a bigger percentage of your signal (SNR will decrease).

I suggest using a "zero-drift" amplifier and paying close attention to thermal EMFs if you care a lot about low frequency noise.

Since the source is low impedance you can also consider a very low noise bipolar amplifier if you only care about higher frequency noise. Some are available with noise in the 1nV/\$\sqrt{\text {Hz}}\$ range, for example the LT1028.

You will have to run the numbers and select an appropriate voltage and amplifier for your particular set of trade-offs. There is no lower limit on the excitation voltage, they're just resistors, only what is imposed by your measurement requirements.

\$\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.