Group Actions
Engagement
  • Author Author: shabaz
  • Date Created: Date Created
  • Views 3096 views
  • Likes 11 likes
  • Comments 28 comments
Related
Recommended

Brush and Large-Area Multi-Meter Probes for PCB Reverse-Engineering: A DIY Approach!

shabaz


This blog post contains some ideas for creating a probe to aid with reverse-engineering PCB layouts. You'd use this probe to temporarily replace one of the normal pointy probes on your digital multi-meter (DMM).

Often, you may want to run a continuity test across various parts of the circuit board in order to figure out which PCB trace goes where. This can get tedious with dense boards, or multi-layer boards where the trace is not feasible to follow by visual inspection alone.

Time can be saved by using a brush-like or pad-like probe with your multimeter. The probe would provide the ability to contact large areas of the circuit board quickly or simultaneously, and then the user could revert back to the usual pointy probe to identify the traces further.

image

I had some random parts lying around at home that I decided to use. I’ll describe it all below, but I wouldn't suggest it's the best way of doing this. You may have far better methods. Also, you may have other spare bits at home to implement the same thing in different ways. A 3D-printed version would be quite interesting too.

Hopefully, the 60-second video explains things:

To start, I cut some Perspex (acrylic) sheets, one piece approximately 1 inch square and another slightly larger, to experiment with making different-sized pads. I drilled holes to fit 10 mm diameter Magnetic Pogo Pin Connectors, which were a random purchase from AliExpress a while ago because they looked interesting.

For the probe wand, I used an eyeliner bottle; the brush tip can be pulled out with pliers, but since the handle is smooth and not easy to hold, I wrapped a sheet of silicone rubber around it to obtain a better grip.

The bottle also has a 10 mm hole drilled at one end, and the stalk end has a 3 mm hole drilled for passing the probe wire through.

image

The assembly of the probe wand is self-explanatory from the photos:

image

Here’s the finished result of the gluing. Loctite “All Plastics” superglue worked well with the Perspex/acrylic. For the wand, I used 90-second fast-set epoxy glue. Take care while gluing to avoid it spreading onto the magnetic connector mating or soldering surfaces!

image

Next, I cut some EVA foam sheet (it is also known as craft foam) to the same dimensions as the plastic sheet.

I had some fine-woven copper mesh sheet. I cut it into a length that would wrap around the EVA foam like a hoop. I also cut some sticky take to apply to the EVA foam at 90 degrees to the copper mesh hoop.

The idea is to tape the EVA foam to the plastic sheet, with the ends of the copper mesh wedged between the foam and the plastic sheet. The mesh makes contact with the connector, and doesn't need to be soldered. It would be easy to disassemble and replace if the copper mesh ever gets damaged, by just peeling off the tape.

image

Here's the finished result:

image

View from the other side, showing the sticky tape attached to the plastic:

image

I also decided to create a brush. For that, I obtained some finely stranded wire. I put it in ferrules and crimped them, then snipped off the plastic, leaving just the metal part. See here to learn about ferrules. 

image

I soldered all of them together by first aligning them in a clamp/vise.

image

Then, I directly soldered the magnetic connector to it and then covered it all with black PolyDoh (moldable plastic that is low-cost and extremely useful; there are many engineering uses for PolyDoh!). You can heat it up to soften it, and then press it into place by hand, and then use a metal object to flatten it into a better shape. The wire bristles were straightened and then given a haircut.

image

They appear to work!

image

When using such probes, some precautions need to be taken:

  • The board must be unpowered
  • Ensure capacitors are discharged, and batteries are removed prior to use
  • After use, ensure the board is free from accidental wire strand remnants!

I hope the blog was useful or that it can spark further ideas. If you have ideas/suggestions or reverse-engineering stories, it would be great to hear them!

Thanks for reading.

Parents

  • Just seen that this sort of thing used to exist (it's not manufactured any more), called a 'short finder'.

    It's basically a probe and all the electronics of a continuity tester all collapsed into one item. Nice that it had the pointy probe too, so the user can flip it over. Plus a cap for the brush.

    image

  • in reply to shabaz

    There's some errors in that circuit that was pasted from the PDF doc. Two of the op-amps have got their inputs flipped around incorrectly. I decided to redraw it in KiCad, so that it can be constructed. It is broadly identical, except the resistor values are different in places just to simplify things. I split the potential divider into two separate ones, to make it a bit easier to experiment with the values.

    The way the circuit works is:

    U1A forms a (approx. 50mV) voltage source, then U1B and U1C are a couple of non-inverting amplifiers. U1D works like a comparator. If the probe input is shorted (i.e. very low resistance), then the comparator output goes low, which will trigger the 555 monostable circuit (U2). I think Q1 is merely there to stretch the monostable pulse as long as the probes are shorted. The output of the monostable is used to pull the 555 stable circuit (U3) out of reset, to generate a tone.

    image

    Any errors anyone can spot? It would be nice to build this circuit, but I wanted to see if there are any glaring errors first. 

    Second question: any simple cheap enclosure anyone can recommend? It needs to be able to accommodate 2 x AA batteries (or 2 x AAA), a switch, and two 4mm banana sockets, as well as the circuit of course.

  • in reply to shabaz

    > Two of the op-amps have got their inputs flipped around incorrectly.

    Are you sure about that? Looks like that turns U1C from  a "logic gate (schmitt-trigger) with positive feedback" into a non-inverting amp with negative feedback?

Comment Children
  • in reply to Jan Cumps

    There's a good chance I could have misinterpreted the original (the black monochrome) schematic intent. It's just guesswork from me how they intended it to work.

    I couldn't see how the third op-amp (labeled U5.3 in that schematic, with C inside the triangle) fitted inverting or non-inverting pattern, so I assumed there was a mistake,because flipping it makes it become a non-inverting amp. And if that is done then there is the knock-on effect that the fourth op-amp also need flipping of it's inputs, because the 555 triggers on the negative-going edge, i.e. the fourth op-amp would need to have a low output when there's a short.

    It's tempting to start from scratch and not rely on the circuit at all. It could be done with just one op-amp and one comparator I reckon. But now I've entered it into KiCad, it may be worth giving it a shot anyway. I found a crazy-cheap op-amp for this, MCP6009, it's just 9 pence per op-amp, in sets of 4 inside a SOIC-14.

  • in reply to shabaz

    It is always difficult to determine if the schematic is correctly drawn in the first place, but trying to read a blurry schematic is another can of worms.

    I used an image processing program to shift the intensity, such that I can read the original schematic a bit better.  From that image here are the differences that I see with your new schematic (minus the resistor changes you made for the change in battery.

    1. U1-C inputs are swapped.  Hard to say which is correct.  U1-C is acting as a comparator, so it all depends on what state you want to the output to go high on. As drawn in the original, the output is high if the probe input is below the bias point. As drawn in your schematic it goes high when the signal is above the bias point.  The 1M resistor seems to be a hysteresis control.  I am not sure what U1-D is doing, as it appears to just track the output of U1-C.
    2. Your C4 is 10nF, where the original does not have a cap.
    3. Your C5 is 10nF, where the original (C7) is 100nF.
    4. Your C7 is 10nF, where the original (C8) is 22nF.

    Good luck!

  • in reply to genebren

    Hi Gene,

    Thank you for this! Regarding your point 1, and Jan's comment, it's dawning on me that I've probably indeed got it wrong. It does make sense that U1C could be a comparator, and the connections that were confusing me, were probably the hysteresis as you say.

    In that case, U1D is probably a red herring; it's not required, but they had an op-amp to use up, so they used it to perform the inversion that they needed, because the 555 needs a negative trigger. They could have just used a NPN BJT for the inversion for instance, but since they had the spare op-amp, they used that. 

    In summary, the circuit in monochrome could be accurate, with the benefit that with the hysteresis, it would not produce a scratchy sound if the resistance varied a bit. With the way I have it in the KiCad layout, the sound could have been scratchy. I'll revert those op-amp connections back to the monochrome circuit layout. I had thought the monostable would have got rid of scratchiness, but possibly with experience in actually creating and testing the circuit, maybe the original authors found that both hysteresis and a monostable worked best.

    I've found a suitable enclosure, and will start laying the PCB... I think it will be mostly complete by the end of this evening.

  • in reply to shabaz

    I have to say, the 'Pack and Move' feature in KiCad is brilliant. It's a massive time-saver. 

    It allows the user to automatically segregate components into sub-circuits. For instance, in the screenshot below, the three sections from left-to-right are the op-amp sub-circuit, the monostable sub-circuit, and the tone generator sub-circuit.

    Very easy to use.. simply highlight on the schematic any sub-circuit area, and then move to the PCB layout, and press the 'P' key to clump those components together.

    image