ADC for strain gauge measurement

Hi Amir,

I do not have any idea about your hardware set-up. May be, providing a little clearer picture can help you get more answers.

When working with uV range of inputs, the result can significantly change based on the hardware setup (especially if it's a prototype board or self-etched PCB).

You may have to consider the effect of "thermocouple" when working on uV range. When two metals are in contact, the change in temperature can change the voltage as well, hence making an overall change in the output. You may have to introduce a dummy component network (to create another "thermocouple") to negate this effect.

My experience in strain gauge is very limited, but a small suggestion:

Try to use a Wheatstone bridge based circuit along with high-gain (Gain=1000) instrumentation amplifier, if not yet used. Use a OpAmp based subtractor circuit to subtract the base volt (at 1K Ohm) if necessary. If you're using subtract-circuit, change in each microstrain can be measured by the ADC.

Reference for ADC is also as important as other things. REF102 will be a good choice for reference. It gives fixed 10V reference, but with a very small error (many other fixed voltage reference ICs are available such as REF5050, etc,). You may use a divider with high-precision (0.1%) resistors to lower this to create any intermediate reference voltage. (Variable voltage references are available, but I found use of REF102 with a trimmer is easy for set-up).

Regards,

Binoy Raphael.

Aamir,

It think it would be best to show some sort of schematic of what you intend to do. I do have a couple of comments though.

First, I think that using a current source to drive the strain gauge and a series reference resistor is ok. In general any noise of the current source would be masked because the measurement is ratiometric (so that the same noise is seen by the reference as in the strain gauge). However, I would warn you that I don't think you'll get to that noise performance. The noise of the reference resistor may still be a factor though.

If you look at the ADS1248 datasheet, table 1, you'll see that in PGA=32 and at the lowest data rate, the noise is about  70nV rms, and about 310nV for peak to peak. That's probably a bit noisier than you want. I'm not sure if there's much that you can get as a better part, although I believe that the ADS1232/ADS1234 might have marginally better noise performance (although not necessarily in the gain that you want and it doesn't have the current sources). While in your case, you want to generate a reference based on a current source and resistor making a ratiometric measurement, it still gives a bound on your noise which will only be worse than that level.

Joseph Wu

Hi, I am planning to use Wheatstone bridge.

So change in voltage is very small so I need some opamp. My circuit is powered from +5V.

1. Can TI suggest some very low power opamp (my application is battery powered), which can amplify the signal for adc. Since I am using wheatstone bridge, so I need a opamp which can amplify the signal from ground rail.

2. next i have to add 2.5V to add it bring it in common mode range of ADS1248 so can Ti have some reference circuit on how to these two steps.

3. What to avoid negative supply since it will add to more circuitry i.e more power