Why is this so F****** hard to get a clear answer on the subject?

I have a Wheatstone bridge that i want to interface with a microcontroller, but cant figure out what way to do it! Maybe you can help?!?

As I see it I have following options (Please correct me if I need anything (components, filters, eg.) or am simplifying something too much)

1)
Connect a 16’ísh bit ADC that can read +-10mv (aprox. Output from bridge when force is applied). To power I need to configure a circuit with a voltageregulator to power the bridge – (to ensure that the fluctuations in the 10V supply wont show in my output, I connect the 10Vsupply to the ADC’s Vref (I am unsure about this???!!).

Then I should have a digital signal I can use………

2)
Same as above but I have seen people make their own opamp prior to the ADC (I am unsure why this is done instead of buying an ADC with internal opamp. Does anyone know why?


3)
The last possibility is that I have a device I can connect to 24V and to the bridge. It will give me an output of +-10V. The thing is that I cant read that with my Arduino. So that leaves me with the option of setting up an other op-amp to bump the +-10V signal to 0-5V or getting a ADC that can read +-10V directly from the device.

I find option 1) or preferable to option 3) but I am unsure.


Thanks, I appreciate it.

 

Use an Instrumentation Amp (true differential with high CMRR), not an OpAmp, to sense the bridge. Set the gain to get the voltage swing you need into the ADC. The programmable gains in the ADCs I've used are single ended so will amplify the noise as well as the signal which is the reason for the IA.

You can use single ended or differential output IAs according to what your ADC requires. I've been OK with single ended with proper layouts etc. YMMV.

Excite the bridge with Vref (buffered if you have to) so that the bridge voltage will always track the ADC reference. To avoid errors due to the buffer, pick the ADC Vref right off the top of the bridge (after the buffer). This approach also reduces the need for a tightly regulated Vref/bridge exciter and negates any problems with voltage drop in the wires driving the bridge. Since the ADC is ratiometric it won't care about minor changes in Vref since they will be reflected in the output of the bridge as well. This approach requires a calibration step in software unless you carefully pick the excitation voltage to fit the range of the ADC. Since calibration is usually a requirement anyway, its not a big deal.

Keep the low end of the bridge at the same potential as the low input of the ADC (point-connection).

Prepare to do some averaging or other filtering of the ADC readings.