Search all of my sites with Google

Sunday, December 14, 2008

Excessive Cargo Capacity

Does not really exist. :-)

In the late afternoon, I was about to make a run to the grocery store, then realized that a tiny oversight in the new cargo pod design prevented me from using the existing trailer: There's no place to attach the "hitch", which was simply a bar-stem. Oops. :-(

I pondered for a moment, and thought of a bunch of solutions, but none I could make before it would get seriously dark and much colder (been nippy all day, and breezy to windy, making it much worse). So I decided I'd just go for what I absolutely had to have for tonite and tomorrow, in case I don't finish the new trailer (not really started yet, except design) and/or don't get a hitchmount made for the old one.

As you can see below, I didn't stick with the plan:

Besides the five 12-packs of soda you can see (each of which is around 10.5 pounds), there's also four 5-pound-ish flats of chicken breast and other assorted heavy-ish things packed in the cargo pod proper. I'd guess total about 80 pounds of groceries. Nowhere near what I carry on the trailer. I didn't intend to get this much, but they had a very big sale on the stuff, and I didn't want to miss the sale in case I didn't get back up there in the next two days (possible), and as cold as it was, getting darker and colder, I wasn't going to go back for a second trip tonite if I could possibly help it.

That's a lot of weight, but not only did it not have a problem with it, it balanced fairly well! The only part I had any problems with at all were mount/dismount and initial start, since due to the left-biased weight distribution, I could not just stand over the frame and push down with one pedal to start moving, I had to sit in the seat so I could use it for leverage to let my body balance the bike, and start in the lowest gear and quickly upshift from there. Would've been really handy to have the motor about then. ;-)

I took the pics right after getting home, before unloading it.

The cargo pod is in a mostly-finished state in the pics, with hinges and lock installed. It's only a cheap lock off the old electric scooter carcass, and will be replaced with two much better locks (and probably a different lid) and piano hinges, as soon as I have time and find the keys to the locks. There'll be more pics posted later of the pods metamorphosis into the current state, once I make the right-side unit. I didn't take pics of the left side one in-progress mostly because I forgot until too far into it to matter much, but also because I was changing things as I built it, thinking suddenly of better ways to do various things even as I was marking or drilling a hole for the previous way(s).

Saturday, December 6, 2008

Synchronicitous Technology

First, some more pics I keep meaning to take but always forget, of the speedometer system using Veloace and a PDA:

The first is the wheel sensor, simply a badge magnet strip ziptied to the spokes. More secure than a single magnet glued on, so far. You can also see the black cable running down the inside of the fork on this side, which has the reed switch inside the end of it just above the wheel-axle-release lever. Second is the PDA and it's hotsync cable, attached to that ex-monitor cable that runs down to the reed switch. Third is a front-view of the PDA while Veloace is running. Since the motor assembly is all off the bike, so is the meter that was behind it. I'm still working on a design for a quick-release mount for the PDA that lets me securely mount it on there for a ride, yet just slide it right off to take in with me (to prevent theft or temptation).

Much better than zipties, I now have a battery cage to hold my lighting battery in place.

It's from inside a UPS, though oddly enough not from one I had already. This one came from a neighbor, who saw me working with my piles of stuff in the yard and asked if I wanted an old dead UPS for parts, and since sometimes the batteries are still good with fried electronics, or vice-versa, I figured it would be worth a shot. Plus I was just about to measure and cut some metal to make such a bracket for my battery, and had the thought that since this type of UPS happens to use a removable, bolted-in (instead of riveted or welded) battery cage, I could just use that. I opened it up and found it did indeed have exactly the cage I wanted, though both the battery *and* the electronics were toast. I have a couple of other UPSs from Mark that also have cages like this, if I need them later. :-)

Some pics of my first test of the cargo pod idea:

With and without flash, to show the bike lighting visibility (looks a bit different from when it was on the baskets), from the left-rear and left-front. You can see the battery cage under the seat. I don't have the lid on the cargo pod yet, and only have the left-side one on for now, since I am first just doing test rides with it to see about clearances and the way I ride--I don't want to end up with it mounted too low to be able to tilt for turns, but I want it as low as possible to keep any cargo mass in it as far below the center of gravity as possible.

Once I have figured out exactly where the mounting should be, I'll make permanent mounting holes in the other side of the left cargo pod, and matching ones on the right side pod (along with a hole for the derailer, and internal cover for that). Then I'll put hinges and the lids on, which will have locks that will hook under the lip at the corners. The lids will swing outward, so that I can have cargo taller than the lids would allow if needed, and still be able to use stuff wider than the rack allows, at the same time. Also I won't have to hold the lids up while putting stuff in. :-)

Eventually I will also run wiring from the lighting system to some internal lights on the cargo pods, which will come on when the lids are open and it is past some darkness threshold, to allow me to see what I'm doing inside them, like a trunk-lid light, just not in the lids.

To that end, I would need connectors to allow the cargo pods to be removed, and preferably waterproof ones. Oddly enough, these old test-equipment boxes from mid-'90s Motorola have some aircraft-style connectors on them, with about 13 pins. Unfortunately I don't have the mating cables and connectors for them. Even more oddly, I *did* have them, just not where I would have expected.

Along with these (and many other) empty boxes and aluminum parts, Mark also gave me some really cool Boeing 737 cockpit panels made in the '60s. They're in pretty terrible shape, overall, given that they were in the desert plane-graveyard for at least years, possibly longer, before he got them. But they're still useful, and perhaps restorable. Most of the wiring on them has been cut off, but still attached to the harness (with the wires cut from wherever they once led to) were several connectors with plastic ziptied down over them (probably from the factory; they were likely test connectors). Some were 3-pin, but two were 13-pin, and the same diameter, pin style, and shell-keyways as the ones on those test boxes above.

That's just...wierd, given how many possible variations there are on these kinds of connectors, both in diameter, shell keyways, pin configurations, pin sizes, etc., that two pieces of equipment from two totally different lines of technology and different companies, made 30+ years apart, would not only use the same connector but end up in the same room of the same house just at the moment when their matching connectors would be needed.

It doesn't usually bug me, because stuff like this happens to me all the time in my repurposing of things--it's what usually leads me to do some particular repurposing in the way that I end up doing it. :-)

Also, some of the unused screws on the 737 panels (that appear to have been used to mount the panels to the cockpit frame) happen to be exactly the right kind to bolt the panel onto the box. Good, because I didn't have any screws around here that had the right diameter and thread pitch, otherwise.

Anyhow, some pics of the pods.

First is the synchronicitous connector pair, both plugged in and not. There's also a wingnut on a press-riveted-in bolt, probably for ground to the test box. I left it on there in case I need to use it for something later, such as securing the cable hanging from that connector with a clamp, to keep it from dragging the ground. Then a pic of the lid as it might be used on the box. Then just the box itself.

The box's top cover (now the side) has air vents in it, seen in the pic as dark horizontal strips. That cover will later be on the outside, using either rivets or security screws to hold it on, but for now I'm using the vent slits in it as adjustable bolt holes while I figure out the fit of the pod vs my riding style, etc. The main reason I won't use those as permanent mounting points is that this panel is thinner than the rest of it, and thus weaker, and is further weakened by those slits. I'm afraid it will just bend outward under load, potentially breaking off at the bolt-on points. A second reason to mount them with the panel facing out is that it has no lip on it for locks to engage, so I'd have to add something. Since the other part of the box does have lips on all three sides, it's better to use what already exists. Plus the lid hinges only need a single flat surface, not a lip, to rivet to, and that panel does indeed have that.

I'm still looking thru my stuff to see if I can get two piano-hinges out of it, long enough to use for this. So far, I've found one, and am sure I have seen at least one identical one somewhere else, if I can find it in all this stuff. :-)

Thursday, December 4, 2008

Removing The Motor System

So, with the motor system removal, I took the pictures I didn't take when assembling it, showing the brackets and such that hold it on:

They're riveted directly to the frame, and that's been trouble-free except for the one time I hit the bottom edge of the motor's mounting panel on the driveway entrance at a shopping center, trying to avoid a crash, which actually broke bottom rivets.

I changed it so the front part of the bottom bracket is bolted to the back part, instead of riveted, and that hasn't had problems since then--it has less flex.

This is the whole motor system, batteries, charger, motor (minus the one that broke the shaft), and controller. Only parts not there are the brackets themselves, and the throttle, which is still on the bike.

I'm working on getting the treadmill motor's reduction system completed, so that I can use it on this bike until I complete the recumbent it will go on. The batteries are still a showstopper, though, so until I get new ones (lighter than SLAs), I still can't *use* the motor.

Cargo Pods, Modular Connectors, and Some Disasters

The new cargo pods, which will replace the baskets, will be made from these aluminum test equipment boxes. I don't know what was originally in them, equipment-wise, except that it required calibration, as the sticker-seals for that partially remain around the edges.

I don't have the original front panels for them, but I do have the sides/top/back, and have a lot of small panels that can be used in place of the front panels (which will now be the top). Those silver-grey panels sitting on top of them will be attached to the outer edge of the case by piano-hinge, and the inner edge will have two IBM AT-case locks (or similar), one at each corner, to hold them shut. They're not theft-*proof*, but certainly *resistant*, as the casual thief can't just reach in and take stuff out, like they can with my baskets. (And did, by stealing my very last two bungee cords, just wrapped around the baskets not even holding anything on, while I was in Jack-in-the-Box a few days ago. I was watching the bike thru the window, seated next to the bike rack, the whole time *except* for the one minute or less that I walked away for a refill. At least they didn't try to steal something more important, or damage anything.

But it motivated me to finally make good lockable cargo pods, which I've had several posts on before, but never made anything beyond the couple of kitty-litter-pail pods that worked ok for a while. I stopped using them mostly because they were such a pain to open and close, especially after the sun had begun to dry out the plastic enough to make them harder to flex around the seals they have. They're also not very roomy--each of the new pods above should be around three to four times the volume of a kitty litter pail. They also won't get destroyed if I lay down the bike--the pails can crack, or get gouged by rocks so badly that they later crack along the gouges.

The vacuum-cleaner-body pods would have looked nicer, but had far less room. I might still eventually finish making those, if only for an instructable on how to do it for those that don't need much room, but do need a tough pod that looks streamlined.

These two pics just show about how the panels might mount. They actually have to go significantly farther back, to allow my heels to clear them, but you get the general idea.

They are so wide and vertically deep that they'll be a bit in the way for really steep-angle turns, if I have to do one, but since I rarely ride fast enough to need to tilt very far during a turn, I don't think I'll worry about that.

They'll be mounted to flat plates that used to be rack-mount ears for some very heavy 10BT LAN routing rack boxes.

Those rack boxes were steel, but the ears were aluminum. I'm trying to use aluminum where possible to save weight and make things easier for me to work on. For the most part, the slightly thicker aluminum I'm using weighs significantly less than the steel plate I'd need to do the same job, and is much easier for me to work with. For instance, I can hand-drill aluminum for field repairs or modifications when necessary, but would have a tough time with steel, using only hand-tools or at best a small 4AA-battery-operated cordless screwdriver/drill.

Here are the mounting panels being marked, drilled, and riveted into place:

You'll see some cutouts in each plate, necessary to clear the axle bolts on both sides, and to clear the rightside's rear shifter cable stop. The rightside cargo pod will have a similar cutout for the derailer, which will waste a small amount of space in that pod. If I could get a NuVinci CVT rear hub, I could ditch the whole rear derailer and cassette, and not have to worry about that. :-) They're a bit on the expensive side for me, though, and highly unlikely to show up used anytime soon.

I first clamped the plates in place to level them out, make sure they all came out straight and level. :-) I just set a flat plate across the top to check that evenness. Then I marked the left side unit with lines, drilled it's holes (using the existing holes from the motor mount brackets first), then used that plate as a template to drill the right side plate, with them clamped together, to ensure symmetry. Then I clamped the right side plate on the frame, lined back up to the left side plate, and drilled thru it's holes into the frame, and riveted both of them in place.

The actual cargo pods will be bolted to these panels, allowing me to remove them if necessary, or to change them to others reasonably quickly. With the heavy loads I have to carry, these will also help prevent side-sway and wobble, which happens to all the common rack and basket solutions I've tried, which only have connections at the seatpost and the two dropout accessory hardpoints. That wobble can cause me to overcorrect when steering, and while no crash happened because of it yet, I'd like to keep it from ever happening by stopping it entirely.

The top of the panels have the rack-mount tabs on them, which I can bolt a top flat panel to for cargo that won't fit in the pods, and can strap down to it. The panel itself I will try to make so it can be carried *in* one of the pods instead of having to be mounted, when I'm not using it. I might not do that, though, as it may simply be more convenient to have always available.

An add-on I'd put on a few days back is a connector so I can detach the taillight assembly from the bike and put it on the trailer.

It's just a pair of old AT-style keyboard extension cables. One (non-coily) I cut the male end off of, leaving about 8" of cable, and that was used on the taillight assembly itself. The female end of that with it's about 5.5' of wire went to the bike itself, wired up to the headlight module and the battery. The female end was used for that so the pins can't be accidentally shorted out against something if it is dangling loose for any reason.

The other is just a coily cable ziptied to the trailer. For the nonce I just zip tied the taillight to the styrofoam fish-transport-coolers I use for carrying the groceries, during that trip. It worked very well; the only bad part was when the chain link broke (unrelated, told below). :( I didn't bother with a mount for it yet, because I will not be using this trailer much more, as I'm building a new one out of the same kinds of materials as the cargo pods.

The grocery-trip ride right before I decided to take the motor off completely, a link in my chain actually popped apart. It's probably one I'd disconnected before with the "chain-breaker" tool, and was looser just because of that.

I found it because it suddenly got squirrely during shifts even when not under load, and sometimes clattered as I pedalled (but not otherwise). When I stopped, I saw what you see in the images above. Had the camera with me, so I took pics. Not something you see every day. Fortunately. :-(

I didn't have my chain tool with me, but did have a small pair of vise-grips and a pair of tiny Craftsman two-setting pliers, so by clamping the link itself into place with the vise-grips, I was then able to use the offset-jaws off the pliers in the expanded mode to squish the pin back into place. The link was tight and didn't flex well but it let me get home without a disaster. :-)

I think that if the motor were working better (well, actually, if the *batteries* were better), it would not have happened, as there would be little stress on the chain. Since I was pulling my trailer full of groceries, perhaps 150 pounds of stuff on it, and had little battery power left at that point, virtually all the pulling power was going thru that chain.

Another disaster that is both worse (in that I can't just patch it together) and yet better (in that it isn't a critical part) is that today, one of my dogs (Fred, the demon-posessed one) somehow got my helmet out of my bike baskets while I was busy elsewhere, and chewed the headlamp off of it. It is pretty thoroughly destroyed. :-( She didn't damage the helmet itself, though. So now I only have the wide-angle CCFL headlight, until I can build a new helmet lamp for distance and spot use.

Wednesday, December 3, 2008

Speedometer Returns, Motor Fails. :-(

Eil sent me a PDA (Palm m100) that works spectacularly well for this project, and runs on AAAs as well, which makes worries about recharging it moot. I can just carry a spare set with me (usually did for my helmet-mounted headlamp anyway).

You can see it temporarily zip-tied to the bars (actually, to the cheap multimeter I used to monitor battery power after the LED-bar-meter disintegrated from something (Vibration? Someone messing with it while I was away from it? Dunno.)). The two views I use mostly are the regular Speedometer Dial view, with just speed and odo, and the Data View that has Speed in large numbers, then Average Speed, Acceleration, Total Distance (Odometer), Max Speed, Trip Time, and Trip Distance. If you zoom in you'll see an absurd max speed of 46.9MPH, which was from when I tested the IR function with a remote.

I did not end up using the IR version of the wheel rotation counter, because this unit came with the serial Hotsync cable, which lets me simplify the whole sensor enormously. I just wired a the reed sensor directly to a male 9-pin cable cut off of an old nonfunctional monochrome monitor, per the VeloAce VAIF-1 instructions. Then I plug in the hotsync cable to that, when the PDA is on the bike.

I use FlipHack to rotate the screen, as it's easier to mount the PDA this way. One of the HackMaster variations is used to manage the hacks and make them come back automatically after any reboots. The buttons on the left, when VeloAce is running, will select which view it uses. I don't keep the stylus in the PDA so it can't fall out from vibration. Don't really need it for this purpose, anyway.

So now I have a great new way to see how fast I am going, not only to make sure I don't go over 20MPH with the motor engaged while on city streets/etc., but also because I am curious about the acceleration and average speeds, as well as to just test flat out how fast the thing can go on motors alone, now that the batteries are dying (compared to how it was when I started). But the day before I set the bike back up for using it, the top motor actually sheared it's shaft, inside the casing, just between the motor section and the bearing, so the roller end of the shaft did not fall off (and did not appear to have a problem, except that the roller didn't spin even though the motor was spinning).

The motor is probably not repairable, unless I can push the shaft out and replace it with another steel rod that's hard enough to bear the load. When I looked at the shaft's break, it appears to have had a crack in it along the slits they cut in it to insert the C-clip (jesus-clip) on the shaft to hold it between the bearings. That could have been there when it was made, or it could have happened from the side-loading placed on it by the roller against the wheel, which was not always even, such as when the tire goes over a bump and slightly expands the rest of the tire a tiny bit.

This is just some pics of the rest of the motor inside:

It's a little different than I imagined, as the magnets are to the sides of the coils (which are cast in some plastic) rather than circumferentially around them. It's a four-brush motor, with 20 coils. There do not appear to be any iron (or other) cores on the coils, which means the motor is not as efficient (or heavy) as it could be. Still, there's a surprising amount of torque for the small size at higher-than-rated voltages, and no damage to the brushes, coils, or contacts has occured from the tripled-voltage power levels I've been using all summer, and the very heavy loads placed on them.

The only problem they've had was the broken shaft, which doesn't appear to be related to the power so much as the side-loading pressure from the roller on the wheel. Pretty good for a radiator fan motor abused repurposed as a tire motor. :-)

Anyhow, it failed on the way home from work, and since the motors were wired in parallel, I got no torque on the working motor, because there was less resistance for the power to flow thru the broken motor. I had to cut a wire to the broken motor to continue home (just a moment's work, but annoying). When I got home I took that motor off and found that broken shaft after disassembling it. I just left it off the bike after that, and used the one motor left for a few days.

Since the batteries are so worn out, they're not really giving enough power for enough time to make it even worth having them on the bike at all, so day before yesterday, I decided to just remove the whole thing for now. I needed to do this anyway to add my new cargo pods, more of which there will be in the next post.

FWIW, the max speed on the single motor alone with no pedalling was about 8MPH. I could easily maintain about 15-17MPH even with just the one motor with generous but not hard pedalling. The real advantage is in accelleration, however, in that I can get to a start speed a lot easier with the motor than without it. The max speed pedalling with the motor, after the first third of my trip to work, was about 20.1MPH, and that left me straining, and the batteries more depleted than typically.

Now that the motor is off, the best I can do pedalling without straining my knees is about 11MPH, and it's more comfortable around 9.5MPH, regardless of what gear I'm in. Starting up from a standstill is harder, and I generally have to make sure I shift down to at least the lowest rear gear (of 7) and the middle front gear (of 3). If I don't, it hurts during startup, until I get to about 5MPH or so. Worse in higher gears, of course.