How to detect shorts, Need to develop a short detector

[Rlima]

   Hi all...

      I'm trying to figure out how to detect shorts on a  circuit board  (15 to 30 pins on a straight line of a circuit board)

      something similar to this    ********************* this is hand solder and we got some short between pins.  I want

      to implement a short detector using LabVIEW and DAQ.  I keep wondering but so far  I can't figure out how to do it?

     I'll really apreciate your help..

     Thanls in adavance...

     Reynaldo Lima V.

     If you have any questions let me know...

[johnsold]

Reynaldo,

What kind of circuitry is on the board? Analog, digital, mixed? In other words: How much voltage can you safely apply to the board without damaging components on the board? What is the minimum impedance that is considered to be an acceptable board? What is the maximum impedance which is considered a short? Are shorts between non-adjacent pins to be checked? Are any adjacent pins intentionally shorted (as in multiple pins connected to the same point for higher current)? How are you connecting to the board for testing purposes? Connector or some type of probes?

Lynn

[tbob]

You would need a fixture to hold the board, and pogo pins or similar to contact the pins on your board.  The wires from your pogo pins would have to go to a switch matrix.  Then you need a DMM (GPIB controlled).  Then your software would use the switch matrix to connect one point of the DMM to one of the pins, and the other DMM lead to another pin.  Look for an open (ohms).  Disconnect the second DMM lead and connect it to the next pin.  Look for open.  Keep on this way until all pins are tested against all other pins. 

NI makes PXI based switch matrixes and a PXI DMM, but a cheaper solution is to get a USB switch matrix.  Measurement Computing makes one, TAMS makes another.  All of this will add up to be a bit costly.  So it only makes sense to do this if you are going to test a lot of boards.

[tbd]

Hi Rlima,

What do you mean by ""

---

"15 to 30 pins on a straight line"

---

Is this a header or IC socket?

Can you post a digital-picture of the PCB?

Must the test-points be probed, or might it be possible to "reach" (electrically-speaking) the test-points via a "header"/ edge-connector?

Can you afford the time and money ( probably $10K US - or more) to purchase switch-matricies and to have a custom "bed-of-nails" manufactured?)

How do you do the test now?

Cheers!

[Rlima]

     Hi to all...

     Thanks for you comments,  now I have some ideas. So far I know I have to nest to hold the PCB then have a pogo pins for each test point.  I'm starting to like the idea of the DMM to measure impedance. but I still need a switch matrix.   I need to keep this proyect on a low cost,

     I test the complete board on a In-Circuit tester with  500 nodes. But the issue to implement this short detector is due to a hand solder board that is added to the main board, I can't  test the board again because is brake apart won't fit the In circuit tester. 

     Does any one have a model of switch matrix so I can take a look at.

     Maybe can use an analog demux (demuxtiplexer)   to handle the pogo pins to the DMM.

     is a  25 pins (thru hole) I cant reach them with pogo pins from the bottom side (this is not a issue) I'm home right now, don't have a picture available by now,   Monday can send it.

Best Regards.                                                                                                                                                        Reynaldo Lima V                                                                                                                                                          

[LV_Pro]

As previously asked, are there components connected to these pins, are they connected such that measuring between pins may be reading across these components? Some components may provide low inital impedance (capacitors) others may conduct if your DMM's measurement voltage is sufficiently high (diode junctions).
   Many years ago we tested (at a previous employer) for VCC to GND shorts on unpopulated boards by connecting them to a high current source! Turn the lights down low and you could see where the shorts  were  (they usually were cleared by this process). Of course these were wire-wrapped boards (actually used in production parts for RADAR systems) and the shorts usually were a result of the plating peeling back on the wire wrap pins when the machine inserted them, but it made for a light show sometimes!  Not recommended for printed circuit boards, particularly the multilayer ones!

[Dennis_Knutson]

Also keep in mind that if you go with the DMM and a switch, you have a lot of tests to perform for thorough coverage. For 4 pins, you have 6 tests (1 to 2, 1 to 3, 1 to 4, 2 to 3, 2 to 4, 3 to 4). If my calculator is still working after trying to finish my tax retrun is correct, you have 435 discrete tests to perform for 30 pins. You'll want a switch matrix and dmm with pretty fast settlings times and data transfer rates. A PXI or pc based dmm is going to be better than GPIB and much better than a serial instrument. PXI is going to be pretty expensive.

Depending on what is on your board, YOU MAY be able to use a much cheaper digital I/O solution. If you drive and sense each pin at the same time, you can quickly detect all possible shorts. For example, if you drive pin 1 high and pin 2 low and there is a short between them, then pin 1 will fail when you sense the logic level at pin 1. It would require 6 patterns using this technique (FFFF0000, FF00FF00, F0F0F0F0, CCCCCCCC, AAAAAAAA, 55555555). This is basically what the simple bare board/cable testers do. I think they are actually a little fancier in that the inputs are comparators with an adjustable threshold. With passive components on the board, you don't hve to worry much about backdriving like is done with the big, expensive ICT testers. There, the pulse are kept to a very short duration and low current. A custom solution is also possible if you want to design and lay out your own test board. USB interfaces from companies like FTDI are pretty cheap and easy to intergrate.

[tbd]

That's a great story Putnam! - reminded me of looking under a hood, at night, to find bad spark-plug wires.

[Joe_Henderson]

Hey Reynaldo

How is the short detector project coming along ? I'm curious to know what you have tried for your project. Did you use a DMM card to measure the shorts? Was it fast enough for your purposes?

George