June 29, 2005

Test Rig

Made up a test rig to evaluate the Z-axis magnetometer from Honneywell. I got three sets of 3-axis magnetometer chips for building a tilt compensated compass. The chips are all micro tiny surface mount chips with the largest one being the Z-axis magnetometer.

I have to fabricate a circuit board to mount the other chips just to test them so they are going to sit for awhile. The magnetometer signal is a small (millivolt) range signal indicating the magnetic field.

Finally started to get the hang out of mapping the signal when I had a slight misunderstanding of the proper method to cycle the reset pin... Over amp'd it and zap - Adios! Lesson learned the hard way. Good thing Honneywell sells those chips on a minimum order quantity of 3.

June 27, 2005


Worked up a quick test rig to get a feel for what the accelerometer signal looks like on a little road test with the Elantra stationwagon. The voltmeter shows about the midpoint of the supply voltage. The Analog Devices two axis unit modifies the voltage on the pin to reflect the applied acceleration for ± 5g's.

This type of sensor appears to be originally developed for tilt compensation of digital compasses, but people have gone on to use them along with rate gyros for attitude/heading reference systems for remote control/UAV helicopters.

June 26, 2005

Test Fit

Cathy wasn't available when it came time to take the canopy off, so I came up with a simple solution to hold the nuts on the inside under the front of the windscreen while I removed the screws. It really is nice to be able to take the plexiglass back off the plane. This should be a real advantage when it comes time to paint everything.

June 22, 2005

Black Hole

Pretty darn small. This is the GPS unit I got from SparkFun electronics. I'm not knocking them but the Silicon Lab's outfit out of Austin Texas that makes that little tiny black chip... Grr! Seems as though they are committed to playing hide the ball on device drivers.

I finally verified the device was working with a GPS diagnostic tool downloaded from the GPS unit manufacturer running binary drivers for Windows (rebooted to a legacy Windows 98 box for this type of problem). So, we know the chips and the board are working, but getting the communications to work accross the last 1/4 inch from the USB connector to the GPS unit in the silver box... A proverbial black hole.

Silicon Labs has releases source code for a Linux driver for 2.4 kernals (I'm running a 2.6 kernal) but some BOZO thought he would be cute by chopping the source file into single token bits and be a real wise guy and convert all the identifiers to variations of "O0100101010101" type strings make the file gibberish. I wouldn't care if it compiled but it don't...

On the bright side, I've been able to use some tools to interrogate the little rascal to get it to identify it's USB-2 parameters. Now if I can get the bulkread function to work, I may be able to get the GPS unit's output stream. Still, it's a cute little rascal. Kind-of funny how the antenna is bigger than the actual unit, and the power for the whole shebang comes off the USB cable.

Oh well, at least I now know the front steps of the house are more or less located at 72.06864 degrees West and 43.71213 degrees North at an elevation about 335 meters above mean sea level.

June 21, 2005

GPS Time

Put some finishing touches on the canopy plexiglass to countersink the holes and draw-knife the edges to get rid of any stress-riser scratches. I tested out installing the canopy with Cathy and we got screws and nuts finger tight in practically all the holes.

Got a call from the post office about 4pm to see if I wanted to pick up an express package that had come in late, which turned out to be the GPS/USB board combo I ordered from Sparkfun Electronincs. Cute little unit, but it turns out the USB chip maker has been playing games on releasing driver information... A very bad start.

Looks like we are in for a little "fall in the ol' black hole" session chasing communication protocols. Looks like I am about to get immersed in the trials and tribulations of Linux USB interfacing (LIBUSB). After burning through the afternoon and all evening I was able to verify the kernal sees the device and knows who it is.

Now I need to figure out how to get the bulk transfer protocol to work and burrow through the GPS chip's documentation. This ought to kill a solid two weeks... Honeywell sent me an e-mail saying they shipped the chipset for the digital compass.

June 20, 2005

Canopy Day

Finally, after months of various details it was time to rivet the canopy frame. It turned out the key to riveting the skirts and skin in place (with the underlying stiffeners) was to rivet the skin/stiffeners on the outboard edges first, then rivet the skirt on and then rivet the center stiffener in place.

To make this sequence work I machined up a modified version of the bucking bar I worked out helping Brian on his 7A. What I needed was a larger "foot" to fit under the stiffener to buck the rivets on the outboard sides of the skin. When the dust settled I was still four rivets short of having them all installed. (I need to make a bar from scratch with a longer "foot" to buck the last four rivets.)

Van's aircraft should really consider providing a little more information on the option of making a metal skirt instead of the fiberglass layup method of attaching the front of the canopy. It really isn't all that hard (just tedious) to make those front skirts using essentially just a stick of oak with a slot cut it in. It may be the old fashioned way of hand forming aluminum parts, but it works!

When I worked at Connor Engineering in Danbury, CT in the 70's there was an old timer in the prototype shop who would tell stories of his early days making aircraft parts for Chance Vought et al. When he started out, he was shown a bucket of mallets, a sand bag, and a sheet of aluminum with the announcement: Sonny I'm going to show you how to make an engine cowling... May be slow but hand forming parts still works.

If you look at the air inlets on the leading edge of EAA's Aluminum Overcast B17 you can see they are hand formed. In any event, I sure had fun working out how to make those skirts and I would recomend anyone building an RV at least consider the alternative to the fibeglass layup method. Think of it as a stress reducer - or producer - depending on your disposition.

Thanks to Dino for getting me the invite to Plymouth Airport to crash their BBQ so I could talk the builder of the RV-9 which had aluminum front canopy skirts. It was a real boost to talk to someone who had actually done it and get some insights into how it should be done.

June 19, 2005

Bright Light

Finally some sunshine! Took advantage of it by priming some canopy parts for final assembly.

June 18, 2005


Finally worked out the sequence for riveting the final assembly (I hope) and it's kind-of tricky as I am trying to flush rivet the forward canopy skirts and the riveting of the braces inside the assembly makes for a sort of dock and dinghy problem...

June 17, 2005

More Plans

Prepping parts and planning out the riveting for the canopy...

June 16, 2005

Final Drill

Finished drilling the front skirts to the canopy frame. It took a little bit to work out the details to zig-zag around some structures. Came up with a simple gauge to check for clearance for a screw to secure the canopy to the front skirt. Once the hole locations were worked out it was time to take everything back apart to drill pilot holes in the skirts, then put it all back together to drill the pilot holes through the canopy.

Now it's time to take it all apart once again for the final drilling to size and countersinking on the canopy etc. and then we'll be ready for final assembly of the parts. Kind of neat to be able to bolt the canopy on and then be able to take it back apart again, hopefully this will pay dividends when it comes time to paint the plane.

June 15, 2005

Some Issues

With the skirts formed and fitted it's time to lay out the holes and drill them. This is still a work in progress, in stages, as there are some issues with a couple areas regarding to how to interface with some of the other structures that make up the canopy frame. This is going to be a riveter's nightmare on final assembly, as we try to avoid having to use blind rivets.

June 14, 2005

Form Fitting

A couple of days back the 2024-0 sheet arrived. So today it was time to form, then drill the skirts into place... Here's a three photo synopsis of the sequence of forming and fitting the skirts.

You basically set the flange with the notched oak stick, wipe the bend to fit the cross section with your fist "ironing" the sheet over the bend in the fusalage. The ironing takes some of the flange angle out, but not all of it. So you trial fit the piece, then do a little more with the oak stick, iron the piece etc. until you get the fit right. After that I trimmed the left and right ends to butt up against the canopy frame sheet metal.

June 12, 2005


Saturday and Sunday finished up some miscellaneous firewall details like making the hole big enough through the firewall so you get the bolt into the engine mount weldment which holds the front gear leg on. Deburring stainless takes a real effort.

June 10, 2005


Canaan has been getting hammered by thunderstorms the last couple days. Closed route 118 with a washout. We've been pestering the town for a couple years to fix the culvert and thank goodness they finally put in a new 2 footer to replace the ancient (collapsed) stone culvert.

Meanwhile, we're finishing firewall and canopy details. The side skirts and canopy latch fingers are ready for priming. If my GPS unit gets here I'll see if I can run some software tests: Of course, without a laptop it's going to look pretty funny to lug the workstation out to the station wagon. If I can mooch an oscilliascope somewhere I'll also take a whack at testing my analog devices gyro.

The Dallas Texas based robotics group's website at http://www.dprg.org has an interesting page on circuit board for use in a two wheeled standing robot and links to expresspcb an internet based prototype board fab shop. Was looking at patent number 6,671,648 on the USPTO's website on gyros which gives an interesting background on just how many types of gyro's there are. I might put a little summary doc together after I do a little more background vacuuming of sources.

June 08, 2005

Skirt Flash

Pumped up with the knowledge it can be done, courtesy of the RV-9 yesterday, I proceeded to work out the kinks (literally) in forming a front skirt molding for the canopy using aluminum roofing flashing as a test media. I figure flashing has to be at least strain hardened to a half hard state to begin with, as the stuff doesn't form worth beans. Still with a lot of patience you can do it... It came out so well I went ahead and ordered some -O temper annealed aluminum sheet from airparts to take a crack at the real thing.

June 07, 2005

Bullwinkle BBQ

Finished mounting the main tires to the rims and assembling everything onto the main gear legs this afternoon, but not before an interesting and productive diversion. Dino Vlahakis called to let me know of a BBQ for retired airline pilots at Plymouth airport and suggested I come over to look at some of the RV's.

I finally got a chance to see the RV-9 I spotted back in May without the canopy covered so I could check out the interesting approach the builder, Malcomb Brawm of Hebron NH, used to secure the windscreen. Normally, the front edge of the canopy on RVs are fiberglassed to the fusalage. This RV-9 uses a sheet metal skirt to do the same thing, only with this method you can remove the canopy (for painting).

Malcomb was very helpful in sharing how he fabricated the windscreen skirt. The shot of the Stearman is Dino's plane on departure from Plymouth. I couldn't see worth beans on the digital camera display because of the glare and took the shot anyway. The young moose are moving around now and it was a real treat on the drive home to get the shot of the young bull moose, but unfortunately I don't think he'll fit on the tail as a mascot.

June 06, 2005

Rolling Along

Finished the last pop rivet details on the center harness anchor this morning. Here's a shot of the the cockpit with all the sheet metal (etc.) in place.

Shortly after taking the shot I also put in the vents and piping. I've been itching to move on to get the wheels done to be able to kick the fusalage out of the house for the summer. The nose wheel and its assembly worked out pretty straight forward but... And there's always a but... The main gear stumped me.

There are a couple of fastener details to the main gear's tubes on the nozzle I need to figure out before I can for sure split the rims, insert the tube, and inflate the tires. The plans for the disk brakes on the mains are a tad sketchy also, so it's going to take a bit to get the order of assembly down pat.

On the avionics front I ordered a GPS unit that I can read through a USB port the other day, and may have a source for an experimental six axis attitude sensor system. The GPS unit is sold by the same outfit that pretty much tracks all the big rigs you see on the highway these days.

June 05, 2005


Used my new pop rivet gun to start installing the center seat belt attachment point. I kept removing a little metal here and there on my old pop rivet gun then ...Argh! Dropped it and disintegrated the nose piece.

June 03, 2005

Side Skirting

NC drilled the rivet hole pattern in the side skirts and fitted them to the canopy frame. CNC drilling the hole patterns looks a whole lot better then my battling with a wandering drill point with a hand laid pattern.

I found particularly with drilling that cleco's that the occasional extra clamp is all that's needed to hold the sheet metal against a piece of wood. What I did here was drill a series of 10 holes, then I moved the piece on the jig and reran the program to extend the pattern... Do that a couple times and you have nice, evenly spaced holes.

June 02, 2005

Canopy Skirt

If I were a master at metal forming arts, I would make the front canopy skirt in one piece. Since I'm not, I'm going for the two-piece approach using a pair of "flat" patterns to develop the skirt.

The upper pattern is for the piece to wrap around the canopy, and the second is for the tie-down piece which wraps accross the canopy frame skin. The upper pattern requires a blank about 54 inches in width and the lower one requires about 48 inches.

June 01, 2005


Doing some work on pattern development. I've been playing around with aluminum flashing to work out fabrication and fastening options. This is going to kill a lot of surface area of some aluminum sheet(s) eventually so I'd really like to get this right the first time.

The idea is to wrap the upper piece accross the plexiglass and with the tabs bent up to intersect the canopy frame's skin... Then put a cap over the combination to rivet the assembly to the canopy frame/skin and then bolt the canopy to the skirt. Cathy says this is like the pattern work to make a shirt collar.