Fixed it!


If you compare this photo to the ones in yesterday’s post you’ll notice three important things.

  1. There’s a blue wire running between R5 and pin 2 of U1
  2. There’s a scratch on the PCB just above capacitor C2
  3. The LED is not on!

I’d already made a new revision of the PCB with the fixes for the problem that was keeping the LED on all the time regardless of whether the loco is present, but I really didn’t want to wait another 3 weeks to test the fix.  Fortunately the required changes are simple enough I was able to “yellow wire” the board (using a blue wire).  Sure enough, it now works, though I have some sensitivity adjustments to test out.

“Yellow wiring” is a term of art, at least at my day job, though a quick Google search did not turn up any outside references.  It originated, I am told, at IBM, where engineers would use bits of literally yellow wire to make repairs or changes to circuit boards in order to fix problems.  Supposedly they had a whole color code of wire colors that would be used to designate whether a particular fix was a “bug” fix, or a design change, or a repair, or whatever.  I’m not really sure of the details — perhaps an ex-IBM’er can chime in with more of the story.

Anyway, by the time I arrived on the scene, the term “Yellow Wire” had evolved to be in common use either as a noun, meaning the repair or modification itself, or as a verb meaning “to fix or modify a PCB”.  We do often use literally yellow wire, but it is increasingly common to use whatever wire color is at hand, or in the case of complex changes, to use different color wires where needed to help with human understanding of the wires’ purposes (red for power, black for ground, etc).  The term has been expanded to mean the whole fix, not just the wire itself. One will often hear hall talk or discussions in meetings about having to yellow wire this or that board, or how this or that yellow wire was easy or hard, or worked or failed.

In this particular case, all I needed to do was cut the trace routing the (wrong) signal to Pin 2 of U1, and re-route the correct signal from R5 over to Pin 2.  All done and tested (successfully!) in about 10 minutes.

Now I am much more confident that the new, updated PCBs I just ordered will work well!

This entry was posted in Uncategorized and tagged , , , . Bookmark the permalink.

One Response to Fixed it!

  1. Brandon says:

    Any updates?

Leave a Reply