In recent months, I’ve been lending a helping hand to a few people who’ve shown interest in electronics. I’m not an EE or an electronics expert of any sort either by training or in my work, but the basics are extremely valuable skills to have under one’s belt, even at the hobbyist level. My assistance has taken the form of giving short demonstrations, watching them do stuff and giving pointers and suggestions, and helping them troubleshoot small projects. Hand soldering has figured prominently in these activities, along with soldering pitfalls and basic touch-up/rework/repair. Reels of soldering braid have been distributed, I’ve handed over gun-style vacuum desoldering tools, and helped people get oxidized tips back to usable condition with tip restoring compound.
A lot of the projects have been kit-based and Sparkfun’s Electronic Dice Kit has been very useful. After I bought out what I think now must have been the last of their stock, Sparkfun retired the product. It’s a shame, since it’s quite a nice kit and the different types of components used, the fact that one needs to work on both sides of the PCB, and the need to think ahead a bit about the sequence of work (some components have to be mounted over snipped-off leads of others) make it an excellent little project for speeding up the process of getting good-enough at soldering.
In the process of up-skilling, before they get to their first fully-functioning fully-assembled kit, people can tend to churn through them, producing non-working finished products. Usually, the problem is obvious and easy to fix and, after explaining the problem and providing a fresh copy of the same kit, I often pocketed the inoperable/buggy one with the intention of using it for troubleshooting practice.
Recently, I dumped the non-working dice kits I’d accumulated onto my home benchtop and worked through them one by one, fixing all but one. The problem here seemed to be that the solder pad for the pin on the kit’s coin-cell battery holder had been damaged in the course of desoldering the battery holder.
First, I cut the narrow end off of a small solder pad (aka solder lug), one shaped like the silhouette of a tennis racket, so that it would be short enough that it’s large hole would fit symmetrically over the pin hole for the positive-terminal pin of the battery holder (which was attached to the other side of the PCB) and so that the stub of its short end would extend beneath the raised copper trace that had formerly run to the solder pad.I didn’t see any other exposed traces on the PCB beneath the footprint of the piece of solder tab. Nevertheless, just in case, I cut a teeny tiny bit of high-temp tape and placed it there as insulation. That’s the odd-looking bit of milky-colored mesh visible in some of the under the pictures included in the composite of the repair process above. The white thing tapering to a blunt end extending downwards from the top is one jaw of a pair of normally-closed, ceramic-jawed tweezers that I used to hold the work in place during soldering. The solder pads for the positive and negative pins on the battery holder are marked with red-circle plus signs and blue-circle minus signs.
The kit still didn’t work, though. All of the LEDs were oriented correctly, the diode was installed pointing the right way, etc. The kit’s little brain, a PIC 12F675 microcontroller (the “1” in the marking on the IC is difficult to make out bc a brief touch of my soldering iron slightly melted the plastic nearby) is getting the correct voltage difference across pin #1 and pin #8, etc. The LEDs, when directly and independently powered from their snipped-off ends, work fine.
Actually, very briefly, it did work. After the repair I’ve described and once I’d checked things out and couldn’t figure out why it might not be working, it occurred to me that the battery might have run low. When I popped it out of the kit’s battery holder, it was able to drive a mangled through-hole LED I had lying around just fine but I was in a what-the-heck mood. I clipped the leads from a benchtop power supply to the correct metal bits of the battery terminal, turned it on, and slowly adjusted the voltage upwards from zero.
It worked (some LEDs lit and tapping the piezo disc caused the pattern of lit LEDs to change). I unclipped the power supply leads, turned the voltage and current back down to zero, and switched the power supply off. There was no clear explanation for why it should’ve worked with the power supply and not the apparently-fine coin cell, but it had.
From my box of batteries, I plucked a brand-new CR2032, popped it out of its blister pack, and inserted it into the kit’s battery holder. Nada.
I removed that battery and hooked the kit up to the power supply again and did precisely as I’d done earlier. Zilch.
At this point, AFAICT, all of the passive parts in the kit (i.e. everything except the microcontroller brainbox) are fine. Maybe I somehow zapped the PIC 12F675. I don’t have any fresh, unused PIC 12F675s, or whatever hardware/software I’d need to program them, or (for that matter) the code used in this kit. I could futz around with my oscilloscope to see if the microcontroller is “talking” to the LEDs. Alternatively, it might be possible to cannibalize one of the other fixed-up, working kits and try swapping the suspect brainbox in this kit and the known-working brainbox in a functioning kit after desoldering them brainbox. The idea of continuing the quest does appeal to me, but I’m leaning towards writing this off as the one that got away.