I have a customer that needs to digitize the output of a strain gauge. Their power source can vary from 1.6 to 3.6V and is capable of supplying no more than 200 uA. Ideally, they want to use a differential input A/D and want a SPI output. Sample rate is super low. Maybe one sample per minute. 8 bit accuracy is acceptable. The design is high volume 100-200K units and very cost sensitive The Maxim MAX1391 is being considered. As an alternative, I offered the AD7468 + AD8237. I have a few questions:
Attached is my initial idea
A couple new thoughts. Perhaps the AD8236 is a better fit for the Instrumentation amp. It is much lower power. Regarding the reference, I looked for something cheaper than the Maxim part and considered the following:
Both are very low power
We ar looking in to this. We'll get back t you soon.
Thanks. Hopefully my question makes sense. The part I really hate is not having a reference solution. I'd like to avoid recommending an Intersil or TI part if possible.
No worries on keeping commenting. It always help because, it can be give a better understanding of the problem.
I found an ADC that is differential that is supplied from 2.5V to 5.5V. The AD7687 one of the candidates I found. The reference pin is external and can be connected with the power supply. The reference input voltage range from 0.5V to VDD+0.3. There is a circuit note for the AD7687 that you might want to check CN0225, http://www.analog.com/en/circuits-from-the-lab/CN0225/vc.html, Analog Front signal conditioner for industrial process control.
From what I understand, the stimulus or analog voltage measured on strain gauge are not dynamic. Usually for application that are not dynamic input voltage,sigma-delta ADCs are recommended.
I would recommend, you may check Sigma Delta ADC I found which is a single chip solution for strain gauge application. To mention are the AD7793 ( has on chip reference), AD7798/99 and AD7190.
Sorry to keep commenting. Perhaps I should have thought about this more before posting.
It occurs to me now that maybe I really don't need the reference. As long as I power everything off the same power rail, the measurements should scale, right? Also, if I have a 1.8V reference/ power supply for the A/D and IA, I won't be able to measure anything above 1.8V. Right now, I'm feeling pretty silly.
At this point, I'll just wait for your potential solutions. Hopefully, I've describe the problem in an adequate fashion.
Thanks Jonathan. I appreciate your input. The problem I have is that the solution must operate from a supply voltage of 1.8V to 3.6V. A solution that requires a 2.5V A/D will not fit this application.