How would that detect a short between two wires?

It would require turning them on one at a time.... wait you're right. Shit.

 

It would require turning them on one at a time.... wait you're right. Shit.

I must add though that shorts will be extremely unlikely in this case... we aren't testing field wiring that may have gotten beaten up, we're testing that the temps put the pins in the right holes. Testing for shorts would still be good though.

 

Yes I'm extremely interested in doing that. May be overkill for what I need, but I like overkill.

Holy crap I never thought about using a multiplexer like that. Mind blown.

 

It would require turning them on one at a time.... wait you're right. Shit.

Pulse the wires one pair at a time? I.e. no cap just cycling the input? sorry if this sounds stupid.

 

How would you use that to test for opens, shorts, or miswires?

As it sequences each wire in turn and powers the wire with LED and grounding resistor connected to each wire any out of sequence LEDs would indicate a cross, any constant ON (or multiple lit) LEDs would be a short and any constant OFF LEDs would be an open. Otherwise, you would get "walking" LEDs in sequence end to end to indicate GOOD. It would sit there and cycle over and over until turned off. As long as the LEDs are lighting up in sequence the cable is GOOD!

 

I'm thinking this will do nicely:

 

That is essentially what the LM3914N-1 and LM3915N-1 do without having to push a button to cycle through each one at a time. Chips made famous by the KITT car on Knight Rider back in the 80s.

 

Personally, I like the Arduino Mega solution. Not only could you program to test for continuity, but it could also tell is something is crosswired and tell you which pins are crosswired. Simple and inexpensive. Plus the test could run faster than you could plug in the cables....

 

Here is another thought, if you want a more robust piece of equipment without the need to solder.

Buy a cheap PLC (AutomationDirect makes great low cost PLCs LINK) with 24 outputs and 24 inputs. Create pigtails that mate to each end of your cable, permanently terminated to the PLC I/O terminals. Write a simple program to turn the outputs on one at a time and read the inputs. You should see your corresponding input turn on and no other inputs. Quickly cycle through all 24. This tests for breaks, shorts, and crossed terminations. At the end of the program, turn on a red or green lamp to indicate pass or fail (or get fancy, add a small HMI that displays the error).


Your upfront cost might be $300, but using a PLC is very robust and you will have it working much faster than learning Arduino or using discrete components and soldering. I assume with 800 cables, you are selling them, so this would just be a business expense to validate the quality of your product.

We do this at work all the time for 100+ pin military connectors.

 

This tests for breaks, shorts, and crossed terminations.

So the PLC can be programmed to do that automatically?
If so, that sounds like a winner.

 

So the PLC can be programmed to do that automatically?
If so, that sounds like a winner.

Yup, the way I've done it before is to write a program with many steps (one for each connection). You turn on one output and read all inputs (usually together as an integer) and compare them to an expected result. If the result matches, then you turn off all outputs and move to the next step. If the inputs ever fail to meet the expected result, you exit the routine and jump to a fault condition which turns on the red lamp and saves the step that failed.

Some quick pseudo code (we do this in ladder logic for simplicity at work):

Step = 0
FailedStep = 0

If StartButton == 1 then
    Step = 10
end if

If Step == 10 then
    SetOutput(0, ON)
    if Test1ExpectedResult == ReadInputs() then
        SetOutputsOff()
        Step = 20
    else
        SetOutputsOff()
        FailedStep = 10
        Step = 999 //This is the failure routine
    end if
end if

If Step == 20 ........ //Repeat the test from step 10 as needed for each connection

If Step == 240 then //Last test
    SetOutput(23, ON)
    if Test24ExpectedResult == ReadInputs() then
        SetOutputsOff()
        Step = 777 //This is the success routine
    else
        SetOutputsOff()
        FailedStep = 10
        Step = 999 //This is the failure routine
    end if
end if

if Step == 777
    SetOutput(GreenLight, ON)
    If StartButton == 1 Then
        SetOutput(GreenLight, OFF)
        Step = 10
    end if
end if

if Step == 999
    SetOutput(RedLight, ON)
    If StartButton == 1 Then
        SetOutput(RedLight, OFF)
        FailedStep = 0
        Step = 10
    end if
end if

EDIT: corrected hastily written code

 

The PLC would be better in terms of durability for sure, but I have to question it being faster to learn. There are millions of Arduino tutorials and examples out there. C is relatively easy to learn... even easier in the microcontroller world. Take the above pseudo code, combine things into loops, add a couple details, and change some punctuation and it's already done.

 

The PLC would be better in terms of durability for sure, but I have to question it being faster to learn. There are millions of Arduino tutorials and examples out there. C is relatively easy to learn... even easier in the microcontroller world. Take the above pseudo code, combine things into loops, add a couple details, and change some punctuation and it's already done.

I think both are easy to learn. I find that people who are not familiar with programming in general pick up on ladder logic more quickly than text languages. It was meant to be read by electricians and you can view exactly what your code is doing while it runs.

Either is perfectly acceptable so it is a discussion of what hardware you want to interface to and what the project budget is.

 

I find that people who are not familiar with programming in general pick up on ladder logic more quickly than text languages.

Good point. Personally have never worked with ladder logic, but looked at enough laying around at a previous job to kind of get the general idea. My concern was more in the variations between different PLC manufacturers, setup, and such details. Probably the same as say switching from Arduino to PIC or something else. Of course there are people familiar with just about anything on here and no matter what choice is made help is only a forum post away.