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").



links

decay & ruin
Biosphere II
Chernobyl
dead malls
Detroit
Irving housing

got that wrong
Paleofuture.com

appropriate tech
Arduino μcontrollers
Backwoods Home
Fractal antenna

fun social media stuff


Like asecular.com
(nobody does!)

Like my brownhouse:
   technology trouble followed by technology trouble
Tuesday, August 24 2021
I took the dogs for a walk this morning earlier than they normally go. We went through the abandoned go-cart track and back along the plateau west of the Farm Road. I don't remember seeing so many red efts so late in the summer as there are this year. At first I thought this was an Adirondack thing, but they're here as well. It probably has something to do with all the rain. Many of the big ones are less red than a dull greenish-orange, though there are also quite a few of the bright red-orange little guys, some only an inch in length (something I never remember seeing around here before and which I first saw a few weeks back in the Adirondacks).

[REDACTED]

Something has been amiss with my solar controller for the past few weeks, usually as evidenced by the circulator pump (which pushes hydronic fluid through the solar panel) being on after dark. For reasons that should be obvious, it's only supposed to run when it's sunny. Initially I replaced the ULN2003 chip, which allows five volt signals to control the higher-power signals needed to drive big relays, and that seemed to fix things. But now I'm having problems again, and they seem to be related to a worn out zone valve. I have a number of Honeywell MZV-series valves in operation because I initially got compatible valves (from a company that must've since been acquired by Honeywell) for cheap and have since been locked into using replacement parts. This is unfortunate, as these valves don't seem to last as long as the more common V8043-type Honeywell valves. So today I ordered a conventional single Honeywell V8043 replacement for one of the bad MZV-series valves, which I will have to drain pipes and desolder to remove. While that's happening, I'm seeing other problems with my solar controller. Yesterday it must've been splashed by some of the water I accidentally blasted out of an accidentally left-open valve, so now some lines of pixels on its LCD are failing. Furthermore, I could no longer use the IR remote I keep down there to change configurations on the homemade menu system I developed ten years ago. That last problem was probably the worst issue, since it meant the only configuration I could do was over a serial connection, which is a major pain to set up and hard to do entirely in the boiler room, the place where the results can immediately be see. And often even difficult problems can be fixed simply by making configuration changes (such as setting the realtime clock to the correct time and date). I couldn't tell if the IR menuing problems were because of an issue with the old universal remote (which had been accidentally drenched) or the solar controller box itself (which had mostly avoided the deluge). To figure out what was going on, I fired up the Arduino IDE and loaded the ten-year-old "slaverman2" sketch that handles IR communication and the LCD display. The sketch is so old that it has a .pde instead of an .ino extension and its syntax is so archaic that it must be compiled in an Arduino IDE prior to version 1.0. I didn't have such a version of the IDE set up, so I went through the bother of migrating the syntax of the old sketch (all of which concerned the Wire I2C library). Once that was done, I could look to see how I'd handed IR reception in the old code. I'd apparently noted the raw integer values of every key I was interested in, put them two arrays (one for digits and the other for commands such as cursor controls) and then had a function translate these long arbitrary codes into simple numbers usable by the sketch to carry out its actions. I noted that the old universal remote was not generating the right codes (perhaps because its configuration had been accidentally altered when it was drenched), so it seemed like all I had to do was get another remote, note the codes for its keys, and then substitute them into the array holding the menu system's configuration. All of this took awhile, and then I had to upload the compiled sketch to a loose Atmega328 microcontroller. The new menu controller seemed to drive the LCD successfully, but it was completely unresponsive to signals from the new IR remote, which made no sense at all. So I decided to shelve the issue for the evening and work on something more straightforward before going to bed.
The other day I'd taken delivery of a brand new micro AT minitower to house the newest version of Wolverine, the computer now based on an AMD Ryzen 3600. This evening I thought I could get an easy triumph by moving the Wolverine motherboard from the big ugly case it was in into this new case. Boy was I wrong! Initially, the power supply and the motherboard fit just fine (though things were a bit tight), but when I went to screw the motherboard down into the standoffs, the screws went in only a little ways and then stopped. It was natural to keep turning them, but that was a mistake, because all this did was strip the threads in the case's sheet metal around the standoffs. I ended up having to remove the motherboard and all the standoffs, since now the standoffs were flapping around loose in their holes. I then figured out that I was dealing with two problems. The first was the thin sheet metal used in this inexpensive $50 Rosewill case; it provided little for the threads of a standoff to bite into. The other was that the standoffs (normally made of brass, though these were probably made of something cheaper) had holes for a smaller screw than the one normally used for securing a motherboard. I've been installing motherboards into cases since 1994, and the screw for this purpose has always been the same: the #6-32 UNC screw, the same kind used to secure a hard drive, expansion card, or a power supply. In this case, though, the screws that worked with the provided standoffs were finer-threaded M3 metric screws, the kind used for securing optical drives. This was why my attempt to secure the motherboard had failed so badly on my first attempt. To remedy the situation, I had to secure each of the six standoffs in a vise and then force a M3 screw into its hole, thereby fixing its threads. As for the problem with attaching the standoffs to the case, I still had to deal with stripped-out holes in too-thin sheet metal. To fix this, I put superglue in each of the bad holes and slowly tightened the standoffs down under low torque until I felt mild resistance. When all of this was done, I could install the motherboard. Happily, Wolverine started up just fine in its cheap new home.
One other thing to say about cases and motherboards these days is that there's way too much superfluous LED lighting in them. I get what's happening; enthusiasts think of their computers as hot rods, and they want the engine to look visually interesting. This is why there's even a fucking a window (made of cheap plastic, of course) on my new microtower case. Otherwise, I might install fancy lighted components and nobody would be able to see them, For me, though, a computer is a black box that does amazing things, but I don't want it to be a spectacle. Ideally, it takes up zero volume, which is something computers are asymptoticaly approaching. I'm annoyed by, for example, all the fans with LEDs. They're using electricity that could do other things to create visual pollution. I'm willing to pay more for a fan that has no LEDs.


For linking purposes this article's URL is:
http://asecular.com/blog.php?210824

feedback
previous | next