Current measurement question

2000:1 is the range. If linear, you will need enough bits to resolve at least your 50uA -- 12 bits, say.

How many volts is your measuring circuit allowed to drop, in-line? At what potential relative to ground is it?

And I may not be able to help that much. I know a few things, but I'm just a hobbyist and probably can only offer a few thoughts.

Jon

Why bother with amplifiers and resistors? Just connect a swinging-needle meter.

Or, do you need multiple ranges, high precision, machine-readable outputs? It'll be difficult to handle 100 mA and 50 uA cases with a single shunt for current sensing, if they both need to be +/-

1% accuracy.

You'll need a 12 bit converter for 50uA resolution on a 100mA range. Make sure you take the effect of shunt resistor voltage drop into account with whatever it is you are measuring. You don't say if you need much accuracy or resolution down at 50uA?, that will determine if you need multiple ranges or not. A high side current shunt resistor with say a MAX4372 amplifier would be a standard solution. Or if your amp is battery isolated you can get away with the shunt anywhere, similar to my uCurrent adapter:

Dave.

:On May 4, 5:06 pm, panfilero wrote: :> hello, :>

:> I have to measure DC current through a range of about 100mA to 50uA, I :> need to make bi-directional current measurements.... so I'm planning :> on getting an instrumentation op-amp and a shunt resistor... but I've :> never done this before, I was wondering if anyone had any experience :> with this could give me some tips on how to approach it... I think the :> range may give me difficulty, I'm going to feed the output of the :> instrumentation amplifier to an analog to digital converter that can :> take -10V to 10V inputs... :>

:> thanks : :Why bother with amplifiers and resistors? Just connect a :swinging-needle meter.

I assume you mean a "centre-zero meter".

: :Or, do you need multiple ranges, high precision, machine-readable :outputs? It'll be difficult to handle 100 mA and 50 uA cases :with a single shunt for current sensing, if they both need to be +/- :1% :accuracy.

e:

:Why bother with amplifiers and resistors? Just connect a :swinging-needle meter.

That's one possibility, but polarity and shunt-value switches (or single switch, if it has enough complexity) can do the same job. It's not totally clear from the post what is needed, but the idea that this task required multiple regulated power supplies, A/D converters, high-performance instrumentation amplifiers... I just had to throw cold water on that.

Some of the best big systems I've used were cobbled together with CCTV cameras aimed at a panel of clearly labeled meters. That kind of system (Hey, it's Post-It-Note compatible!) just can't be beat.