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

On Fri, September 12, 2025 2:04 pm, Justin Richards wrote:
> Nothing is set in stone.
>
>
> However, we are on the road for an additional 5 weeks and I have a small
> selection of parts that includes a shunt, LM324, esp8266, thru hole vero
> board etc and like to think I am resourceful.  There maybe electronic
> stores along the way with a limited range of components. (JAYCAR and
> Altronics)
>
>
> While investigating current sensing solutions I discovered there are many
>  purpose built devices but acquiring them without a fixed address is
> challenging and they are SMD.
>
> Curious if the diode solution to lift the reference is viable as the
> challenge of using opamps without a split rail seems to come up often and
> LTspice seems to suggest it is.  I can just make it and find out, I just
> dont have the luxury of breadboard, Bench PSU, Scope etc so I was
> interested to know if the solution has merit or has been used in the
> wild. I have not found any such examples.
>
>
> Jusin


Because the LM324 uses PNP input transistors, the input voltage range
includes ground. That should eliminate the need for adding a bias for a
unidirectional current sensor. But, I believe you are trying to detect
current in either direction, so the output has to be biased up so it can
go down when the current is in the opposite direction. A typical
differential amplifier has a resistor to ground on the non-inverting input
of the op amp. You've connected this to the top of two series diodes to
provide a reference voltage. This resistor is 100k in your original
circuit. I wonder if you could replace the 100k with a 200k to ground and
a 200k to Vcc. This would bias stuff up half way, use fewer parts, and
probably be more stable (not depending on diode voltages which are pretty
temperature dependent). The junction of the 200k resistors would have a
Thevenin voltage of Vcc/2 and a Thevenin resitance of 100k.

Good luck!

Harold




-- 
Not sent from an iPhone.