Better is a matter of opinion.
For me, I like to reduce the chance of error stack-up so here is how I do it:
* Wire in the CA tap to your controller
* Hook up a power source to your controller - your battery is fine but a constant voltage supply is will supply more consistent readings.
* Power on the controller and confirm that everything is working.
* Set the controller board on the table. Be extra, extra, extra careful not to set the board on any of the screws or random pieces of metal.
At this point, what I do is a little Ghetto, but you will get the idea.
Basically I short the +V buss through an ammeter and a load to ground through the shunt.
* Hook an alligator clip to the +V buss
* Run that clip to your load. In my case it is a big oil heater. There are 2 switches and 3 modes.
The "hottest" mode gives me 10 ohms and 2KW power handling. I attach to the load by simply clipping the aligator lead right on to the plug.
* Hook another alligator lead to the other prong of the plug (or other end of the load) and run that into the positive input of your ammeter.
* Run the negative input of your ammeter to a banana style terminator (or something that can be used to "poke")
At this point I fire up my spreadsheet (downloaded here). The spreadsheet has a column for Control (ammeter reading) and Measured (CA measurement).
You enter into the spread sheet the starting value in the CA. For simplicity, lets say you set it to 1mOhm.
* Get everything ready. Turn on the Ammeter, open the spreadsheet (so you can write down the numbers before you forget), turn on the CA.
* Set the ammeter right next to the CA so that you can easily read both.
Here is the sketchy part:
* Poke the lead that is exiting the ammeter into the PHASE side of the current shunt. This will force the current to run through the shunt.
* Watch the readings on the Ammeter. As soon as they settle record the reading on the ammeter and the CA for the same moment in time.
This is where you want a constant supply voltage. If you have one the only variation you will see are heating affects of your cabling.
* Fill in the spreadsheet with the ammeter reading and the CA reading.
You will be given a %difference calculation and the spreadsheet then tells you what the new value is for the CA.
Lets say it was .865 mOhms.
* Enter the new CA value into the CA.
* "Poke" the probe on the phase side of the shunt again and take another measurement. These should be within a few percent.
* Enter the numbers in the spreadsheet and read off the new CA value.
* This value will be within a percent or two. Repeat if you have OCD.
The reason I like this method is because you read the current straight from the horses mouth.
Nothing gives me a warm fuzzy like seeing 10.65A on the ammeter and seeing 10.65A on the CA.
If you see this, there is no doubt that you have done everything correctly.
If, on the other hand you use a method that involves reading the voltage from the CA tap (or worse... please dont say reading the voltage right off the shunt
) there is an entire new set of errors introduced. The error of the voltage reading, the possible probing error (in the case where people read directly off the shunt instead of at the taps), etc.
I am just that kind of engineer though... Any time I can prove that something works directly I will always take that approach over an analytical approach. I know that mathmatically measuring the voltage at the CA tap and computing the resistance with Ohms law is just as valid... It just does not take into account any "funny business" inside the CA.
You could employ a hybrid of my method and Justin's method. The biggest problem with my method is that the larger battery voltage forces you to run a larger load.... like a MUCH larger load. The inductance of that load can come back to bite you, but that is another story.
Some small part of me also feels better about reading the current coming from applying 100V instead of the current coming from 12V.... I know there should be no difference but then I also know that all sorts of stray currents are running around in the circuit. At 100V there simply must be small offsets here and there that could affect things. I guess in my mind I am convinced that I am better approximating the real system response.
This is just my method though.
There are many methods that are equally valid.
Mine is actually much more dangerous
-methods
