Saturday, September 05, 2009

And so it begins...

Some of you may know that I've been in the research phase of a new vehicle project for quite a while. Using the excellent resources of the Expedition Portal forums, I've been scheming on the most affordable way to build a 3-4 passenger Overland/Expedition Camper. This should be the start of a series of posts as I being the transformation from8-passenger midsize van to what a friend has dubbed an "Astrolander".


Sunday, June 07, 2009

Baby's First Hack

Even before Kimberley and I knew we were having a baby, I knew I was looking forward to being that "cool dad" who can build anything for his Kids. Science projects and Halloween costumes are going to be epic in our household in the years to come. I've always enjoyed hacking for myself, and I always knew I'd end up hacking for my kids, but this little project sort of sneaked up on me. Its appropriate that this "Baby's First Hack" fits perfectly in the nature of most good hacks: It was a simple fix for a problem that needed solving only because of limitations in the original design of a product.

As Kimberley and I were starting to move into the nursery after all the paint and wainscoting was finished, she ran into a problem installing her preferred mobile, the Tiny Love "Sweet Island Dreams". She preferred this one because it played real classical music, recorded with a real orchestra, not dumbed down MIDI or electronic noise. However, our new crib has a very wide cap rail (click first pic to see), and there wasn't enough "bite" in the mounting system to attach the mobile. Examining the mount, I could immediately see where the manufacturer could have solved this problem for us by making the mount for the threaded post more adjustable (a few more molded in slots for the "lock" in the back of the plate would have sufficed), but as is so often the case, this was apparently designed as a "one size fits some" solution.

Enter Baby's First hack. I cut a mounting plate from some 3/16" aluminum sheet I had in my shop that allowed me to drop down the mounting spindle to allow the mobile to fit over our wide crib rail. The resulting mount and its clearly Autobot-logo outline was purely a bonus. At least my daughter will grow up to fight Decepticons.

The most complicated part of this whole mini-project was getting the mounting flange off the back of the mobile. Everything in this toy is secured with a weird triangular anti-tamper screw that I've never seen before. It gave me an excuse to finally buy the 100piece anti-tamper bit set I've been eyeballing, and fortunately one of the tri-wing bits included gave me enough leverage to remove the screws, even though it wasn't the correct bit.

After that, it was simply a matter of screwing the mounting spindle to my new mount, then attaching that where the locking flange used to be on the back of the mobile. Since this mount covers the battery compartment (and will thus need to be removed for battery changes), I replaced the tamper proof screws with some Phillips-head screws from my collection of parts removed during previous hacks. In a rare move, I went ahead and discarded the tamper screws rather than save them, as they're a subtle affront from the manufacturer. Remember the Hacker's creed: If you can't open it, you don't own it. Anti-tamper fasteners are just another way that manufacturers try to trap you as a consumer. Devices are designed to be hard to repair or modify so that you have to buy another one as soon as it ceases to meet your needs.

As you can see from this side shot, I've increased the vertical bite of the mount by about 1-1/4". In case it isn't obvious, the threaded spindle passes between two vertical slats in the crib, and is then secured with the giant plastic wingnut. Combined with the flange at the top, this provides a surprisingly secure yet easy to install mounting system. What probably isn't obvious in the first photo (mounted on the crib) is that I've chamfered the edges on the plate and sized the mounting screws for the spindle so that they're flush to the plate, so this is nice and smooth and baby-safe!


Thursday, April 30, 2009

Weekend Project: BMW Double VANOS rebuild

Double-wha? Simply put, VANOS (Variable Nockenwellen Steuerung) is BMW's name for their fancy oil-driven system for variable valve timing in their 6-cylinder engines. Oil is pumped into pistons which move the camshafts fore and aft to hit slightly different cam profiles, thus increasing engine performance and fuel economy. The problem? Due to a failure of materials design, basically every VANOS engine from 1992 to 2006 will eventually develop leaks that prevent the pistons from functioning correctly. If you're under warranty or willing to spend the money, they will happily fit a rebuilt VANOS assembly with the same design flaw and the same failure mode. A full explanation can be found at Beisan Systems, written by the clever bloke who figured out the problem and the solution for his own car, only to be told by BMW that "no further development" would be done on the design.

Luca, one of the members of the local San Diego BMW Z Car Club (SDBMWZCC) had done the research into this problem for his own stable of BMWs. Luca is an awesome shadetree mechanic, and I've never seen him do something half-assed, so if he was going to fix his wife's Z4, it was going to be done right, which meant fitting the upgraded O-rings and Teflon seals from Beisan Systems into the VANOS pistons. And in a display of car-club-generosity unmatched in many months, he also organized a group purchase to bring the price down the price of the parts, then hosted a clinic at his condo where several other Z-owners could turn their own wrenches under his guidance!

With a Sunday appointment, I readied my tools and headed off to the clinic. None of this work is terribly difficult, but quite a few parts have to be stripped back from the head in order to get the VANOS assembly off, including the valve cover, spark coils and harness, and fan. Since the cars are low, this isn't exactly easy, but this process is really good for getting familiar with the more intimate parts of the engine, and I got my first really good look at how the VANOS system interacts with the cam shafts.

Once removed, we drained the VANOS (onto a diaper for easy cleanup, seen here), disassembled and cleaned it, then reassembled it. The actual upgrade takes only a few minutes, its just too bad that you have to spend several hours pulling things off the engine to get to it. After the new seals were fitted to the pistons, everything gets reassembled and refitted, re-torqued, and replaced.

Sadly, getting this intimate with the Z did reveal some of her other flaws, namely some crusty guff near the thermostat housing that indicates a coolant leak. I'll be dealing with that next, but I'll be passing that one off to a professional because I lack the tools and patience to deal with draining, refilling and correctly purging the cooling system on a BMW.

And finally, after a test drive, I can report that yes, the fix was worth the effort and makes a noticeable difference. There is a noticeable increase in "urgency" from the engine, especially at the lower revs. The transition into VANOS can be felt but is subtle, but overall everything feels more powerful now. Most importantly, I know I won't have to look forward to doing this again in 20k or 30k miles as I would with the OEM seals.


Sunday, April 12, 2009

Guest Room Remodel - Budget Murphy bed with Ikea Pax cabinet frames

Over the last couple of weeks, I've undertaken a complete re-do of our guest room.

One of our key goals was to keep the room multi-functional, so we really wanted a Murphy Bed (folding wall bed) to make room when the bed wasn't needed. Unfortunately, even the cheapest Murphy systems we could find were well out of our price range. With a Queen bed, side cabinets and a rudimentary headboard, delivery, and installation, we were looking at nearly $5000, and that's just for a thin white-melamine cabinet with vinyl-foil doors.

I knew I could do better (price-wise) if I did something myself, but I don't have the equipment or time to make full cabinets or raised-panel doors. Wanting an alternate solution I did a lot of searching and scheming and eventually hit upon the idea of using Ikea Pax wardrobe cabinetry (somewhat modified).

This was a long-ish project, so I won't go through every step here. You can see the evolution of the project along with descriptions of many steps here on the Web Album.

Since the main reason for undertaking this myself rather than just purchasing a basic melamine Murphy Bed was cost ($5000 or so, as of November 2008), let me break down the main expenditures: Queen-size Murphy Mechanism + Foundation .... $390
Floor saver (attaches frame to wall instead of floor)............$80

2 x Pax Cabinet Frames, 20x23x93.................................$180
1 x Pax Cabinet Frame, 39x23x9...................................$100
4 x Pax Birkeland Doors, 19 5/8 x 90 1/8.........................$480
2 x Pax Birkeland Glass Doors, 19 5/8 x 90 1/8...................$240
3 x Gravyr Handles 2-pack........................................$24

Melamine Sheets (Home Depot).....................................$60
2 x Piano Hinges, 72" (ACE Hardware).............................$30

Dimmable Halogen Lights, door-switch, wiring, etc................$40

Assorted Hardware, Hole covers, magnetic catches, etc............$10

"Wrought" Iron Headboard from our old bed........................$0

......................................... Rough Total ..........$1634

Even accounting for a few tools that I had to buy (<$50) and a big pile of Komplement accessories and sales tax, you can see that I came out way ahead.

So, what did it take to pull all this off? Surprisingly little.

As far as tools, I needed a Circular saw, a Table saw (borrowed from a friend), Cordless Drill, various drill bits including 15mm and 35mm forstner bits (the tools I bought), hammer, screwdriver, and various measuring tools, squares, and a few clamps.

This whole build was based on using standard Ikea Pax wardrobe frames. This meant I could fit stock Birkeland doors, which in my mind was the biggest hurdle to overcome. The only real trick to this is that the Pax frames only come in two widths: 19 5/8" (1-door wide) and 39 1/4" (2-doors wide), whereas a Queen bed and frame needs roughly 65" clearance. The solution was to "extend" a wide cabinet to a "double-wide" cabinet and make the doors bifold (4-doors wide).

To do this, I made new top and toe-kick panels in the adjusted width. Using the original top as a guide, I machined my new top to have the same pockets for the Ikea cam-lock hardware so that everything would assemble the same way as the original cabinet frame. I then added a horizontal stringer bolted to the wall to keep the top piece from bowing under its own weight. The toe-kick was similar, except that instead of a full bottom to the cabinet, my toe-kick is only topped by a strip that fills the gap to the Murphy base frame (the center cabinet is "bottomless" and the Murphy Bed frame sits directly on the floor, bolted to the wall.)

Attaching the doors is nearly identical to the Ikea method, except that given the added weight of a second door panel, I felt compelled to add three additional hinges on each side. These were mounted just like the regular hinges into additional recesses I drilled into the doors.

The lighting is actually not Ikea, because I came across some Dimmable Halogens for a great price at my local Dixieline. Rather than just put in the lights and a dimmer, I also built a switched outlet into the cabinet frame. A momentary switch is mounted such that power to the lights is only "On" when the cabinet doors are open, since I didn't want the hot halogen lights left on by accident when the bed was folded up into the cabinet.

The other big improvement in this implementation vs. the cheap commercial Murphy Beds was the headboard. Since the Pax wardrobe cabinets are deeper than I need (even accounting for a pillowtop mattress), there is room between the wall and upright bed frame to accommodate our spare "wrought" iron headboard. When the bed is lowered, the headboard can be slid forward and bolted to the bed frame with two wingnuts. This provides a REAL headboard for our guests to lean against for reading, etc., something that was not provided by the cheap commercial beds we looked at.

Final results? We're very pleased.

The actual Murphy mechanism from went together very easily and is well built, and their "Floorsaver" is a unique offering, as far as I can tell. Being able to bolt exclusively to the wall studs instead of into the subfloor made things much easier on me. My only knock is against their "foundation", which is really nothing more than a pine wood frame, some corrugated cardboard, and a quilted covering stapled on. For the $110 difference, I would just order the mechanism without the foundation and just make one from plywood, were I to do this again.

The Ikea cabinetry isn't as nice as real hardwood, but it actually feels less cheap and more sturdy than the thinner particleboard/melamine on offering from commercial Murphy Bed sellers. For our room, we actually wanted the white cabinets, but we could have also selected one of the other Pax finishes. NOTE: If using a cabinet finish other than white melamine, you'll probably want to buy one additional Pax cabinet frame. Use the side pieces from this extra cabinet as the raw lumber for the new top and toe kick panels. It will be slightly more expensive that buying melamine in sheets, but it gives you the option of "other-than-white", another difference versus the commercial Murphy Bed makers.


Wednesday, March 04, 2009

Four more lives that could have been saved by PLBs

So apparently this is turning into the 406MHz Personal Locator Beacon advocacy blog, but dammit this is important. Since I wrote last week about the out-of-bounds skier who died due to a cascade of missteps both personal and professional, two more stories of tragedy have hit the news. Once again, both stories could have been very different if someone had bothered to make the investment in a PLB.

Image courtesy of cnn.comFirst, there was the boating accident that saw two NFL players and their friend lost at sea. One man survived by clinging to the overturned craft, but as of today the Coast Guard search has been called off for the other three. The rescued boater spent nearly two days with the craft before rescue. Even with poor weather, the search could likely have started many hours earlier had a PLB alert been triggered.

ACR MicrOFix, just like I carryNext comes the story of two off-piste snowboarders in the Alps. Once again, one of the party survived, this time by using a GPS-equipped iPhone and what seems to be a lucky strike with a WiFi signal from a nearby resort. His friend died, though that may have been the result of a fall. In either case, a PLB might have sped up rescue operations for one or both of them. This is also a good reminder that the COSPAS/SARSAT network is a GLOBAL cooperative effort. In this case, the same satellite constellation would have been notified, but rather than US Coast Guard, or Park Rangers, or the RCMP, the appropriate Swiss authorities would have been notified.

Seriously people, if you're going to be spending any amount of time away from the comfort of your urban/suburban environment, take one minute to think about how much you'd spend to be able to call for rescue in an emergency. How much would you pay to call 911 if you were trapped in a burning building? Lost at sea or stuck in a blizzard is just as serious and just as deadly. Why wouldn't you spend a little money to make sure you could call for help?


Thursday, February 26, 2009

One skier dead, one finally rescued after 10 days

According to a couple of stories from CBC News and, a Woman has died after she and her husband got lost while skiing and spent 10 days in the B.C. wilderness.

The tragedy of this story is massive. I won't recount all of the details (see the above articles for that), but it seems that critical failures of preparation were made by the skiers, and critical failures of response were made by the local volunteer S&R team and the RCMP.

Sadly, any one of these missed steps by the skiers might have changed the outcome:
  • They didn't let anyone know to expect their return that day. No ground search was launched until they were actually reported missing nine days after getting lost.
  • The couple had no survival gear, just two granola bars and no water. Even a basic PSK (Pocket/Personal Survival Kit) would have provided for fire, and thus melted snow for water. An improvised shelter of even basic materials like a couple of garbage bags or a mylar space blanket would have reduced their exposure significantly also, reducing the need for additional calories to keep warm.
  • No GPS or Map & Compass. Getting lost is best avoided by "staying found", but even someone who is massively lost could likely have found a path back to safety given a little time, and assuming they knew how to use those tools.
  • No Personal Locator Beacon. A $450 PLB would definitely have alerted the highest eschelons of both American and Canadian SARSAT command and would certainly have kicked off the search nine days sooner, no matter what mistakes were made at the local government level.
As for the local volunteer S&R and RCMP organizations, I'm sure they will be looking long and hard at their own response procedures to determine why multiple SOS signals were ignored. The actions taken by the Golden & District Search and Rescue may even qualify as criminally negligent, and their response to the RCMP inquiries certainly hampered the RCMP response.

< Soapbox Mode > Its obvious that mistakes were made all around, but the skier's failure to prepare was the set that might have had the biggest direct impact.

Please please please, if you ever spend ANY amount of time in the outdoors, even as "civilized" as a ski or beach resort, consider your surroundings and circumstances and prepare for these kinds of events. And while you're preparing your few essential survival elements, please consider investing in a 406MHz PLB. I'm sure Ms. Fortin's family would agree that $450 is a small price in exchange for a life saved. If you have any doubt, let me point you to this other timely article: NOAA Satellites Helped Rescue 283 People in 2008

< / Soapbox Mode >


Monday, February 23, 2009

Science Olympiad: Junkyard Challenge

After I coached the "Robot Ramble" team for Mount Carmel High School's 2008 Science Olympiad team, I suppose there was no chance I'd get out of coaching an event for 2009.

The Science Olympiad folks served up a new event for this year, "Junkyard Challenge". The event consists of students partially pre-building an autonomous device to complete a pre-defined task, but the students must incorporate one or more "Mystery Materials" into the final build that takes place on-site at the start of competition. Its a great way for the competitors to demonstrate the ability to improvise, and potentially exposes them to a lot of different concepts.

The task for the regional-level competition was "Tipping the Scale", in-which the students must build a device to determine the mass of a challenge object provided during the test.

There were a number of ways to build a simple mechanical device based strictly on Hooke's law or a balance scale, but I asked my team if they wanted to try something more advanced, and they said "Yes!".

I'm proud to say that my team of High Schoolers built themselves an electronic scale, complete with a hand-built 5-stage ring oscillator biased by parallel plate capacitor acting as a transducer to measure the compression of the scale as mass was added. Besides my usual safety lecture and introduction to various machining operations for the mechanical part of the build, I had to introduce the team to some of the basics of electronics, including RC-circuits and the basic concept of an ideal parallel plate capacitor.

The concept for the capacitive transducer is based on something I was exposed to during an internship back in college. Obviously there are better (cheaper, easier, more accurate) ways to detect mass, such as using a strain-gauge load cell, but the competition rules were very explicit in disallowing any components harvested from commercial scales. To avoid any possible problems, I helped them build the whole device from absolute scratch.

I couldn't be more proud, as my guys took 4th place in their event, and MCHS took 2nd place overall, meaning a select portion of the team is headed to the State Competition in Long Beach on April 18th.

See more pictures of the build process in my web album.

Thursday, January 15, 2009

I've gone meta!

I seem to have gone slightly meta, in that I'm writing a blog about my blog, but this is the sort of thing I want to document since I don't want to forget what I've done, in case I accidentally update a template and forget to merge this change back in.

You'll notice some of the long posts here will now get just a summary, followed by "Read More...", which is your indication to click:

I know I'm a long-winded chap, so I wanted to start breaking up the posts into pieces.

After a lot of searching and experimenting, I've gotten it working. I'm using the help provided by the blog Pinkish, but I found an error in their instructions which added a lot of time.

Without duplicating the entire content of the post, I'll point out Step 4 and Step 5 say:

Your splitted posts should look like:
This is part one of my post

<span class="fullpost">
This is part two.

But the correct span tag is <span id="fullpost">. This is critical.