2
\$\begingroup\$

I am a recent graduate working on a side project that involves creating a test jig/method for a large PCB backplate. Unfortunately, the original layout and diagrams are no longer available, so I need to create a schematic and map out the connections before developing a permanent testing method. There is no strict deadline for this task, and I would appreciate assistance with the following questions:

  1. Is there an efficient way to map the connections on the backplate so that I can recreate the PCB layout and document it?
  2. What would be the best method for testing the backplate, and how can I implement a quick and permanent testing solution for it?

Thank you in advance for any help you can provide.

Photos of the backplane:

enter image description here

enter image description here

\$\endgroup\$
4
  • \$\begingroup\$ Are you trying to test out new unpopulated PCBs, or existing populated PCBs? If new PCBs, then you should be able to get hold of the production files (Gerbers and netlists) from the manufacturers, which is what are typically used for "flying probe" PCB testing. \$\endgroup\$
    – colintd
    Commented Sep 26, 2023 at 9:48
  • \$\begingroup\$ It is quite an old PCB. I will check with the manufacturer if they can supply Gerber and Netlists for it. Unfortunately, the original schematics are missing, and we do not possess the means to test it in-house, although they do test it. I am looking for a way to create something that can plug into it and check all the connections. \$\endgroup\$
    – mw599
    Commented Sep 26, 2023 at 10:32
  • 1
    \$\begingroup\$ Hello and welcome. "Large" is very much a subjective term and depends on one's experience, it leaves us guessing. A measurement or even an estimate goes a long way; a photo would be great too. \$\endgroup\$
    – jonathanjo
    Commented Sep 26, 2023 at 10:57
  • \$\begingroup\$ Please add more details, including photos, overview of unit function, test volume, reliability requirements, etc. \$\endgroup\$
    – colintd
    Commented Sep 26, 2023 at 11:13

1 Answer 1

Reset to default
3
\$\begingroup\$

In terms of regenerating a schematic, the netlist and parts list is the key. The manufacturer should be able to supply these if you are still having PCBs built. The combination can give you a full schematic with work, and there are a number of tools (such as this one from Elgris) which can help the process.

You can at a push regenerate from the gerbers+partslist, but this is much more of a pain.

If you manufacturers are a good quality house, they should already be doing flying probe testing of the bare PCBs (using the netlist) to guarantee connectivity and lack of shorts. They will do a better job of this than you can do post addition of components.

In general whole board testing focuses on functional testing of the assembled board. You will need to put together a jig (or test cards + harness if it is a mid-plane being testing), which allows you to verify the correct operation at all the external interfaces.

I've designed and built a number of these over the years, and the goal for production use is normally to minimize the manual operations (insertion, removal, etc) and operator judgement requirements, whilst maximizing the extent of functional testing.

Exactly what is best for you will depend on the unit function, size, complexity, cost of missed faults and required end-quality.

Very hard to go much further without considerably more information on your specific case.

\$\endgroup\$
5
  • \$\begingroup\$ Hi, I am linking the pictures of the PCB in question Image 1: i.imgur.com/cO8HJR7.jpeg Image 2: i.imgur.com/Pm38P12.jpeg \$\endgroup\$
    – mw599
    Commented Sep 26, 2023 at 12:44
  • 1
    \$\begingroup\$ These photos should be added to question (via an edit) not in comments. To help out I've made this change. \$\endgroup\$
    – colintd
    Commented Sep 26, 2023 at 12:53
  • \$\begingroup\$ You have a very significant backplane there, with lots to test! \$\endgroup\$
    – colintd
    Commented Sep 26, 2023 at 12:55
  • \$\begingroup\$ Thank you for editing the pictures into the message, I will keep it in mind for the future. Yes the PCB is quite significant :D \$\endgroup\$
    – mw599
    Commented Sep 26, 2023 at 13:29
  • \$\begingroup\$ Worth thinking about if you need testing to specifically locate issues, or just identify an issue exists. For instance, I've used connectors in the past which have small pcbs "hair pinning" one set of signal paths to the next. With careful design this allows a single signal path to be established across dozens of backplane traces. You can then inject one signal and measure its return. The upside is most connectors are passive and you need a single test channel. The downside is that if there is a problem you have to track down manually where the signal path is broken. \$\endgroup\$
    – colintd
    Commented Sep 26, 2023 at 14:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.