Ignore them: they're just flesh wounds [entries|archive|friends|userinfo]
sprockets, sockets, grommets & gaskets

[ userinfo | livejournal userinfo ]
[ archive | journal archive ]

(no subject) [Jul. 22nd, 2014|08:46 pm]
manintheboat was reading one of those glossy magazines and found an ad for a 7 acre island off the coast of Vancouver Island. (Well, technically, off the coast of Salt Spring Island.) We could afford it. Like, the whole island. But there are some issues, like it's not clear there's any source of drinking water. Or electricity.
But hey I'd finally get my floatplane license.
There may be issues with rising ocean levels, as well.
Still, fun to think about.
Link33 comments|Leave a comment

(no subject) [Jul. 21st, 2014|11:36 pm]
New throttle/accelerator cable in the car. In essence, I turned it around. Now the swaged end is at the carb, as Datsun intended.
I went to several bike stores to try to find the right cable. In each, I walked in and said "okay, what I need is a brake cable for a pre-1985 mountain bike, to use as the throttle cable for a Datsun engine in an old Triumph" because I do enjoy seeing people's eyes pop out a bit.
Well, nobody had one, until I got to the store just down the street, which sells Cannondales out front, and has an old British motorcycle garage in back. I talked to the pimply kid in front, and he's all ".... uh you better talk to the old motorcycle guy." Well, the old motorcycle guy was all "we just make our own custom cables" and whips out this enormous box full of cable ends, and gives me some instructions on soldering to stainless cable.
Although what I ended up doing, when the terminal he gave me didn't quite fit, was cutting my own out of brass, that pressed into the recess on the carb, with a transverse hole drilled for the bike cable, and then countersunk halfway for the swaged end of the cable.

Have I mentioned lately how dangerous it is to drill brass? That stuff catches like nobody's business, no matter how solidly I bolt the workpiece to the table. Slightly dull drillbits, or zero-twist drillbits, or even a handmade ream, is always a better idea than a nice new drillbit. (Handmade ream: take your drillbit, turn it around, and grind a 30 degree flat on it. Hone. Now you have a safe brass drill.)

The original throttle cable ended at the gas pedal, which has a piece of tubing welded to it, with a slot in the end, that the cable goes through. It doesn't come out of the slot because there's a ziptie around the end, pulled really tight. Or, there was. Now there's a nice little piece of aluminum with a hole bored through it the same diameter as the tubing, and two pinch bolts that capture the cable with a bit of steel bridging between them. I'll probably back it up with another pinch on the other side. Belt and suspenders.

Work assignment to build a heavy base to hold circuit boards flat for testing, when they have heavy cabling sticking straight up out of them. Yay, a chance to get someone to buy that chunk of 1.5cm thick plate steel from me. I suspect it will never, ever tip over. It may break someone's foot. I'm cutting a piece of acrylic to go over the top and prevent electrical shorts. I may even cnc out an outline of the circuit board so it snaps down into the acrylic, with a couple of toggle clamps to hold it in position.

Thinking wistfully about another restoration project.
1958 Morgan? http://bringatrailer.com/2014/07/21/bat-exclusive-1958-morgan-plus-4-bustleback/
Or 1985 Tatra? http://bringatrailer.com/2014/07/19/1985-tatra-t815-8x8-dakar-tribute/

Alas, I bet that Tatra doesn't even get 5 miles to the gallon. But you could drive it through a river.
Link30 comments|Leave a comment

(no subject) [Jul. 20th, 2014|09:59 pm]
Starter on. Boy was that a hassle. Alternator on. Boy was that a breeze. Transmission wiring (reverse light) and speedometer attached, clutch bled. I went to the local NAPA to get some 1/4" fuel line. They had none, at all. But I've been over there so many times and talked to them so much, the guy gave me, gratis, some 3/16" line, in the hopes it would work. It worked better than the 5/16", that's for sure. So, fuel line is connected. (I think. I have no idea which fitting on the fuel pump is input and which is output, but I'm making the assumption that the lower one is the input, knowing how fuel pumps work.) I did a massive amount of rerouting on the coolant hoses, to reduce it to something less insane, which also meant that I could use both orifices of the rather dirty-looking water pump outlet, one to handle the radiator and the other to handle the heater core inside the car. (This deletes a brass tee that was clamped into duty as a splitter.) The upper radiator hose is still awful: it comes out of the front of the head, pointing to the right, and goes into the radiator with a right-pointing inlet on the far left-hand side of the radiator, meaning I need a very tall J-shaped bit of tubing. I managed to saw a short J-shaped piece of tubing out of the middle of something the people at NAPA dragged out of a box, that would have worked for the previous setup, but it was too short for this, so reluctantly I lathe-cut some aluminum tubing to extend the J enough to reach to the radiator. Reinstalled the shifter lever, wired up the distributor cap, pulled out the electric transducer senders on the water and oil and replaced them with the mechanical ones that drive the original gauges in the dash. (I sort of hate doing that.) I lathe-cut a tophat-shaped bit for the throttle cable stop.

So now I'm down to: the throttle cable needs to be captured by the throttle bracket. It originally used a cable that terminated in a disc of lead crimped onto the cable. It is, in fact, identical to a bicycle shifter cable. Unfortunately for me, it is identical to a shifter cable from the 1970's, and once bicycles moved to aluminum rims, sticky pads, and higher-leverage calipers, they also moved to much thicker cable with larger terminating discs, and I don't have anything that old, so I have to jazz something up.

And, the distributor. Which has two wires coming out of the bottom. I can handle dizzies with no wires, and ones with one wire. But two wires? So, taking ilcylic's advice, I opened it up... and no points. Just a reluctor wheel and a hall effect sensor. So it's intended for an electronic ignition, which I don't have. But it does have a vacuum advance on the distributor, which means it was a very stupid electronic ignition system, and it may be possible to bodge something together to act like the missing brains. Or I just get very busy finishing the firmware for my Ford full electronic ignition system and skip the distributor entirely.

Also a lot of walking, a lot of pulling up cottonwood tree sprouts.
Link10 comments|Leave a comment

(no subject) [Jul. 19th, 2014|10:54 pm]
Engine is in on the new mounts, with the belleville washer stacks acting as the elastic connection to the frame. I ended up scooting the engine back about 2 cm from where the first engine was, which meant I also had to build a new transmission mount. That was okay, as the original one was made of a stack of pieces of rectangular steel, bolted together, some of them with wood hammered into the tubular insides so they wouldn't collapse. One thing I didn't count upon was that the original was so flexible, it contacted both the frame rails and the transmission mounting point flat, because it was distorted. My new one is roughly 200 times as stiff, maybe 1000 times, I don't really know, and since the one I pulled out was parallel, I made mine parallel too. Well, it's only touching along one edge. I don't think that's really a big problem, under the circumstances, so I'm just going with it.

New clutch flexible hose is installed. I'll bleed it tomorrow when it's light out.

Then, starter installed, alternator installed, both those wired in, radiator back in. That may be all I get done tomorrow, but with any luck I'll also manage to plug all the holes in the intake manifold (it had an absolute riot of spigots and hoses for emissions control, which I'll attempt to reimplement at some point but not right now) and then, in theory, the car will be able to move under its own power.

The difficult step is going to be figuring out how to hook up the exhaust. That's changed significantly. There's also a lot of work to be done, reinstalling the stuff inside the car, where I've removed the transmission tunnel and the dashboard support, and I'd like to improve some stuff there. However, I'm a lot closer than I was.
Link3 comments|Leave a comment

(no subject) [Jul. 18th, 2014|11:11 pm]
programming question: how to pass values between functions in c++Collapse )
Link24 comments|Leave a comment

(no subject) [Jul. 16th, 2014|10:47 pm]
My version of xubuntu is old enough that the standard package manager isn't updating automatically anymore (or at least I assume that's the reason, since it did work and now it's stopped) so I have to remind myself to apt-get update every couple of weeks. I should set up a script to do this.

Work is kind of depressing. I'm not engineering right now, in any sense of the term. I'm building neat optomechanical things, and programming, but neither of those advances my resume. The track I'm on right now is pretty documentation-heavy and almost entirely lacking in actual electron-wrangling.

But I'm making progress on the car. My fussy method for making engine mounts is slowly bearing fruit: I have the final one done for one side, and tack-welded up the trial one for the other side tonight. I'll finish it tomorrow and maybe have the engine actually on mounts rather than sitting on a piece of wood on the frame.

This AM I went to the dentist. I had an old filling that he disliked and a couple of years ago he replaced it. (He disliked it because it left a small gap between it and an adjacent tooth.) Well, he decided he disliked his replacement, too, so he convinced me to come in for a free replacement, which he did this morning. He replaced it like whoah and now my face hurts. Plus I suspect this, too, is going to be problematic, because now it's too tight. He said it would equilibrate within a couple of hours. It is not equilibrating anywhere nearly that quickly.
Link14 comments|Leave a comment

(no subject) [Jul. 16th, 2014|10:18 pm]
On our lunchtime ride, we got pulled over by the police. Not, alas, for speeding -- that would have been funny. We were pulled over because one of my coworkers ran a stopsign. The rest of us stopped, and he slowed quite a bit, but apparently not enough. He just got a verbal warning. He told me later it was his first ever moving violation and he felt pretty badly about it.
LinkLeave a comment

Infrared camera question [Jul. 15th, 2014|08:23 pm]
Resharing because A: it's a cool problem with neat progress already and B: someone else here may have some ideas about hardware that would work better.

Originally posted by rebeccmeister at Infrared camera question
Okay, folks. Science time.

Last year, I spent a chunk of time working on developing a setup for filming cricket activity. The challenge is that crickets are very sensitive to disturbance, so I had to come up with a way to film them remotely. Actually, there were multiple challenges. I also needed to film them both at night and during the day. Here's the process I came up with: I used a Wanscam webcam connected to a computer's ethernet port, set up inside of an environmental chamber (aka incubator). During the day, the setup looked like this (note the 3 crickets in the red square and also the use of plastic lids and duct tape on the sides to reduce light reflections on the glass pane covering the cricket cages):

Filming setup - day

At "night," here's what I had:
Filming setup - night

Those three light sources all have red safelights in them, under the assumption that the red light isn't visible to crickets. The Wanscam has an infrared setting, its own set of infrared lights, and a manually-adjustable focus, which allowed me to create lovely* videos that looked like so:

(this is a still frame)
Sample video still

The camera filmed at 100 frames per second, for 15 minutes, recorded directly to a PC. I then had an undergraduate laboriously work away at a data-processing workflow, to subsample the videos down to ~500 frames, generate background images, and use BioTrack to generate tracking data (timestamp, id, and xy-coordinates). I wrote a brief script in R to convert the tracking data into a measure of the total distance traveled, and also into pretty graphs that look like so:

Sample cricket track

So now, here's my question. The camera sucks. It was pretty much the only thing I was able to scrap up, after trying to look through a lot of different options and weigh various factors. I had access to a "fancy" HD video camera that produced mega-huge video files, but it was wretchedly terrible not only for that but also because its autofocus was garbage, especially at night. Five minutes after pressing "record" it would come unfocused.

Ideally, I'd like: slightly higher image resolution, infrared capability, manual (macro) focus, ability to manually control the camera in Ubuntu (this one had to be controlled through a Windows-based web browser), and less lens distortion. Do any of you have any ideas on where to look for such a thing, or how much it might wind up costing? I mostly just want to price out options at the moment.

Also, I am not interested in anything that has anything to do with EthoVision.

*Here I am using italics to indicate sarcasm font.
Link9 comments|Leave a comment

Department of Bad Ideas In Action [Jul. 15th, 2014|06:14 pm]
As I was driving home today I passed someone in a Kia sedan towing a motorcycle.
Like, towrope, motorcycle, and a person who pretty clearly didn't know how to ride a motorcycle under its own power, trying to keep it upright and not run over the towrope.
Worst idea I've seen in several months.
But hey they made it several hundred meters during the time I watched, without any of the manifold ways in which this could go hideously awry actually happening.
Link17 comments|Leave a comment

sleeping dreams [Jul. 15th, 2014|08:23 am]
I was back at Reed College and we were having a game night, playing something that was a combination of Risk and Global Nuclear War. There was an ambiguity in the rules that allowed people to vote on some aspect of game play, and I used that to try to Nomic the game rule structure into a wholly different game, proposing rules that we could start new cities on post-it notes and then hide them, for instance, but the other players were all "oh, no, we know where this is going" and since changes required (at least initially, as I pointed out) unanimous votes I made no progress.

Then a whole bunch of pigs walked into the room and everything dissolved in porcine chaos.
Link4 comments|Leave a comment

[ viewing | most recent entries ]
[ go | earlier ]