One of the problems of buying stuff for a vintage car is that except for the most basic items, you have to order parts from one of the very few dealers of those parts. Combined with the fact that dune buggy magazines have (apparently) promoted a lot of additions to Bugs that are a bad idea (chrome parts don't cool as well as black ones). So there's a range of vendors, from bad to good.
Unfortunately, as much as I like their facility, Mid-America Motorworks is one of the bad ones. There are definitely good stuff to be had there, and I really like the VW events. I like their on-line catalog too. Unfortunately, their catalog has some stuff that's poor quality, and some things that are just the wrong parts for the job.
I recently ordered an oil pressure gauge setup from them. Here is the relevant page on their site. It has two oil pressure senders listed, a single post and a double-post. I think the the idea is that the single post one can control the gauge, and the double-post one can run the gauge and the dash warning light. However, I've heard that the double-post ones don't drive the gauge properly, and that the correct way to do it is to have "T" attached to the oil passage and have the stock VW pressure switch run the warning light and a separate sender run the gauge. So I bought the "T" fitting from that page, and the oil pressure gauge and the single-post sender. Let me stress here that this catalog is specifically for air-cooled volkswagens.
So here's the stock oil pressure switch in my engine that's sitting on my work bench:
The switch has a terminal that sticks out. When there's no oil pressure, the terminal connects to the switch body and so to ground, making the oil light turn on. When the engine is running, oil pressure enters the switch and compresses a diaphram that disconnects the terminal from the switch body, making the light go out. If oil pressure drops, then the terminal grounds out and the light goes on again.
So, here's the T fitting screwed into the place where the switch normally goes.
Here's what the sender looks like.
A narrow end that screws into the block (or into the T) and then the other end has the electrical contact.
Here's what the final setup would look like. One fitting, connecting the oil supply gallery to two different sensors.
Unfortunately, this is not all wine and roses. The sender that I have doesn't seem to want to screw into either terminal in the T fitting, nor into the block in place of the stock switch. Something's up. Here you can see that the pitch of the sender threads don't match those of the brass T.
Wheras the stock switch and the T match perfectly
Huh. One thing that I noticed and got me thinking; it looks like the sender has a marking that indicates its thread specifications:
What's interesting is that alhtough it says 1/8", the diameter the screw part is more like 3/8 of an inch. It turns out that's just the way it is; as shown in this table, diameter of 1/8 NPT threads is .405 inches. The beetle is a german designed car, and most of the screws and fittings on it are metric, so this is surprising.
Well, it turns out that the thread spec on the sender that Mid-America sends out is just wrong. I figured it out at California Import Parts:
here's the sender that I need, with M10x1.0 threads (metric)(which they mention fits the stock location)
Heres' the sender that I got, which does NOT fit the stock location. I would be really nice if Mid-America could at least mention that the sender they have is not the stock thread set.
So I guess I'm going to sell this one on ebay, I guess, and then buy one that fits.
It was about time for me to get a new phone PDA, so a week ago, I upgraded my cell plan and got a new phone, a Palm Centro. Like my last phone, it's about at the end of its sales life. The Centro is smaller and lighter than the Treo 650, and so I think I'll like carrying it more.
The big thing was whether I could sync it with the Linux Jpilot tool. It was very easy; transferring things over from the old phone was harder.
To get the Centro to sync with JPilot, there are two things you need to know.
One is that one of the connection settings needs to be changed. Go to the hotsync application. "Options" menu, and go to the "Connection Setup" item. Select "Cradle/Cable" from the list, then hit the "Edit" button on the bottom of the screen. On the "Edit Connection" screen, push the "Details" button on the bottom. Set the "Flow Ctl" option to "on" (default is auto).
When you plug the sync cable of the Centro into the computer, a library will automagically connect to it. Unlike other PDAs, once the USB cable is plugged into the computer, it will connect.
The second thing you need to do is to set the location that the port to the phone will show up. Due to this automagic library (I got this info from this blog post) the Palm device won't show up as /dev/pilot. To use JPilot, you must tell it to connect to device "usb:" (without the quotation marks).
Once those two things have been done, open the JPilot application and hit "hotsync", then push the software sync button on the phone, and you're off to the races.
As far as moving the files from the old phone to the new, that was a bit tricky. Jpilot stores its state and data in the directory .jpilot. I created two separate directories, one called .jpilot_old and .jpilot_new to keep them separate. I wanted to be able to transfer all the informattion from the old one to the new one. That was tricky. I was able to get the calendar and memo information across by "exporting" them from the old and "importing" them into the new jpilot account. The contacts list wouldn't export properly, I would just get an empty list of lines. I used the Palm-OS IR "beaming" function to send the contacts across; that worked fine.
One of the reasons that I wanted the Centro was because I like applications that come from Palm OS. One of the principle ones is the RPN calculator program that I use. After getting all the basic information synced up from the last phone, I installed the RPN calculator and even dug up the unlock code to get it working. I now have a fully fledged phone again, how nice.
Some things don't live up to their descriptions on paper. On the other hand, some things end up being profoundly more than what they're cracked up to be. A couple of TV shows come to mind:
A several-hundred-year-old vampire is a police detective. He works the night shift of the homicide division. Subplots include his sort-of friendship with his painfully mundane partner and his friend, the medical examiner. Lots of flashbacks to earlier times before he'd turned to good. [Forever Knight. Ran for 3 seasons in the 1990s while I was in college. Very good writing.]
More recently: They're going to re-make the 70's sci-fi show Battlestar Galactica. The star is going to be the police chief from some 80's cop show. Oh, and the first Cylon we see is going to be played by a model who's really hot.
I have to say, the idea of microblogging, and the particular site twitter, had felt like those descriptions to me. Not realy something that I would get into. Although I certainly do see the appeal of text messages, that's very useful for small amounts of fast communication, without having to occupy someone's time exclusively.
However...I got a new phone recently, for unrelated reasons. Which I'm pleased with; I'll probably talk about that later today. And I got unlimited text messaging, because that seems to be the way of things. And recently I started posting occasionally on twitter, because other people I know do so and it's interesting.
So when we took a shopping trip to the Big City yesterday, I found myself having fun tweeting about being in the AT&T store, and shopping for Rock Band. It was satisfying and fun in a way that I would never have predicted. I think that my tweeting at home will probably be pretty low key, but when I'm out or travelling, I think I will use it to do travelogues.
Feel free to check out my twitter page. I will try to hook it in to this page at some point, so that you can get updates here as well. And I expect that I'll tweet a lot when I'm on the road, but less so otherwise.
By the way, if you're interested in seeing the movie _Coraline_ in 3D, this weekend is your last chance. Next week another movie takes over the theaters that have that projection capability.
I've seen it twice, once in 3D and once not; it's definitely worth seeing it in the original.
This has been a public service announcement from Neil Gaiman central command.
So--I realized in the fall that the right side cylinder head in my beetle was messed up, and had had a very high compression ratio and that was bad. Well today I worked on figuring out what to do next.
First, yesterday, I had to clean the garage to get to things. Here I've moved the table with beetle engine bits out of the way:
which makes some space on the work bench:
The goal here is to set the compression ratio to something sensible. First, find out what it is now, and then adjust the effective heights of the cylinders to give the correct combustion chamber volume. First, use the plexiglass disk and measure the volume of the combustiion chambers that are going into the engine
Measuring all four chambers, I'm fortunate in that they'e all basically the same, within 1cc of each other (they're about 50 or 51 cc).
Next, measure the "deck height", which is the space above the piston in its cylinder when it's at top dead center. So for each cylinder/piston, set it to top dead center position
and measure the (small) distance from the piston crown to the top of the cylinder:
So take the head chamber volumes and the volumes from deck height, do some math
and come up with...
The good news is that things in my engine are pretty much consistent--my best guess of the compression ratio of all four cylinders is that it's about 7.9. Probably higher than I want it; the correct value is 7.3.
Now I'm curious why it's not set right. The deck height seems to be too low (this article says that it should be greater than 1.5mm; mine are around 1.2mm). The deck height is adjusted by putting shims in between the cylinders and the engine block. For the very first time, today I pulled a cylinder open to expose the inside of the block so that I could check to see what shims they had installed.
Well, it turns out that the reason is that the person who put the engine together didn't put any shims in at all. Here you can see the paper gasket that's peeled partially away, revealing the magnesium sealing surface. No shim.
As I go through the process of debugging this, I think back to 2006 when it was brand new to me and I drove it from Illinois to Kentucky fairly fast on a hot July day. Every time I find out something else wrong, I marvel that I didn't weld the engine into a solid block that day. A brief run-down on things that were wrong at the time: - the engine was put together without any barrel shims, making the base compression ratio 7.9 rather than 7.3 - the right side cylinder head had been machined too many times, lowering the compression ratio even farther - there was a leak around the throttle bushings, letting air into the induction system and making the car run leaner - The timing was advanced way to far at the time - there was a piece of engine tin missing, allowing the hot exhaust air of the oil cooler to be sucked in by the engine cooling system - the oil pump had been put together incompetently, with the wrong gaskets
All in all, amazing I made it here. (Not without delays; that was the day that the starter started failing). With all that, though, it's nice to know that one of the most important pieces of the beetle engine is indeed there:
That's the cooling air deflector that forces air coming down between the cylinders from the main cooling fan to pass through the fins of the clyinders, rather than just going between them and out. This is a piece that's occasionally left out, with disasterous results.
I really love shopping at Harbor Freight. Although they don't necessarily have the very very best deals on everything, they carry tools that are impossible to get at any other consumer-level retail store. And unlike discount stores, you can usually get JUST what you want, not a blister pack of 24 associated tools.
I was very pleased on Saturday when I went down to the Big City to see Coraline in 3D, we stopped by Harbor Freight on the way. In the front of the store, they had a bin of digital multimeters:
for THREE DOLLARS. Usually they would cost ten times that at any store. I bought four of them. I took one out of the packaging and it works.
I'm going to be putting together a tool box for each of our cars, and I'll put a multimeter in each one. They're incredibly useful, and that way I won't be without one wherever I go.
They're an infinitely useful tool if you every work on anything electrical, or on cars, or really anything.
I didn't, unfortunately, take any photos of the re-assembly process. Once I had gasket sealant on my hands, I didn't want to stop and take photos; besides, that would have mucked up the camera. I took the car on a test drive yesterday (Saturday), and it was back to its old behavior. Huzzah!
One interesting thing about this repair--the uppermost bolt on the thermostat housing is incredibly difficult to get to. It's in behind a couple of other things, with no direct way to get at it (there's something in the way of a standard socket extension). However, it turns out that at the Oshkosh air show a few years ago, I bought a 1/4-inch drive flex driver, which was perfect for this application. Here it is in action:
The flex drive is the black snaky thing, which is flexible but it transmits twisting force very well. The bolt that I'm turning is near the upper left of the photo, under the spark plug cables.
Here's a slightly closer view; the shiny thing is the socket on the end of the flex drive, and the bolt is immediately to the left of that.
Well, I think I found the problem with the Escort's heating system.
Yesterday evening, I took took the thermostat housing out of the escort, since all the diagnostic indications I could find pointed to a faulty thermostat. Just for background, the way the cooling system in a water-cooled car engine (all of them, except pre-1980 volkswagens and Corvairs): There's a water pump (probably a centrifugal impeller) that draws water from the radiator and pressurizes the water jacket inside the engine block that surrounds the cylinders, usually at the bottom of the block. On the other side of the block, near the top, there's an outlet that goes to the upper radiator hose that carries water to the radiator, so the coolant (water/antifreeze mix) flows in a closed loop. At the outlet point on the block, there's a mechanical thermostat valve that opens when the cooling water is up to temperature. This is a remarkably stable technology; it's been done the same way for tens of years, at least. The engine computers have temperature sensors for for air and coolant, but the cooling system is actuated by this purely mechanical valve.
Here's what one looks like; this is the one that I took out of the engine:
The ring forms the seal between the water jacket in the engine block and the outlet hose. The spring at the bottom sits in the water in the block, and when it hits the proper temperature, it pulls the core of the thermostat down, opening the way for water to flow up.
I think the problem with my car was the rubber seal was distorted and didn't seat right, letting some water get past it no matter what its temperature was. It's possible that's because they installed the wrong thermostat, which was slightly too big and had to be jammed into the thermostat housing. Here are the two thermostats, the one of the left is the one I removed, the one on the right I bought at the store the other day
The outer diameter of the old (nonworking) one is slightly bigger. The new one on the right slides into the thermostat housing pretty tightly. On the lower right corner of the left thermostat, you can see where the rubber seal is distorted.
Here's the thermostat housing, looking from the point of view of the engine block if it was installed
and here with the new thermostat installed
It's supposed to get warm here this afternoon; I'll probably take half a vacation day and put this all back together, and hopefully things are better.
With all that I talk about my beetle, one might think that's my only car. We also own a Ford Escort, which is what I drive when the Beetle is being worked on (like now).
I had some work done on the Escort recently, and on the next big drive, instead of seeing this on the temperature gauge,
I had this:
; oh, and the "check engine" light came on.
One thing about these newfangled cars (starting in the 1980s and anything after 1996) is that they have a computer that controls the running of the engine. The computer has inputs from sensors that tell it temperature, pressure, and flow and it sets the output of, say, the fuel injectors appropriately. If the computer detects something that happens outside of its parameters, it logs the error and turns on the "check engine" light. Mechanics have widgets that can read the error codes and that will help them figure out what's wrong with the engine, if there is something.
I decided that in case the error conditions were something that could be read out, I wanted to have a code reader, so I picked up the cheapest one that O'Reilly's had:
It was about $70, which isn't too bad for a tool that I'll use a lot.
It has a keyed multi-pin connector that connects to the diagnostic plug on the car.
The diagnostic plug is typically in front of the driver's knees at the bottom of the instrument panel. Here's where the plug is in the Escort:
The reader is powered by the car's battery. When it's plugged in, it reads codes out of the computer's memory automatically.
I'd just like to get the word out about Coraline, a book by Neil Gaiman that's been made into a movie by Henry Selig (I think) who made _A Nightmare Before Christmas_. The book is a spooky adventure story about a girl discovering a very creepy alternate reality accessed through her house. The film is entirely stop-motion animation. From the clips that are out, the film captures the wierdness and creepiness and fun of the book very very well.
I ask that if you're intereested at all in true, creative film-making by people who are true artists that really love their craft and are really good at it, then please see this film as soon in its run as possible. Opening weekend statistics are very important to movie distributors, and so if you want to encourage distributors and movie theaters to get small, good films like this, then please see it as soon as it's out.
Neil and his publishers have put the book itself free online if you want to check it out.
Coraline comes out this Friday, February 6.