Re: [EE]: Bi-Directional 20A current shunt Using Single supply OPAMP

Harold,

Reducing the parts count as you described was an interesting exploration.
It simulates nicely.

However, with the selection of resistors on hand (listed below) I was
unable to exactly match the feedback resistor (output to inverting input
(Vin-)) with Rth of the divider.  If not matched then the output when
measuring fwd/rev current is no longer symmetrical.

I now have a circuit I am happy to attempt to build.  Except, there are
three unused modules in the LM324 so obviously feature creep such as
measuring and conditioning the 12.8V rail by mapping 15V-10V -> 3.3V-0V.

Thanks.
justin

>
> I assume you are using a pair of 1k in series to form the voltage divider
> and then driving the 100k to the non-inverting input from the voltage
> divider.

Yes

> You could reduce the parts count by just using a pair of 200k
> resistors, one to +3V3, the other to ground, and the junction going to the
> non-inverting input. This would have the same Thevenin voltage and
> resistance. There MIGHT be an issue with noise on the 3.3V line being sent
> into the amplifier through the voltage divider. This could be avoided
> using the circuit you are apparently using and adding a capacitor to
> ground at the voltage divider output. You could increase the resistances
> used in the voltage divider some (as long as they are small compared to
> 100k). But, if the 3V3 is quiet enough, I'd just use the two 200k
> resistors.
>

Collection of resistors on hand roughly :-
10x 1R
10x 4R7
10x 10R
10x 47R
10x 100R
10x 220R
10x 330R
10x 470R
10x 680R
30x 1K
10x 4K7
10x 6K8
30x 10K
10x 22K
10x 39K
20x 47K
10x 68K
10x 82K
20x 100K
10x 330K
10x 470K
10x 680K
10x 1M
10x 4M7