Eleven Hours in Owl Box Hell

I brought the owl box down around 6 PM on Sunday to remove a fox squirrel. This squirrel seemed familiar with the routine; with box open on the ground, I just stepped to one side, explained the situation, and the squirrel obligingly bolted out the other side. Simple enough.

It was a short time later that I made the terrible mistake. As followers of my screech owl cam' may remember, the illuminators in the current side camera compartment have been a bit flaky since I built it. That's because I selected the wrong way to mount and wire the infrared LEDs – using some screw terminals intended for mounting on printed circuit boards. I stuck the lead of an LED and a wire that led to a lead of the next LED in series into each hole in those two-hole terminal modules, tightened the screws good and tight a counted on that to keep every LED lead and wire properly connected. It worked just well enough to fool me into deploying the design, but not well enough to keep those LEDs shining all the time. They would dim, or stop working entirely, at random, and then start working again.

When I first observed the problem years ago, there were owls nesting the in the box, so there was nothing I could do about it. Since then, I've variously ignored the problem, convinced myself that it had gone as mysteriously as it had manifested, and occasionally even considered finding some way to fix the problem. Until this afternoon, I always snapped out of the "I must fix that" mindset before I could do anything foolish, like, say, tearing apart the most complex component of the nest box and rebuilding it in some better, but uncertain, new configuration. But this afternoon, with the box out of the tree, and having recently observed the complete failure of the side compartment's illuminators (they spontaneously fixed themselves a day or two later), my better judgement deserted me and I decided to take apart the side camera compartment, remove the front bezel/illuminator assembly and put an end to the bad connections once and for all.

Ouch.

It started simply enough, with the notion of stripping the insulation off of some wire-wrapping wire, wrapping the end of each LED lead, and dabbing a bit of solder on that in order to increase the diameter of the lead's end, and make it impossible (or, at least, a lot less likely) for it to fail to make good contact with the wire (whose ends were already tinned with solder) with which it shared its screw terminal. What I got was a lead that was large enough that it couldn't be crammed back into the terminal along with its wire – not with any assurance that they were both properly seated, anyway. And thus a simple, but tedious, plan exploded in my face and became eleven (going on twelve) hours of uninterrupted hell.

After finding that every store that might sell the components I needed to rewire the bezel from scratch had already closed for the day, a long and desperate search of my collection of electronic components, both new and salvageable, commenced. Eventually, I satisfied myself that I had just enough parts to do what needed to be done, and the tear-down of the bezel and its illuminator array began. It almost goes without saying that I was wrong about having all the parts I needed.

However, with all of the improvisation required to create a new way to mount the LEDs, having to improvise around the absence of a suitably large connector barrier strip fit right in with the rest of the problems. Like the mounts for the LEDs, the solution to the barrier strip problem was a strip of plywood. OK, the plywood strip provided no barriers, but spacing the screw terminals such that the wires that attached to them couldn't possibly come in contact with their neighbors was good enough.

The end result was every LED soldered to its own pair of wires, with each pair leading back to a central (improvised) barrier strip, that linked them in series, and then joined the ends of the series to the camera compartment's power supply. It meant a lot of extra wiring to replace what had been a nice, simple series wiring job, but all that wire also created a configuration that could be taken apart and laid out flat on a work bench for servicing, which wasn't quite the case with the original (unreliable) wiring scheme.

And the LED mounts? More strips of plywood. Specifically, a two inch strip screwed to the plywood back of the bezel in order to clamp the leads of each LED firmly in place. Fortunately, I'd recently re-stocked my supply of no. 6, ½" metal screws, which, oddly, I've always found ideal for this sort of woodworking, since their heads don't try to sink into the wood, unlike the heads of wood screws. That means that they project into the wood just as far as they say they will (½"), and no more. Their flat heads, drawn down tight onto the surface of the wood, also make them ideal stand-ins for screw terminals.

One Especially Nasty Lesson

I learned (or perhaps re-learned) something while performing continuity tests on the infrared LEDs as I soldered wires to their leads, and crimped screw connectors onto the ends of those wires. Inexplicably, those assemblies, all of which had passed their continuity checks when I built them, started failing later when I re-tested them both in series, and individually. This was a major contributor to the "hell" aspect of this night's experience. What I eventually discovered (or re-discovered – I now very vaguely recall maybe noticing this in the past) is that the infrared LEDs would pass their continuity tests if exposed to light, but fail those tests when they were in any shadow at all, or were, for instance, sitting mounted in their final assemblies, inverted on my workbench. Much serious unpleasantness was experienced before I figured that out.

The rewired bezel, disassembled (inverted) on my improvised workbench.

The rewired bezel, seen from the back, after reassembly.

The rewired and reassembled bezel, seen from the front.
The owls will never know the difference.

Now, I just need some owls. I heard one calling from my backyard early in the evening while I was beginning all of this work. Fortunately, I'd already returned the box (with a spare side wall module replacing the side camera compartment) to the tree, so if that calling was about enticing a mate to my nest box, there was, fortunately, a nest box present for the enticing.

Unfortunately, it's a bit late for my local screech owls to begin their nesting, if past years are anything to go by. So, it's entirely possible that they already have a nest somewhere nearby, and the calling was just routine communication between the pair.

If not, though, my newly improved nest box stands ready for them.