2005-04-26

HOWTO Build Your Own Bicycle Rotisserie: Part II

In Part I I discussed the inspiration for the bicycle rotisserie and laid out a bill of materials. Now we get down to the nitty gritty of actually building the thing. So without further ado… The first thing you need to do is pick a master bike and a slave bike. The slave bike doesn't need to have a functioning chain or sprocket setup, or a rear wheel. So pick whichever one has a nicer chain, sprocket, and dérailleur to be the master bike and remove its shifters from the handlebars. Next, remove the front forks from the bicycles. The way to do this varies from model to model, but typically there is a small allen-bolt in the center of the top of the fork tube, which you unscrew. This releases the tension from a small wedge mechanism that keeps the handlebars and fork together. With this bolt loose, you should be able to remove the handlebars and fork from the bicycle. Sometimes a gentle tap on the protruding bolt head helps unstick things. You'll also want to remove the seats from bicycles and the seatpost from the seat itself. The seatpost is usually held to the bike with a small pinch bolt, or with a quick-release fitting, depending how fancy a bicycle you're dealing with. Loosen the bolt or quick release and pull up on the seat. This may require a bit of coaxing, twisting, or WD-40 to accomplish. The seat itself is usually also held on with some kind of pinch bolt, which should let go of the seatpost when loosened. If you're lucky, the top of the seatpost will nest nicely inside the spit you've acquired. To allow for easy removal of the spit from the rest of the contraption, I used the seatpost ends to allow the spit to be unbolted. Basically this involved inserting the end of the seat tube into the spit and drilling a 1/4" hole through both. Later I inserted the 1/4-20 bolts in this hole and fastened it with the nuts. Next you'll want to cut the tops of the seat posts off so that only an inch or so of full-diameter tubing remains. You'll also want to cut off the pedal crank on the chainring side of the bottom bracket and weld the stub onto the chainring where the crank used to be. Putting some effort into getting it welded on perpendicularly will pay off later, but it doesn't need to be absolutely perfect. The remainder of the seatpost needs to be welded onto something that will give the bicycle firm footing when placed upside-down. I used an 18" section of square tubing welded crosswise onto the end of each seatpost. Once this is welded up you can re-insert the seatpost. Ideally you'll want the seattube and downtube on each bicycle to form and isosceles triangle with the ground, but you can vary the height somewhat by using the built-in seat-height adjustment. You need to find a suitable motor to power the thing. In my case I was able to visit the local recycled building materials place and pick up a used blower motor for $5. You want a single phase 120V induction motor with a rating of around 1/3rd horsepower (250W) and a speed of around 1500RPM (a 2-pole motor). A small pulley might be helpful, but I got away with just using the shaft of the motor to drive the belt with a little stick stuck through a hole that happened to be in the end of it. You'll need a fairly long belt (around 10ft/3m) to go around the rear wheel (from which you'll need to remove the tire and innertube. What worked for me was a length of surgical rubber tubing (available, oddly enough, from fishing supply stores). To make it into a loop I just stuffed one end inside the other and super glued it. Amazingly, it held for the duration of the cooking event. About the only thing left to do to the bikes is to defeat the freewheel and lock the chain tensioner in an appropriate position. You'll be turning the wheel in reverse, and driving the wheel rather than the pedals. This is a sort of double negative that means the freewheel doesn't really come into play. But the meat you're cooking is rarely perfectly balanced, so it'll want to lurch forward once every rotation unless you lock out the freewheel. This is as simple as running some wire around one of the sprockets and through the spokes. The chain tensioner doesn't work right when driving things backwards, so you'll want to lock it out as well, which can likewise be accomplished with some creatively placed wire. At this point you'll be ready for a dry run, followed by the actual cooking event. Stay tuned for Part III.

2005-04-15

HOWTO Build Your Own Bicycle Rotisserie: Part I

A couple of years ago one of my good friends married a very nice young lady at a lovely ceremony in his back yard. Dan is Jewish and wanted to have a kosher wedding feast, which meant, among other things, that food had to be prepared at the house (well, strictly speaking, any kosher kitchen, but those aren't very common outside the Jewish community). I'll let Dan explain:
When Dan's parents first moved to Fargo, ND, they celebrated their new home by roasting a whole sheep--a gift from the NDSU agriculture department--behind the house.  Dan loves the pictures of this event, and is only sorry that he was too young to appreciate it.  To compensate, he'll be re-enacting the affair, roasting a whole kosher lamb (actually, six forequarters of lamb).
Anyway, we needed some way of actually roasting said lamb parts. I should say Dan needed some way of roasting said lamb parts, but he's quite good at getting other people to do work for him. Real management material, that Dan… So in any case I was charged with the responsibility to devise a device to roast eight 20-pound forequarters of lamb on the wedding day. No pressure… Now the first thing any sensible engineer would think is "outsource." Not as in rent-a-coder outsourcing, but as in "buy something off-the-shelf that will work." But it turns out the only thing really commercially available is a pig roaster, which just doesn't quite cut the mustard in the kosher department. It also carried something like a $150 per day rental fee. So the next thing any sensible engineer would think of is "improvise." Any improvisation, however, would have to have a relatively low likelihood of resulting in catastrophic collapse, since we're talking about the main dish here. I don't remember the exact thought process, but somehow I hit upon using a couple of old bicycle frames to do the heavy lifting, as it were, in building the rotisserie. It turns out to be a relatively ideal solution in case you need to crank out a sturdy rotisserie in your spare time given a month or so to prepare. The advantages of this approach are manifold, but primarily they are:
  • The bottom bracket is designed for a heavy-duty offset load
  • You get built-in gearing (or as my friend Craig would say, "it come widdit"). If you run a belt around the rim of the rear wheel, you get an even greater gear-down.
  • The seat posts allow for height adjustment
  • Useless old bicycles are available nearly free of charge
What you'll end up with is both bicycles upside-down, resting on the fork and the seatpost (with a little "foot" attached). The spit will run from one bicycle's bottom bracket to the other's, and one of the bikes will have a belt driving it's wheel (sans tire), which will in turn drive the spit using the bike's chain. So here's what you need:
  1. Two bicycles of roughly the same (preferably small) size, and permission from the owner to render them useless (as bicycles anyway). I used a couple of Murray kids' mountain bike frames
  2. A spit -- one-inch-diameter (preferably stainless) steel is ideal, with a wall thickness around 1 - 1.5 millimeters
  3. A few scraps of steel, such as tubing, I-beams, or angle iron
  4. A hacksaw or reciprocating saw with a metal-cutting blade
  5. A welder and associated paraphernalia
  6. A drill with a 1/4-inch bit
  7. A pair of 1 1/2-inch 1/4-20 bolts with nuts and washers
  8. Fire!! Fire!! heheheh
Now that I've teased you with the ingredients, I'm going to end Part I. In Part II I'll talk about how to put the thing together, and in Part III discuss the actual logistics of cooking with it.

2005-04-08

Product Idea: GPS Business Trip Logger

This is something my dad asked about as a potential birthday gift, and I haven't found anything out there that meets the criteria for what he wants. My dad, like many people that do a lot of driving on business, needs to keep track of how many miles he drives on business-related trips. In the U.S., you can deduct most business-related trips at something like 35 cents per mile driven. Typically the first and last trip of the day (i.e. driving from home to the job site and from the job site back home) aren't deductible, but the intervening trips are (with the obvious exceptions of non-business meals and personal errands). The IRS isn't satisfied with a total mileage in the event of an audit. Rather, they expect the starting and ending odometer reading for each trip to make their job possible in such an event. There are a couple of solutions that exist for this sort of thing, but none that are entirely satisfactory. The simplest is to keep a little notebook in the car, and before each trip mark down the starting and ending odometer reading. This isn't particularly difficult, but it's quite easy to forget, and something that you're likely to forego entirely when you're in a hurry. Another possible solution (one that I did a bit of work on a few years ago) is something like the Davis Instruments CarChip. This is a little device that plugs into your (1997 and later U.S.-spec) car's OBDII port (a little socket under the dashboard that looks like a plastic version of the parallel port on the back of old printers). OBDII stands for On-Board Diagnostics Level II, and it was originally intended for emission-control-related tools to be able to download diagnostic information from the vehicle's ECU (the "brain" that decides when and how much fuel should be injected into the engine, and when to fire the spark plugs, among other things). When a scan tool is plugged into this port it can talk to the ECU using a special serial protocol called J1850 after the SAE specification that specifies it (or ISO-somethingorother on some vehicles). It can ask the vehicle a bunch of questions as defined in another SAE spec (J2178 for those keeping score at home). Among the questions a scan tool can ask the vehicle is "how fast are we going?" The vehicle answers (within 100ms, but typically much faster), "20 miles per hour" (as far as I know they still use customary units). By doing this repeatedly, the CarChip integrates how many miles you've gone (at least I think that's how it works -- there may be some way of directly querying the mileage of the car at any given time). It also has a real-time clock, and it can know when you've turned on and off your car (when it receives or loses power). By putting all these variables together, it can get a pretty good idea of what constitutes a "trip", when it started and ended, and how far you drove. Unfortunately, there's no way to know for sure where you went. What I think would be an ideal solution would be to use a GPS receiver to log where you went. A GPS receiver is very good at telling you what time it is and relatively good at telling you where you are. It can't directly measure when you start and end a trip, but really, that's not what we're concerned with -- We want to know whether you're leaving home, work, or a client's site, and how far you went, and whether its the first or last trip of the day. What I'm picturing is a little dongle you plug into your cigarette lighter (I guess they're calling them "power outlets" these days) that simply records where you went and when. At the end of the month, you take it out of your car, plug it into your computer, and some semi-sophisticated software finds places at which you tend to stop a lot, shows them on a map, and asks you what they are (i.e. "Home," "Work," "Strip Club," etc.). Once that's been established, said software could translate that into trips, and either determine automatically or ask you whether a particular trip was deductible for business purposes. As for how to implement/prototype it, there are relatively affordable Bluetooth™ GPS receivers available. There are also very small computers available with Bluetooth interfaces. One could run a small Linux system that would save all of the waypoint data off to an SD card that you'd plug into your PC and download the data. I'd be interested to hear any alternative ideas for how to approach this.

2005-04-05

KCA-420i Experiment/Review

I recently bought an Alpine KCA-420i iPod adapter from the local Best Buy and an Alpine 9825 CD player (as a closeout from Circuit City) to control it. I don't do a lot of driving during the week, but on weekends I often head up to the Great White North, eh? Radio reception gets pretty crappy between here and Vancouver, especially the NPR station that I know and love. And right now they're doing their spring fund drive, so the iPod feature has gotten a bit of extra use. The good news is that the KCA does what it says. The bad news is that the usability is crap, even if you haven't been spoiled by using an iPod. The reason for this is that the device is essentially a hack — it makes the iPod look like a CD changer to the head unit, and since the iPod is not a CD changer, the usability suffers. Give a literate six-year-old an iPod (better get AppleCare!), and he'll probably figure out how to use it in short order. In fact the UI is so intuitive that the instruction manual that comes with the iPod is basically a waste of paper. I've got a hard-earned Bachelor of Science in Engineering and I couldn't figure out how to get the thing to work without reading the instructions (and let's just say that even if seven languages weren't included in the manual, it'd still be too thick to staple). By far the most infuriating bit is trying to scroll through a selection of songs, artists or albums. The scrolling is done through a little rotary encoder on the faceplate. But no matter how quickly you twist the thing, things scroll by at around 2 lines per second. There are three ways to "search" (read: scroll through at a 2 lines per second) for songs: all playlists, all artists, and all albums. The playlist "search" is more or less usable (more on that later). The other two search modes are strictly emergency-backup material: scrolling through a hundred artists at 2 lines per second while driving is an invitation to having a horrible accident. Same goes for albums. And both are made worse if some or all of your library is downloaded (legally or otherwise) rather than ripped from CD. When you go into an artist playlist, the songs come through in one big array, ordered by album and track number, so if you have a few hundred tracks by one artist, it's still a PITA. To their credit, Alpine acknowledges the above and recommends making all songs accessible through a playlist. I could do smart playlists by genre, but I'd end up with around four slightly less useless playlists ("Electronic," "Alternative" and "Rock," plus "Everything Else"). Instead I've decided to create eight (point one) playlists that are alpha by artist, then album, then track number (think telephone keypad). I'm not sure if this solves much, but I'll get back to you in about a week after I've had some time to try it out. In any case, there's a glaring business opportunity for some boutique car stereo manufacturer to improve on this. I picture a nice fat rotary encoder (maybe the force-feedback one I proposed on halfbakery.com), an FM tuner chip, a bitmap LCD display that doesn't suck (IOW, not like the 9825's), and an optional amp (since quite a few high-end audio types use discrete stuff nowadays). That and either an iPod dock on the front of the thing or a cable to run into the glovebox. The iPod remote protocol has been reverse engineered and in any case you could probably get it from Apple if you look like a perfectly legitimate enterprise. This is the sort of thing you could program a Gumstix to do. Why not leave out the iPod altogether? Well, first off you need a way to get the music from your (friend's) CDs or the internets onto your car stereo. Secondly, you can take your iPod with you (both to keep it from being stolen or to listen to music whereever you've driven to). Third so that you can buy from the largest online retailer of digital music and not have to commit a felony to listen to your music on the road. Another thought was why not "remanufacture" an iPod into a detachable-face-plate-like-thingy with more car-oriented controls? But in the end most people spend a fair amount of time listening to radio in their car, so unless you can come up with a good way of making an iPod control an FM tuner, I think it's a lousy idea.