Manual versus Automated PCB Testing

Started by jvthorsley, February 05, 2021, 03:01:25 PM

Previous topic - Next topic


I repair & test a range of circuit boards used on agricultural machinery.
The two main PCB's have a lot of plugs & sockets on the boards.
These connectors all have the connecting pins very close together, and in the humid, dusty environment shorts can develop and with lots of  available AMPS on a 12 V.D.C. system.   Burnt, carbonized circuit board results shorting out CAN BUS lines to power lines.
The resistance of these shorts can be very high, 400 K Ohm +.
And as all the components (Relays, Flasher modules, & Fuses) used on these PCB's are plugin I made an adapter to test using my ProTrack  (or 3200S would do the same) for shorts after removing all the plugin components.
The adaptor I made using Stripboard consists of two 64-way IDC connectors.
One to connect to my ProTrack Scanner, and the other to whatever plug I wanted to test using ribbon cable.
I also put sixty four 47 K Ohm resistors on the strip-board, each one connected to each of the 64 pins of the IDC connector. The other end of the resistors are all connected together and connected to the Tracker COM socket.
Each pin on the two IDC connectors are connected pin 1 to pin, and pin 2 to pin and so on.
Set the Tracker range to 20 V, 100 kOhm, 1 KHz and very tight tolerance, say 1 to 6 %, and this can see short circuits of nearly 400 K Ohm.
But I was to find that some circuit boards came bouncing back to me because they had a resistance of 1 M Ohm or higher between connector pins, that should be open circuit as nothing else was connected to them.
That is until all the plugin components are installed.
So as the more recent versions of Workstation support DMM's, and various other bits of test equipment, time to spend more hard earnt cash on a decent DMM.
So it came down to probing one pin at a time or sell the house and buy a Huntron Prober and work+ live in a bicycle shed!!
But using the above adaptor, this time with 10 M Ohm resistors, and an Agilent 34410A DMMI was able to easily see if there was a resistance between connector pins of the order of 200 M Ohm.
I disconnected the Ch. A BNC plug on the back of my Scanner from the ProTrack, and connected the Scanner Ch. A output to my DMM.
And Tracker COM connection to common on my adaptor stripboard. **NOTE CHANGE. Only necessary to connect the COM connection on my adaptor to the DMM Lo socket. Connection to the COM on the Tracker not needed here.**
I then, and note here no PC connection to my Tracker, ran the Scanner from the Trackers menu and was able to very quickly run through all the pins on a plug or socket to test for shorts between pins using the DMM in the very high G Ohm range.
Note this method can not be run on a 3200S that has the Relay Matrix & Tracker as one unit so not possible to access the Relays output.
I ran this first with Workstation that just gives you a nice big readout on the PC Monitor.
But Keysight who have taken over Agilent test gear have a program called Test Flow in there Benchvue DMM control program that can I am told gives you warning if the measurements you are scanning are outside of the set parameters.
But with another PC running workstation controlling the Tracker,  it might be possible to control the Scanner and run the test a bit like running a Prober.
Have yet to run this test procedure a bit more to see how successful it proves to be.

The attached photos show how a high resistance short of 250 M Ohms can develop into a low resistance short below 100 Ohms and start burning the PCB material, turning it into carbon that makes an even lower resistance.
And photo of the High Resistance Adaptor with 10 M Ohm installed.
The missing twenty 10 M Ohm resistors awaiting customs clearance due to BREXIT!!

Using the the Scanner on my ProTrack menu and turning the Ecoder Knob I can check 37 pins in approximately 37 - 39 seconds.


Great write-up, James!
Those connectors look pretty nasty and I can see how high resistance shorts could be an issue.
You have an interesting workaround to use the ProTrack Scanner. I would not have thought of that method. :)

Also, the Workstation software can be set to Fail if a measured DMM value is outside of a nominal value you set. This is set in the Range by selecting the Comparison type as either "Nominal" or "Nominal %". Then you indicate the Nominal value you want in the Nominal field and the Tolerance (by value if the Comparison is set for Nominal OR by percentage if it is set to Nominal %).
The most important troubleshooting tool is your eyes.



The most important troubleshooting tool is your eyes.