Your leaking thatched hut during the restoration of a pre-Enlightenment state.


Hello, my name is Judas Gutenberg and this is my blaag (pronounced as you would the vomit noise "hyroop-bleuach").


decay & ruin
Biosphere II
dead malls
Irving housing

got that wrong

appropriate tech
Arduino μcontrollers
Backwoods Home
Fractal antenna

fun social media stuff

(nobody does!)

Like my brownhouse:
   actually found the human remains
Wednesday, October 18 2006
The reward for enduring the warm front was a beautiful mid-October day, with temperatures reaching up to around 70 degrees Fahrenheit.
I had a housecall up near Saugerties where I found myself trying to do something I'd never done before: getting a Macintosh to print on a printer attached to a PC that was accessible across the network. There seem to be two ways to make this happen, but both were so flawed that the goal eluded me. To get started, I needed to do a Google search, since the "Print & Fax" System Preference hides network printers behind a button that must be clicked while holding down "option" on the keyboard. (Does that sound like a sound user interface decision?) Then, if for some reason you either don't give a username and password or make a mistake and give the wrong ones, there is no way to fix your mistake: you are stuck with "Access_Denied" errors for all the rest of eternity. That's with the "Add a Windows printer" option. If you go the "Add a Samba Printer" option route, you're expected to know the name of the "print queue" on the host computer. I defy anyone to give me the name of the "print queue" on their computers! That information may be somewhere, but it's not in any of the most likely places on any Windows machine in front of which I've ever had the misfortune of sitting. All this is just another way of saying: adding a Macintosh to your household network is not exactly the Ralph Nader-electing, sunglass-wearing & Radiohead-blasting cakewalk Apple would have you believe. (Did Apple make the video on that page crap out on my PC deliberately? And no, I won't be upgrading my Quicktime!)
I had the dogs with me on this excursion, and at some point I drove to 9W and bought some groceries at the Hannaford. As usual, I took the dogs for a run in the field just north of the supermarket. There's a brand new construction site at the far end of the field, with all sorts of digging and fresh new towers made of stacked and mortared concrete blocks crenelated (temporarily, here in the First World) with whips of rebar. Behind this construction site is a small patch of neglected real estate that has reverted to forest. The floor of this forest is full of human artifacts: here an old refrigerator, there a pile of empty 22 ounce beer cans, and over there a pair of cheap bicycles, stripped nearly bare of parts. I half expected to stumble upon a skeletonized human corpse, but it was Eleanor who found the human remains. And when she did so she naturally rolled in them. I didn't notice until I came back to the car and realized it had assumed the fragrance of my 8th grade English teacher's breath. My suspicion is that one of the workers at the construction site had to use the little boy's room but decided to use the woods instead. Sometimes when Eleanor rolls in something unspeakable I can let it slide, but not this time. The moment we got home I gave her a thorough shampooing, and in so doing I managed to get just enough of that halitosis smell on my hands to give me unexpected flashbacks for the rest of the day.

Yesterday when I was doing that boiler rewiring I was also looking out for an easy way to send a low-voltage signal to my Arduino board telling it whether or not the boiler was firing. With that information included in the data stream going up to Woodchuck (my main computer), I could log it and get a sense of how much oil I would be burning. Unfortunately, though, there didn't seem to be any low-voltage circuitry in the boiler aquastat circuitry (unless you include the thermostat signal, which now should be completely decoupled from boiler firing). So I decided to make a compact little AC-to-low-voltage-DC converter using just a resistor divider, a bridge rectifier, and a low-value capacitor. I used high value resistors (one mega-ohm and 100 kilo-ohms) to divide the voltage by ten, then rectified it and smoothed out the peaks of the waveform with the capacitor. In theory this should work fine, and in fact it did work fine in the laboratory. (I love being able to say that!) It worked so fine that I submerged the riot of diodes, resistors, and solder connections in a vodka bottlecap full of epoxy. I did all of that last night. Today I went to test this new contraption down in the boiler room. I hooked up its low voltage section to my Arduino board and the high voltage section to a 120 volt power cord. I plugged it in to see whether the Arduino would change a 0 in its data stream to a 1. Nothing happened, so I quickly determined that some of the improvised connections were failing. I plugged it in again. Bzzzap! Somehow a circuit breaker had been tripped and the lights were out.
Now I'm not sure what actually happened. The gizmo I'd built wasn't smoking and neither was the Arduino board. The Arduino was still actually working, cycling through its four LEDs as it produced varying analog levels to represent temperatures on the VU meter and considered whether or not the solar panel was collecting heat. But something was wrong; it turned out that some electronics had been damaged by the jolt. The ULN2003, which changes digital levels to the kind suitable for activating relays, was partially damaged and needed to be replaced. This wasn't surprising; it was the only chip that my ill-fated resistor-divider gizmo had been attached to directly. More annoying than that was the failure of the USB link to the laboratory. Evidently the electrical spike had jumped from the ULN2003, passed somehow through Arduino board without damaging it, and then killed off one or both of the "active nodes" on either end of that long USB cable. It had been a very inexpensive Chinese USB cable; evidently industrial hardiness was one of the features cut from its design to save on expense. I won't actually have to restring that cable; once I get a replacement I only have to replace the active nodes at either end. These take the form of six inch long mini-cables. (I'd love to be able to buy these separately, but this isn't possible.)
In thinking about how this mini-catastrophe happened, I suspect it had something to do with the fact that in non-isolated AC electrical systems one has to pay careful attention to the phase of the current. If the "hot" wire entering my circuit was the one going to the 100 kilo-ohm resistor, not the 1 meg resistor, that would place the divided voltage output of my gizmo at, no, not the intended 12 volts with respect to ground, but 108 volts with respect to ground. If I then ground the most negative of the output lines (as I did), well, you can imagine that the other output line isn't the sort of thing you want running up your urethra. My guess is that some sort of thermal runaway thing happened inside the electronic components of my gizmo and this led to a short big enough to blow the circuit breaker. These sorts of accidents are thankfully rare with me, but if you do as much tinkering as I do they're bound to happen occasionally.

For linking purposes this article's URL is:

previous | next