Characterizing parts (INA126 common mode input range)

[cellularmitosis]

Hey all,

I have a few INA126 instrumentation amplifiers on hand, and was wondering how useful they would be as high side current shunt monitors (at the time I ordered them, I didn't know that high side current shunt monitors existed .  I thought I'd post my experience here for other newbies.

I wanted to evaluate the part for the use-case of dropping no more than 1 volt across a current shunt at up to the supply rail (36V).

I rigged up a pot as a 36V divider so I could sweep Vin+ from 0V to 36V, and Vin- trailed Vin+ by two diode drops to simulate the current shunt drop.

I used three meters and monitored Vin+, the diode drop, and Vout from the INA126 (don't laugh at my 3rd meter, the harbor freight cheapy

I stepped the input sweep at 1V steps, and took a picture for each data point.  I then put the whole mess into a google spreadsheet ( https://docs.google.com/spreadsheet/ccc?key=0Aps_jTQV0tKndDJTc19UOE9mM2h1MmVfZ0puajl2Zmc&usp=sharing ), made a few calculations and then plotted the gain.

So, what we are looking for is where the gain diverges from its expected value of 5.  That will give us the range where this part is useful for this purpose.

In my case, it looks like I can safely use the INA126 from 2V to 29V (for Vin+).  The next step would be to take more data points near those edges to get more exact edges.

Experienced engineers, any feedback on glaringly obvious things I'm missing or getting wrong here?  This is my first time characterizing a part

Now, to hack together a micro to take these measurements automatically and spit out CSV...  mwuahahahahaha....

[cellularmitosis]

I have to say I was quite shocked at how linearly the harbor freight meter performed.  All of those middle values of gain calculated to exactly 4.93, and the harbor freight measurement was part of that calculation!  Amazed that thing was useful at all for $4.99.