Bridge Excitation Accuracy

[j777]

Hello,

I'm trying to decide how I will provide excitation voltage to the wheatstone bridge of a load cell. The output of the load cell will be amplified by an LT1167 In-Amp. How concerned should I be about the precision/accuracy of the excitation voltage? Since the In-Amp will reject common mode voltages (which should include any fluctuations caused by change in excitation voltage) would a standard +-5% linear regulator work without introducing and real error in the output?


Thanks

[Bystander]

Are you looking at bridge output as instrument output, or are you balancing the bridge and looking at the balance as instrument output?

[j777]

Bridge output as instrument output.

[Bystander]

Then, accuracy is a linear function of excitation voltage; output is proportional to excitation voltage. 5% in excitation is 5% in output. If you're going to use that as input to a control circuit and drive that to a "zero set-point," no sweat. If you're buying or selling bulk material by weight, you're screwed.

[j777]

Wouldn't any variations in excitation voltage be common mode since only the difference between IN+ and IN- from the load cell is amplified?

[Bystander]

What's the output of the cell under load at a zero excitation voltage? At 1 V? At 10? The bridge is a passive device. It's output is the difference between the voltages at mid-points on two voltage dividers. If by "common mode," you mean that the amp gain is a linear function of excitation voltage, you're gonna have to post the specs on the 1167. "In-amp" ain't something I've run into, but I ain't been in the game lately. I've been talking on the assumption that "In" is short for "instrumentation amp," a souped up op-amp. Without some input giving it "information" on excitation voltage, it has no way of knowing what scale factor to apply to gain, given a variable bridge output for fixed load.

[j777]

Thank you for making some sense of what I was saying. I wasn't thinking clearly but I understand what you are saying now. Your explanation was very helpful as I am definitely not an expert with this stuff.

[Bystander]

Welcome. I failed to point out that the usual use of "common mode rejection" refers to "stray" inputs that appear simultaneously at both inputs to an op-amp --- inductive pick-up of power line noise, or other environmental phenomena.