Search all of my sites with Google

Saturday, October 31, 2009

Low-Speed Stabilizers, Mk I-a Alpha

A few days ago, I started on this alpha idea for the stabilizers, but I forgot to post about it.

I have a couple of very similar kick scooters, with various little defects. One given to me and I think the other was maybe $5 at Deseret Industries thrift store, a couple of years ago.

They come apart very similarly, into a pile of components that's pretty much the same:

One has easy quick-release-skewer type assembly for all of it, and one uses allen-head bolts and Jesus-clips, which I happen to have the tool for this size, making removal much easier (and safer).

The part that I wanted out of it is the entire steering tube and "fork", with wheel, which also includes the pivot point for folding.

I don't need the handlebar tube, as that's not how it'll be raised/lowered. It'll be self-steering as casters, turn around backwards from the way the scooter uses it.

The pic above shows a wheel from each one, but since they're slightly different sizes I'll put a pair of the same wheels onto the two different forks, most likely. That's about where and how they'll look, just wide enough to hold the bike up by itself, but not quite as wide as the bike is. Just outside the cargo pod rails, under the seat, is how they'll fold up:

That puts them about the same distance as my feet on the pedals, as seen below:

I haven't worked out the lowering/raising mechanism nor the actual pivot axle, but the above pics are the bike standing on it's own with them holding it up, just temporarily bolted to the cargo pod rail (too weak to actually do the job on the road; will probably use dropouts instead), and sitting so the edge of the top of the steering tube is catching on the bottom edge of the cargo pod rail, thus supporting the bike.

With a stiff axle pivot, and a latching mechanism, these should work fine as an alpha test to see if it is worth pursuing further.

The battery fairy does wheels, too:

They're not super-soft, but they are slightly compressible urethane casters with good bearings, and fairly light with plastic hubs. Also are larger diameter (about 4-5"), and over an inch and a half wide. That should roll much better on pavement than the kick scooter wheels, should the idea prove out good enough to continue development.

If I use these for the stabilizer wheels, then I would probably make a stub axle bolted into a single tube that simply pivots on a titanium wheelchair bolt from the dropouts, for each side. Then the two tubes would be welded in an "H" with the bar very low down, and one of the tubes much longer than the other, so I can reach down and use it as a lever to lower and engage the wheels or raise them.

The wheel pivot for caster I am still considering, but probably a second axle with bearing that joins the wheel section of tube with the pivot section of tube, inside the tube diameter.

I might even be able to bolt it all together instead of welding, using pieces of old aluminum crutches, of which I have a few in poor shape (for crutches, but not for materials).

Friday, October 30, 2009

Technical Difficulties Mark II; 4-Pole Motor Installed

First, the good news: The 4-pole motor is finally installed and tested as well as expected.

The bad news: today was filled with "technical difficulties":

--Tire has been going flat so something isnt' being plugged by the Slime. Took wheel off and tube out, to find one of the other patches has a split down the middle--poorly made patch, I guess. Glued a new patch from a different brand, waited the requisite time, inflated it to test, and the new patch has a tiny hole in it's exact center, which of course is right over the original hole. Scrubbed that off and put another patch on successfully this time.

--Put wheel back on, reinflated, rode around the block and SPANG! a spoke breaks. :( Not that that is unexpected, given how heavy this thing is and the fact it's all made of old junk. ;) Since the rim itself is bent anyway, I chose to just start with a different wheel entirely.

--This wheel doesn't have an axle (came that way, with gummy grease; must be pretty old). Take old axle out of old wheel, and find axle is BENT. Again, not a huge surprise, but I never noticed it causing a problem.... Got axle out of an old 10-speed wheel instead, and managed to get the wheel together and trued and whatnot, and reinstalled.

--Another test ride, works kind of ok, but something wierd is happening with the assist motor. I am not sure what it is, but it sounds odd, and that could presage a real problem I don't want to have on the road somewhere. I decide to finally just change the 2-pole motor for the 4-pole I've been putting off for months.

--After much ado, get the 4-pole installed and the bike back together for another test ride. Chain comes off the motor sprocket about three trips around the block later. The torque of the motor (which is MUCH higher than the old one, at least twice as high) is pulling so hard on the chain that it actually is pulling the motor backwards in it's mounting slots despite being bolted down very tightly. To fix that, I'll have to make a different mounting plate that doesnt' have slots but rather only fixed bolt holes, which won't allow for tensioning the chain in a normal way. Not happening today.

--Since the chain will only come off if I pedal harder than the chain pulls, I decide to test it longer without any pedalling, just resting my feet on them instead, as they go round. Seems to be working fine, so I head out of the neighborhood for a longer test. First big bump I hit knocks it off anyway.

--Some 15-20 minutes of messing around with some plastic and tape and zip ties makes a sort of chain tensioner, hopefully enough to last me on the way home (it did), and head back. It's now dark, and I can see sparks in my wiring harness along the top tube. :shock: That's VERY bad. I find that one of the batteries is shifting around and a post is touching the frame, shorting across one battery. :oops: Tightened down and fixed, but now the wiring harness is damaged--some of it has melted together, and my front headlight no longer works. The midships one still does, so I'm still road-safe.

--The lefthand shifter starts to act funny, and wont' stay shifted to the highest gear unless I hold it. Dunno why yet; I ziptied it in place as I only need the lower two for hills.

Speaking of shifters, I also found when doing the 4-pole install that I had inadvertently gotten the cable for the rear shifter inside the hose clamp for the front of the seat-frame connection. It slightly crushed the housing, which explains why it has been getting harder to shift the right shifter, for the rear cluster, and less reliable when it does shift (skips several gears, etc). I'll have to replace that part of the housing. That's going to be annoying.


As for the 4-pole motor, it works quite well, and is pretty zippy off the starting line from a dead stop. I can actually feel myself being pushed back in the seat, with my feet not even on the pedals, if I startup from say, 5th gear from top (of 21).

I am certain that it takes significantly more current than the 2-pole, but both of them are beyond the meter's capability to read (12A on the analog panel meter). I have to put the HF DMM in there to see if it's less than 20A (it's max displayable amount).

This motor should be 650-750W actual power handling capability, designed for 24V. At 36V, it should be able to draw less current to do the same work, so it should be able to handle 27A steady at 24V, and 18A steady at 36V, assuming the low end of that range, 650W. Of course, 18A steady would kill my batteries in less than 15 minutes, assuming I don't want to drain them past 50% DOD. I try not to drain them even that far, where possible.

It does have a distinctive noise though: It sounds like the servo motors in robots on TV shows and movies. It's a little annoying, but I'll get used to it, especially for the better acceleration I get from it.

Because of the higher RPM of this motor's gearbox, I used lower gearing (21 teeth instead of 24) at the motor sprocket to keep approximately the same speeds. I think I must've miscalculated though, because unless something else is wrong and the rear cluster isn't shifting all the way up to the smallest ring only under load/on the road (possible, but didn't seem like it), then it simply isn't going as fast as it used to. 17.5MPH is about it.

Since I had to weld the chainring onto the motor hub, then I'll have to weld a different one on the backside of the hub and flip it over, to get the right max speed (20MPH) out of it. Probably a 22T ring. Gotta go back thru the calculations again.

Ooops.

Wednesday, October 21, 2009

Low-Speed Stabilizers Idea

I've been inspired by another bit of one of AussieJester's projects--stabilizer wheels to let him ride at slower speeds, as well as a stable stand to hold the bike upright when stopped. Last time I was able to nearly directly take the idea and run with it almost just like he already used it (the brake-arm chain tensioner), but this time it'll have to be quite different in execution.

Because my motor would never handle the power needed to climb hills at speeds over 8 or 9MPH, and neither will my legs, I have the bike geared so I could crawl up a hill to do it. However, the bike isn't *stable* at the very low speeds (1-3MPH) that I could climb long hills at, so I would be unable to help the motor by pedalling, as I'd have to keep balancing the bike by touching the ground all the time with my feet.

So I need stabilizer wheels (or a trike design, which is the works but not possible yet) for really low speeds. I also need them to not both touch the ground at the same time, since in Arizona, a bicycle is only a bicycle under the law if it is three or less wheels in contact with the ground. As long as I keep only three of them touching, I'm safe from any legal hassles (it's already wierd-looking enough, I can imagine some problems with misunderstandings already, if someone looking for a reason to issue somebody a ticket that day picked on me).

Since I'd be riding potentially a mile or two up some hills like this, I don't want to use solid wheels, so I'm looking for a busted dolly that has good inflatable-type tires, 6" to 8", even if they have plastic hubs, as long as they have bearings in them. I have the old electric scooter wheels but the hubs are big and clunky; I think they're 10" wheels and would ride nice, but those hubs...plus the tires are relatively narrow, and wouldn't give as much shock cushion on bumps and stuff, but the dolly wheels would, since they're usually 2-3" wide.

I don't have any sort of actuators that'll do this kind of work, so it'll be just a manual lever along the center frame I reach down and pull, along the lines of the emergency-brake lever in the middle console of some cars.

The pic below shows about where they would go, and how they'd raise/lower.

Basically, the swivel point would be at the former rear dropouts of the Magna frame I'm using for the front half of the bike. They're dropouts, so they are strong enough to take the strain, even though I no longer have the seat stays that went down to them that you can see in this pic of an older revision of the bike (the batteries now rigidly fill part of the space between the top tube and the base of the bottom triangle/diamond formed by the chainstays).

The handle would go up and around the battery area, to clear them and the cabling, on the left side (even though it's the right side in this pic), so I could keep my hand on the throttle and shifter if need be.

Both wheels would be joined at that dropout, and angle out kind of like yours, so that there would be around 1/4", maybe 1/8" on either side below one wheel when the other was on the ground, without me on the bike. With me (and posssibly cargo) on the bike, it'd be less to nearly (but not) none, depending on front/rear tire pressure and front shock load. They'd pivot together as I move the one handle.

Not yet sure how I latch them in place; possibly some sort of mechanical detent device on the frame that the handle "snaps" into at full extension, with a harddisk magnet or three to hold the system in retraction when not in use.

One problem I have to deal with using midships stabilizers like this is that they probably will have to be casters--they must be able to turn with the bike, or else I could end up with the one on the ground causing sideways resistance to any turns, lane changes, etc.

If I'm lucky, they'll work as solid axles, not needing any steering/caster pivots, so they'll be built that way first. But I foresee the likelihood Murphy will smack me around for doing that, and make me build casters instead. ;-)

Friday, October 16, 2009

DC-DC Converters

It's been mentioned on the Endless Sphere forums that some AC power supplies might work from 48-50VDC and up, and be able to power various low voltage things (like LED lights, bike computers, etc.) from a bike's traction pack.

I may have found a way to power them from even lower voltage traction packs (like mine, at 24-36VDC) pretty easily, via a cheap (free, actually) source of DC-DC converters, in old celphone chargers and other portable-device AC adapters.

I started going thru my collection of ac adapters, including a couple acquired from Freecycle today, and found a couple of Samsung celphone chargers (TAD137VSE) that output "5.0V @ 0.7A" for "100-240VAC @ 0.7A" which also output the regulated supply at as low as 23VDC input. I'm sure the current draw would be higher at the lower voltage, but I didn't test it at it's full output, only with a 1Kohm load on it (~5mA), which drew pretty much nothing on the source side according to the Sorenson's ammeter.

I tested by hooking the AC input of the charger to the DC output of my Sorenson 0-60VDC adjustable linear bench power supply, the 1Kohm load to the + and - output terminals of the charger, and the DMM on VDC across the resistor. Adjusting the Sorenson's output from 0V upward resulted in output from the charger around 3.5-4VDC starting at about 18VDC input from the Sorenson, swinging up to 5.14VDC output at a 23VDC input, stable at that voltage all the way to the 62VDC max output of the Sorenson.

It is stable at 5.14V with that load or no load; might regulate better to 5.0V at a higher load. One diode in series with it should drop it enough to not worry about for the LED boards on the bike lights.

It's around 3.5W output capability at 5V, about 3.1W with the diode dropping almost half a watt across it at full load. So depending on what the LEDs draw, one of these might easily supply that LED board. Two paralleled would work if one won't, and they're quite small and light, even if left in the original little wallwart cases. If used like that, the long lower-voltage wire from it could be run to the light, rather than the higher full-pack voltage, making it safer if a wire were ever shorted somehow.

Alternately, they could be built into the lights' housings, and just run full pack voltage there (with a fuse or micro-breaker at the pack end).

I have a bunch of others not yet tested that range from 4.5V to 6.0V outputs, and from 100mA to over 1A output capabilities.

Some of the ones tested wont' start until they reach 50V or more. The Samsungs were the first ones I found that work at such a low voltage.

Sunday, October 4, 2009

Nuts and Flags and Pots and Chains

On the stolen nuts problem, a friend (John from Teamdroid) got me some temporary replacements a couple of hours or so ago. They're standard 8mm x 1mm pitch nuts, so they don't do what car lug nuts do, to help center the bolt in the hole. The others were like car lug nuts, with the beveled-in center on one end, which slightly engaged the slots in the sprocket to pin it centered and in one place, no shifting, drifting, or loosening.

The ones he got me'll keep the sprocket on to get me back on the road, at least.

Eventually I need to use the lug nut style again, because those will keep the sprocket from shifting back and forth as the chain tension pulls it around during rotation. That back and forth will eventually cut into the bolts, which is one reason I replaced the original 1/4-20s I was using to start with--I was afraid they'd cut thru far enough to shear off during acceleration, when I need the motor most.

With the lug style, it doesn't do that, because the beveled-in center pushes just a little down into the slots the bolts are in, and prevents them from shifting like that.


In the pedal-chain tensioner pics (which I forgot to post a few days back, I think), you can see the old nuts, but the angle shot is too blurry to see the bevelling I described above.

It's just a large plastic derailer idler wheel bolted thru a derailer idler sleeve onto the old bedframe bracket I had used for the derailer-tensioner before (though it's bolted on in reverse of how it had been before.

The larger idler is much quieter than the smaller one I used to have in that spot, but still makes enough noise to be annoying. Still, it prevents the problem of the chain coming off, so I'm happy enough for now.

I picked up a box of assorted items from Freeycle on Friday afternoon, and among the items were some very bright orange "construction flags", in various conditions.

Just because they fit (perfectly) I stuck two of them in the seat back posts; I could actually ride this way, and then pull them out and lock them into the cargo pod when I leave the bike, so no one steals them.

I also took the wooden handle dowel out of six others, and temporarily duct-taped them over portions of the bike. One up front, one on each side of the middle.

Thinking about cutting one in half and putting on the front of each cargo pod, and another on the rear.

I also put one on the side of each cargo pod. Gotta see if I can find that stick0n lettering I had around here someplace to put the blog address back on the pods, right on the flags.

Might be able to iron-on a logo and lettering using a mirror-image laser printer output, then wash the paper off (same way many home-made PCBs are done).


Also posting some extra pics of the throttle assembly as it is now, for the curious:

The pot was a horizontal control out of some old monochrome monitor. The knob came off some old medical equipment modules. The bracket it's on used to hold a seatpost rear reflector. The spring came off of...something. I forgot what.

I just use the side of my thumb to move it and hold it; my bars are essentially like beach cruiser bars, but since I am seated leaning back some, they are not horizontal; more upright, making it more comfortable to have my hand in that position.

Note that my 2QD has a setpoint pot for throttle, allowing me to use quite a range of throttle control types and still adjust the input to the ~4V full scale needed for the actual throttle input at the comparators.

Vandalism Could Kill

While I was at work today (Saturday), with the bike parked inside our warehouse at the back of the store, someone decided to steal the nuts off the motor output sprocket, which could have gotten me killed if the sprocket had come off when I was trying to accelerate into traffic.

Fortunately it came off during a left turn at an empty intersection, so when the chain tangled up because of it, I could let it glide harmlessly to a stop.

I debated whether to walk to a nearby house and ask to use their phone to call the police at that moment or not, and have them come by and see if they could dust it for fingerprints or something, but I realized that A) they would probably take hours to come since I was in no immediate danger, which B) would put me in danger of someone coming up to me and mugging me (as happened in 2003 when walking home one night), and C) they probably wouldn't do anything like dust for fingerprints or check for DNA but would probably instead just take the report, shrug, and do nothing about it. So I decided not to waste the time of either of us.

I checked my cargo pods to see if perhaps someone was playing a very dangerous and impractical joke on me, by putting the nuts in there after taking them off, but no, they did not. They were removed completely from the bike, and it appears my own tools that I carry on the bike may have been used to do it, because the toolbag was unzipped. I am uncertain if any of the tools are also missing, as they are not kept neatly in there, but they did not seem to be all present. I just don't remember which ones aren't there, and I won't find out until I try to fix something later only to find I no longer own a tool to do it.

At that point, whatever it is will be unrepairable and the bike will be scrap, along with any other project I have requiring that tool, as I can't afford to buy new ones.

Since the nuts for this were special in shape, and of a thread and size I have no others of, the bike is currently useless. I can't just go out and buy them somewhere--I don't have money for that, even if I knew for sure what they came from, or exactly what size and shape they had been. I'd have to guess, and try different ones until I got the right ones. Dangerous and even more money.

If I was willing to live with the knee pain, I could still pedal it, but it is so heavy even without the batteries and motor (which I would have to remove) that I can barely get it started while balancing it, on pedals alone. Sometimes I cannot balance it in this situation, and it curves wildly to one side while I try to recover, which on a street with traffic is likely to result in a collision.

So kudos to you, whoever you are that vandalized my bike. You've essentially rendered moot 3/4 of a year of work, and several other people's contributions, and destroyed my ability to travel around, get groceries, and go to work. Thanks.

If you see me limping for miles around town in the heat, wave from your comfortable transportation, ok?