6

I need to reverse engineer this small board (7.5 x 28 mm) which serves an inductive proximity sensor. I suspect some of the components are dated/obsolete but the board works well. Ideally I would end up with an equivalent schematic.

Determining the various resistor values is easy enough but I lack the tools/knowledge/skill to identify the other components, measure capacitor values and trace the overall circuit.

I did some hunting but had no luck finding a service that would do this.

Anyone aware of such a service and/or options to decipher this little guy?

Much thanks for any and all advice.

UPDATE

A year later and after some prompting from Bruce Abbott, I finally battled through creating a schematic for this board.

While I believe I was able to identify all the components, I'm unsure which ones support the oscillator, detection circuit, and output circuit.

I would be enormously grateful for some expert input on which do what and whether the circuit below looks functional.

Much thanks!!

board schematic

coil connections

both sides of board ...

More useful view.
Better still would be a vertical view of each side with sensor bent at 90 degrees to the board so we can see under it.
Here the bottom side / right hand image has been flipped horizontally so the layout matches that of the other image.

enter image description here

puckhead
  • 65
  • 5
  • 2
    Google **three pin hall effect sensor**. – Janka Oct 20 '19 at 22:53
  • 1
    Where do you get them? I could almost do the reverse engineering from your photograph, but the boards aren't facing the camera squarely, and the sensor is blocking a lot of one side. Assuming you can afford to destroy one to get the information, the best way would be to carefully remove all of the devices and measure them, then take high-resolution photographs (or scans) of both sides of the depopulated board. It'll still require some detective work to track down the details of any active (semiconductor) devices. – Dave Tweed Oct 20 '19 at 23:33
  • Looks easy to trace the schematic out. Fat traces, very likely only two layers. Determining the function of the big 3-terminal component will be "the hard part." – vicatcu Oct 21 '19 at 01:25
  • You can probably buy functionally identical ones on eg AliExpress for far less than you can make them for. It looks as though the sensor ID is readable. What is the component code? – Russell McMahon Oct 21 '19 at 05:35
  • Have you removed the sense coil? – Spehro Pefhany Oct 21 '19 at 07:53
  • Thanks for the replies to all. Boards came from our friends over seas. The large 3-pin component is a 2SC2655 2A, 50V, NPN, Plastic Encapsulated Transistor. The sensing coil is not connected. – puckhead Oct 22 '19 at 01:43
  • 1
    We appreciate you taking the time to come back here and post the result you got! However, I feel it might be more appropriate to make your update into a separate answer instead of an edit to your question. It's perfectly acceptable here to post an answer to your own question; I'd upvote. – marcelm Oct 10 '20 at 19:38
  • 1
    Good point. I was also thinking a new question might have been a better move. I've got some great help from Bruce Abbott since adding the schematic and am close to completing the reverse engineering I set out to do a year ago. I'll add an answer with the proper schematic once I declare victory. Much thanks for the feedback! – puckhead Oct 11 '20 at 01:06

1 Answers1

14

I lack the tools/knowledge/skill to identify the other components, measure capacitor values and trace the overall circuit.

Now is a good time to learn!

Get a large sheet of paper, draw well spaced out symbols for all components and any test pads or board connections, and name each one (R1, C2 etc.). For any parts you don't recognize draw a generic box with the device's pins in their physical order. Obtain datasheets for any parts you can identify. Knowing what the device is supposed to do also helps, so gather any information you can such as user manuals, wiring diagrams and specification sheets.

Then use a multimeter on continuity test to 'buzz out' all the connections, and draw the wires between components. To find where a trace goes you can follow it by eye (a strong light shone through the PCB may help) or just rake over the connections until you hear the meter 'buzz'.

Next try to identify common connections such as ground and power, and redraw the circuit neatly with conventional layout (input on left, output on right, ground towards the bottom etc.). This could also be a good time to transfer the circuit to your favorite schematic capture program. Most designs are based on conventional circuit configurations and manufacturer's application notes. Identifying these patterns and laying out the circuit in the same way can help in understanding how it works, as well as providing a check that you have traced it correctly.

Now you can do on-board testing of components such as diodes and transistors to determine their orientation and type. Your multimeter's 'diode test' function can identify diodes and bipolar transistors (which read like two diodes back-to-back). ICs can often be identified once you have determined what the pins are connected to and their probable functions.

Capacitors may have to be unsoldered to read their exact values, but don't do this unless you really need to. Their likely values can often be guessed once you have figured out the rest of the circuit.

Most important thing is to treat the task like a crossword or jigsaw puzzle - enjoy yourself as you try to solve it! And as with any puzzle, if a part is too hard then do another bit and come back to it later.

Bruce Abbott
  • 55,540
  • 1
  • 47
  • 89
  • 4
    this is terrific answer – vicatcu Oct 21 '19 at 01:20
  • 1
    This is, indeed, a well crafted, articulate answer. Still, I'm actually looking for a service that would do exactly what Bruce describes here. I'm in a "more money than time" situation. That said, if there is no such service this is the path I will take. Thanks Bruce. – puckhead Oct 22 '19 at 01:48
  • Shopping questions are off topic here. Any competent electronics technician or experienced amateur should be able to do it. Try your local electronic repair shop or component store. Offer big bucks and I'm sure you will find someone to do it. – Bruce Abbott Oct 22 '19 at 06:08
  • Bruce, I'd love your thoughts on my update. Thanks! – puckhead Oct 05 '20 at 22:51
  • 1
    Your schematic is a bit of a mess. The most obvious errors are C3 which has a short across it, and Q2 whose Base-Emitter is connected directly across the power supply. LED1/R4/Q3 also looks wrong (cannot light the LED). Redraw the schematic so that +9V is at the top and local ground (through D1 ) is only along the bottom, and orient the transistors Collector-Base-Emitter from top to bottom (following the expected voltage drop). Then verify the circuit by checking continuity between the components at each node. – Bruce Abbott Oct 07 '20 at 06:41
  • Thanks very much Bruce. I will make those changes. I added an image and some additional blab to the original post to try and explain the connections at C3. I'm VERY grateful for your input and time. – puckhead Oct 08 '20 at 19:25
  • 1
    Q1 and the coil appear to be part of a [Hartley oscillator](https://www.circuitlab.com/circuit/utc376/hartley-oscillator-bjt/). The 'short' across C3 is actually one half of the coil. – Bruce Abbott Oct 08 '20 at 20:17
  • Thanks so much Bruce! I did a major rework and definitely discovered some errors. I've replaced the schematic drawing with the revision. The tip on the Hartley oscillator was enormously helpful. I believe there is something like a Schmidt trigger also involved here. Thoughts on the revision? Again, much thanks for your patience and help. You're the best!! – puckhead Oct 10 '20 at 18:34
  • 1
    Looking much better, but there are still a few obvious errors. The LED is backwards. C5 and C6 are shorted out. Once you figure where the capacitors actually go you may find other anomalies in this area (can't see any indication of a Schmitt trigger yet). – Bruce Abbott Oct 10 '20 at 19:43
  • Yep. C5 and C6 should have had R6 in parallel versus in series. I didn't uncover any other errors, but did move a few lines and components for clarity. I posted rev 4 of the schematic if you get a chance to take a gander. I would never have discovered all these errors without your help. I'm extremely grateful Bruce. – puckhead Oct 11 '20 at 00:59