current excited wheatstone bridge

[mercurial]

I came across a post on this forum where the consensus was that driving current through a wheatstone bridge was not the preferred way for signal conditioning of strain gauges.

Some users cited linearity and some cited temperature related errors.

So if I use a full bridge topology (all 4 resistors in the bridge are strain gauges) and somehow mechanically couple the strain gauges such that they are equally affected by temperature and drive the bridge with a 1mA current.

Would it still be the non-preferred mode of measuring a strain, what kind of errors does one expect and why?

I did a small simulation test and results and attached.

[voltsandjolts]

With semiconductor based strain sensors, where bridge resistance increases noticeably with temperature, constant current drive can help maintain temperature stable bridge output. It might be beneficial with metal strain gauges if you are operating over a wide temperature range and you're not making a ratiometric measurement of bridge output.

I recommend you use ratiometric measurement and then bridge supply becomes less critical and constant voltage is fine.

https://www.eevblog.com/forum/projects/load-cell-(wheatstone-bridge)-excitation-voltage-hurdles/msg3134330/#msg3134330

[mercurial]

BTW i was thinking of using current excitation  of bridge(metal strain gauge) + ratiometric to remove drift errors from current source. I was just wondering what are the cons of current excited bridge vs voltage excited bridge. Are they any? or whether I interpreted the responses in the other posts wrongly.

[voltsandjolts]

Creating a constant current source may require extra effort and components, and if using a ratiometric measurement I don't see any advantage over voltage drive.

[Kleinstein]

Constant current drive could simplify the wiring / connectors to the bridge, as one does not have to take into account the drop the wires to the bridge, which can be an issue for long wires and high accuracy target combined. It may still be easier to have extra sense wires for the drive than making ratiometric current drive.

[David Hess]

Practical load cells have a temperature sensitive modulus compensation resistor in series with the excitation which would be defeated by using constant current drive unless sense wires are used, and if sense wires are used, then why bother with constant current drive?

The presence of the modulus compensation resistor can be checked by comparing the resistance across the excitation with the resistance across the signal wires.