You could use a shunt but I have found that if the length of wire from
the switch/distribution panel is sufficient and the wire is not totally
overly thick you can get good results just measuring the drop across it.
Typically it is not unusual to have a drop of 500mV from the battery
to the load but this will include the cables to the distribution panel
so the wire you are voltage drop you are working with will be less. On
my boat the wire was so thin the drop was over 1V at the dis panel when
the fresh water pump kicked in. Copper wire is not be as thermally
stable as a proper shunt would be but copper is pretty good anyway
(within a few percent for a 10 degree C temp change). Your electronics
may not be this stable anyway. You will need to calibrate the wire using
a known load or ammeter/volt meter and will still need the differential
amplifier to boost to get your full 2.5V range.

Obviously you would be best using the negative side of fridge circuit or
you will need to bring the voltage down from 12ish volts to 2.5. As a
warning I have found with some data logger circuits that when the
voltages are close to 0 there can be a issues. Might be worth
considering when you test. If your logger power supply is tapped from
somewhere closer to the battery than the fridge you should at least
avoid getting a relative negative on the logger input. Negatives
voltages can be as bad as over voltage I have found.

Steve

Glenn Ashmore wrote:
I should know this but I just can't get my brain around it.

I have a data logger that will accept 4-20 mA or 0-2.5VDC. I want to track
the current going to the refrigeration compressor which I figure will be in
the range of 0 to 25 amps (to be safe for start ups).

What is the best way to accomplish this?