Testing connection resistance on a finished assembly

[OwenH]

Hi,

I have a 24v product that is essentially an intelligent power switch with 5 connectors on the output.  A microcontroller monitors voltage and load.  A PowerProfet load switch from infinion is used as the switch to the outputs.  All 5 outputs are tied together to the load switch.  The maximum current is 9 amps.

There are a number of difficult solder connections and crimp connections that I would like to have load tested after the PCB is assembled into its enclosure.

What I would like to do is build a test setup that can do the following:

- Calibrate the ADC and voltage divider resistors on the product being tested
- Calibrate the current sense function of the Profet power switch
- measure the voltage drop across the device under maximum load (To test for any issue with components on the PCB)
- measure the resistance of the 5 outputs individually (to test for a bad solder or crimp connection)

What I am proposing is to build a test device that can switch the input of my programable load between the various outputs of the device being tested.  I would basically build a MUX using a pair of MOSFETs for each output and then have a multichannel ADC that would measure the the voltage at the following places:

- Input to the device tested
- output of device before the 5 connectors
- output of each connector before the MOSFET MUX

This would be controlled by a microcontroller or maybe a raspberry pi.  The controller would be isolated from the MUX, ADC, and device being tested.

I realise I would need to calibrate out the resistance of all the wiring from the device being tested to the MUX and Programable load.

Questions I have:

-Does this seem like a reasonable approach or am I missing something conceptually? 

-Would It be advisable to measure both the positive and negative seperatly at each connection? Or can I assume that if there is an issue with the negative side of the connector it will show up when I measure the positive side? 

- What kind of protection might I need for the MUX, ADC, ect (maybe this is too broad of a question)?

Nothing is designed yet so I’m just trying to brainstorm possible ideas here.

Thanks!

[Terry Bites]

If you use a force-sense resistance (AKA 4 wire) measurement you dont need to calibrate out anything. You can build a super cheap adaptor for a standard DVM- eg https://www.maximintegrated.com/en/design/technical-documents/app-notes/1/106.html 

[tooki]

hello

read this doc: http://ygdes.com/LiIon/charge/AppNote106.pdf
it may be helpful for you

Literally the same thing someone posted 10 minutes earlier.