With Wings As Eagles: Craig P. Steffen's Blog

the smell of glue in the morning

2007 August 30 06:06

I think I'm done with the major painting on all the pieces of the model. Since I painted the parts on the sprues, I'll have to touch-up the points where they get cut from the plastic.

I got 4 windows glued into the two halves of the fuselage while watching Mythbusters last night. The next time I sit down it it, I'll probably glue the body together.

I haven't written lately, because nothing to write about. I have my pilot's license, so I'm not driving to lessons, and so I'm not driving the beetle. I've also been keeping my nose to the grindstone at work.

golden ticket

2007 August 22 08:05

When you pass a test for an pilot rating, you receive a "Temporary Airman Certificate" (form 8060-4) that grants you the privileges of that rating. In my case, on Sunday, I used that temporary certificate so that I was legal to fly the airplane back home to Morristown.

In case you're interested in what one looks like, see page 11 of this document on the FAA's web site. Mine was filled out by hand, section IX says "private" and section XII says "airplane single engine land".

what if they don't need seats?

2007 August 21 10:22

The model is coming along. I'm in the process of gluing and painting the internal sub-assemblies. Here's the engine on the engine mount:
I noticed in the process of making the photo that there are a few touch-ups that need to be done.

Here is the cockpit floor and all but one of the seats glued in.
I think I like the bronze color that I'm using. If this model were to be examined up close, then it would be too garish of a color. However, these seats will only ever been seen through the windows in the model, and so I think the very intense color will compensate for looking through the windows and will end up looking about right.

this time the slip wasn't pink

2007 August 19 22:35

I passed the check ride re-take today. So I'm officially a pilot now. This has been a very long time in coming. I've wanted it as long as I can remember.

This is me, with the Beetle, and the airplane in the background that I took the test in. In fact, it's the same airplane I took my very first training flight in on July 26 of last year.

I'm way too tired to write anything useful. It's time for bed. Model uploads soon.


2007 August 19 06:46

Heading to take the check ride again. With any luck this will be the last time for a while.

The model is coming along. I'll upload photos when I get back tonight.

I think it looks good in Bronze

2007 August 16 23:29

Back from vacation/work trips, so I'm trying to get back going with model painting/building. The seats of the Cessna that I've been flying are 1970s light brown. The closest color in the pains that I'm borrowing from my Model Building Friend is "Dwarf Bronze", so that's what I'm painting the seats. I like the way it looks, and I wanted to play around with painting larger surfaces with one of the metallic paint colors.

Here's the engine after a round of paint to clean up the ends of the engine cylinders where the sprues were cut away.

read the fine FAA manual

2007 August 16 23:08

On monday I flew with my instructor again to practice the stuff that I didn't do right in my unsuccessful check ride. We did soft field approaches (which is what I failed for), steeped banked turns, and power-off stalls. Here's the Beetle parked at the airport, to show my wife that I really do go to the airport when I say I'm "flying". :-)

Although I have a copy somewhere in the house, while I was at the airport, I bought a copy of the Private Pilot Practical Test Standards:

I advise anyone preparing for a private pilot's license to own and read this book. Then read it again. It's the prescription and the limitations on what the examiner can ask and what you need to do in response. It contains all the criteria that you will be tested on. I'm not an advocate of stuyding only for the test itself in a situation like that, but having at least studient the book a couple of times means that there won't be any surprises.

For instance, steep banked turns that you'll be asked to do in the check ride are once around, 360 degrees. Apparently, the steep banked turns used to be 720 degrees, which is how I'd practiced them, and I ended up screwing one up.

Another interesting point that I came across while reading it recently, is that on a re-test (which I do on Sunday), the examiner can test you on any aspect of the test, including items that you've already passed. It doesn't say whether or not you can fail for items that you passed the first time. It'll be interesting to see what exactly goes on at the subsequent test. I'll report back.

hesitation of 34PICT-3

2007 August 13 20:24

I've read in a lot of places that the Solex 34PICT-3 carburetor (stock on my 1972 Beetle cars around that same age) is fiddly to keep tuned with the stock timing. Many people take as gospel that this combination just has problems.

Well I'm now wondering about that I wonder if it's possible that the problems that are experienced with that carb is just a characteristic of the carb itself. In June, I put in a carb that I bought on ebay and cleaned out thoroughly and rebuilt. Right before I left for work and Oshkosh in July, the car started hestitating a little bit off of idle again. Over the weekend, I tried to effect a quick repair to make that better, and I think it may have done the trick.

Since my car had no hesitation with a fresh carb, then developed hesitation again, and now the simple cleaning fix got rid of the hestiation, I wonder if the hesitation reputation of that carb is mostly due to the passages getting clogged up.

Here's the scenario. Below is a diagram of the 34PICT carb with dots to show where air and fuel flow at idle (air is blue, fuel is red, air/fuel mix is purple). The throttle plate is closed. All the air that is getting to the engine is going through the bypass drilling on the right.

Here's the same sort of diagram with the throttle slightly open. When the throtle opens, there isn't enough airflow to cause the main jet to discharge fuel into the air stream. So if the idle bypass and the main jet were the only mechanisms for fuel delivery, as the throttle was opened the engine would run too lean and perhaps stumble or stop. To make up for this, there is a series of small metering holes to inject fuel into the air getting past the throttle plate as it opens. This diagram shows the flows with the throttle slighly open. There is air getting past the throttle, and the fuel is being metered by the holes near that opening.

This is taken on another 34PICT-3 carb that I have. This is taken of the engine end of the carburetor. The throttle is slightly open, and at the edge you can see the metering holes.

My theory is that the hesitation that I (and perhaps others) experience is because one or more of those holes are clogged. What I did, which seemed to have worked, was to take the top of the carb off and shoot carb/choke cleaner into the jets that go to the passages where he holes are. One is the upper of the two brass passages in the bottom half of the carb here:

The second is the jet that leads to the metering holes. In the above photo, it's the thing that looks like a hexagonal-headed brass bolt. The passage that goes down from the barrel of that jet goes to the metering holes.

If my supposition is correct, and the fault is gunk in those metering holes that will dissolve with carb cleaner, then all I have to do to fix it is to remove that jet and shoot cleaner into the port at the bottom. I'll try that if the problem recurrs. The jet is on the side of the carb where you can get at it without removing anything else. You can see the jet here, with the carb and everything re-assembled, the jet head below the automatic choke:

So the car ran fine today, no hesitation. If the hesitation comes back, I'll try this again.


2007 August 12 12:30

We saw the movie Stardust last night, based on the novel by Neil Gaiman.

It was very cool. All involved enjoyed it, at least in our car load. I stayed up late last night going through the book and looking up various scenes. In the process of translating from the novel to the book, things got changed, and expanded, and of course lots of stuff got left out.

I highly encourage anyone who's interested in fantasy-type stories to go and see it, today (Sunday) if possible. Opening weekend ticket figures are very important to the studios, and I'd definitely like to encourage this kind of new and fresh movie making. If you can't see it today, then please see it anyway in the near future; it's very fun and well worth seeing.

And of course, buy the book.

copy that into your spell book

2007 August 06 23:49

Using software under Linux is quite often like owning a vintage car (I've done both). There are things that don't quite work the way you think they would, and so you have to sort out how to do something new.

There is most likely somewhere an open source software package that will scan in a bunch of images from a scanner and put the resulting images together into a pdf. The machine I have hooked up to my scanner is Knoppix, and so the scanning software is pretty basic. I scanned the pages of the document that I wanted as .png files, and then moved them to my laptop to put them together.

The imagemagick tool "convert" is quote capable of this. Here is the commands I used to batch images together into pdfs of four pages at a time:

convert -crop 3250x2550+257+0 kscan_0001.png kscan_0002.png kscan_0003.png kscan_0004.png -density 300 -adjoin mp1.pdf convert -crop 3250x2550+257+0 kscan_0005.png kscan_0006.png kscan_0007.png kscan_0008.png -density 300 -adjoin mp2.pdf

The -crop directive cut off some artifacts from the left side of the scans. the "-density 300" is to tell "convert" that the photos are scanned at 300 dpi, to get the embedded image characteristics right. The reason that I didn't just put all 14 scan pages on the same commmand line is that convert seems to have trouble doing that without using so much RAM that it brings the machine to a halt. I tried to use convert to put all the pdf files together at the end of the process, and apparently, due to the odd dpi setting, it couldn't do it.

Fortunately, this problem was no problem for my google-fu. I found a page that talks about contatenating pdf files: The ghostscript option they listed worked for me, it only used a little bit of RAM for a 14 page document, and the output file is exactly what I wanted. Here was the final version of the command I used:

gs -q -sPAPERSIZE=letter -dNOPAUSE -dBATCH -sDEVICE=pdfwrite -sOutputFile=out.pdf mp1.pdf mp2.pdf mp3.pdf mp4.pdf

There's something for your toolbox.

keys to the kingdom

2007 August 04 16:20

I'm lazy in a lot of (most) ways. A lot of times I don't bother to set things up on my computers in a sensible fashion that ends up costing me time, but I never quite have the time to do the job right.

My phone, my Treo 650, is a very bad example of this. I've had the thing for a year, and I still haven't updated the firmware which I hope will fix a bunch of the problems I have with it. Ah, my naive optimism.

So finally, today, I'm getting my Ubuntu 6.10 laptop set up to sync to the Treo so that I can back things up, and install software, and so on. There are a few tricks involved.

The kernel that I currently have installed grabs the Treo out of the box. That's nice. When I plug usb sych cable into the laptop and hit the synch button, the Treo is mounted as /dev/ttyUSB0 and /dev/ttyUSB1.

I've always been a big fan of the JPilot software, which provides a syncing facility to some Palm-based PDAs. JPilot can connect to the Treo with no problem, you have to set the serial port to be /dev/ttyUSB1.

However, Palm devices apparently have a "username" and "user id", I suppose to prevent you from accidentally saving the information from one PDA onto another one. The Treo 650 apparently has a NULL username by default, which JPilot doesn't like. To install a non-NULL username and ID (the ID is just a random number), you must use the install-user command which is part of the pilot-link package. The command line goes like this:

craig@traal 16:13 ~ > install-user -p /dev/ttyUSB1 -u "craigtreo" -i 99521

Listening for incoming connection on /dev/ttyUSB1... connected!

Installed User Name: craigtreo Installed User ID: 99521

craig@traal 16:14 ~ > install-user -p /dev/ttyUSB1 -l

Listening for incoming connection on /dev/ttyUSB1... connected!

Palm user: craigtreo UserID: 99521

The second command reads the username that's currently on the device.

Apparently the proper way to have more than one Palm device sync to the same user account on your machine is to have a different directory for each one to save the JPilot status files. By default it uses .jpilot in your home directory, but the environmeht variable JPILOT_HOME overrides that. So now my .jpilot directory has the files to sync to the Treo and .jpilot_old has the files to synch to my old PDA.

I went ahead and installed my favorite application, an RPN calculator on the Treo. Now I'll have a good calculator application to carry with me again.

Now to go figure out how to update the firmware on the Treo.

idiot light

2007 August 03 07:58

In an ideal world, all vehicles would have at least one gauge, preferably two, indicating the state of the electrical/charging system. Ideally one measuring the main bus voltage and one indicating either the charging or discharging of the battery.

For reasons of keeping the instrument panel simple, and for economy of manufacturing, many cars don't have either of those, but instead have just a simple light that warns that the charging system has something wrong with it. This light looks like a battery, and in my apparently naive, optimistic little world, this light goes on to indicate that the charging system is having problems. Ideally, the battery light going on indicates battery discharge. In principle, this happens any time the battery is discharging. I assumed it was normally implemented as a voltage sensor; if the charging circuit voltage is above a set value, the light is off; below, and the light is on. Say, if the threshold is 13V, then the normal 14.4V charging circuit value has the light off, but 12V (running on just the battery) it's on.

After having alternator trouble with my 1996 Ford Escort, I found out that apparently this scheme is way to simple for the engineers at Ford that the time this car was designed. What I was told at the Ford dealer the first time I had it fixed (incorrectly) was that the engine computer will turn off electrical components to try to preserve battery life, without ever having turned on the charging light. This defies all logic and reason as far as I can tell. Does anyone know if this is true or not? That the voltage light won't come on, despite the fact that the alternatur clearly isn't working properly, because there are mechanisms that are supposed to be making the battery last longer? If you've had this experience, please drop me a line and confirm this.

The symptoms that happened to me were that the cruise control stopped working, and then the tachometer turned off (engine still running). Not too long after that, it started having trouble staying running, then quit.