7.4.3 :
mobi`dahon`drive
adding a front
derailer
7.1.13:
. this model of the dahon`folding bike doesn't come with a front
derailer,
but since I had to buy new chainwheels anyway (for mountain bike
gearing,
to get mobi`s 500lb's up the hills ), I had to figure out
how to make and place a cable hold for the front derailer
. I reused a ring that usually applies cable holds around a
bottom.tube
and that was wrapped around the dahon`s rear rack mount and held in
place by a hose clamp and 2 screws .
. install front derailer -- most time was spent figuring the cable
anchor for that .
. the screws were replaced with wire wrap
.
7.3.5: crankset
allignment
. apparently to accommodate the wide tires,
mountain bikes will have wide frames and wider axles,
so that their cranksets are alligned only when farther out than usual
. hence,
to keep the chain from wearing out my chainwheels sooner than later,
I would have to modify the crank
to make it fit closer to the center of the crank`hub
. first I tried ceramic-hacking the crank`hub,
but after that went suprisingly slow
(I remember it being easier getting through a
kryptonite lock!)
I decided to reform the aluminum crank instead,
even though I had no notion of a square file at the time,
and my other tools were going to make a huge mess of it
. after trying to ream it with a large drill bit,
I figured I should first make use of the hacksaw as much as possible
first
by removing the shim-like projection that was built into the design
. after removing all the
chainwheels, that hack job went by smoothly
--. mishap: wish I'd noticed the orientation they used on this fancy
irregular shape;
not sure what diff' it makes . --
. then I used the ceramic-hack rod to do what I should have gotten a
square file to do:
reaming out the inside of the crank`arm .
. after getting the arm alligned,
I realized things would be loose because the hub was poking past the
edge that the nut sits on
. again I tried ceramic-hacking the hub to turn more of it`s axle part
into it`s bolt part,
but there was so much to do as this axle metal is very hardened,
and had no tools for removing the pedal`axle from it`s bearing set
so I kept being slowed by trying to avoid sawing the frame and derailer
. this activity was hacking the bolt diameter too thin to be safe,
so, then I thought about ways to instead raise the edge that the bolt
sits on
. I tried filling that space with tiny ball bearings from an old pedal,
but when you just hope the little bearings land right, they don't;
and, you get a wobbly crank .
. at first I had the idea to keep the bearings in place by packing them
in foam
but then I got lazy and tried to hold them in with tape
... later I realized that the foam wouldn't work anyway
because
cranking down on the nut would cause the whole bearing-foam ass'y to
roll out of position .
. tried forming the assembly out of a perfectly fitting rubber gasket
but
then realized I'd need a stack of them
. tried drilling holes for screws to rest nut on,
but then got interested in bearings being stronger solution,
after so many problems:
finding that I'd have to do some tiny filing work getting the
screw.heads to fit in that space,
drilling on the crank`arm was slow,
nervousness about ruining the crank`set if the holes got too
big,
and being unsure about how big the holes needed to be, or what angles I
could get
away with .
7.3.5: todo" getting the wobble out: [done]
. may do better with a reuse of bearings holder ( clip it to re-form
into right size, then
pull any bearings in the race that are in the wrong location for this
use)
. or,
find things that just fit; eg, putty epoxy .
7.4.11:
. now that a square file is available, see if that could help get the
chainwheels even closer to center,
so the small, front chainwheel is lined up with the large, back
chainwheel .
7.4.15:
. used the square file to widen the hole and grossly overestimated the
removal,
so that the pedal could actually touch the frame of the bike and it
would still be very loose
. I could have removed more of the side of the pedal,
but the left-right positioning of the pedal was not bad,
and I wasn't sure how much more of that I could do without weakening
the crank arm`s grip on the spindle
. a big break came in finding a shim that didn't require machining tiny
parts
(big hassle with my limited tool set)
. after sorting all the connectors the other day,
it was now easier to see that 4 of the square shelving nuts would just
fit (see pict)
. they need to slip between the pedal`crank`hole and it`s spindle
affectively filling the well that the nut would set in,
so that while there would be room for the spindle inside that well,
the nut would be prevented from going down
and then further turns of the nut would instead
bring the spindle
further up the well .
. to recover from having removed too much material from inside the
crank arm,
I applied epoxy with the crank`arm carefully alligned like so:
. the hub-side of the crank`arm is shaped as if there was a washer
welded
to it;
so to get that side both flat and perp' to the axis of the bolt`hole,
I filed until that washer projection was exactly removed .
. I would make the hole right-sized again by filling it with putty
epoxy,
and bolting it into place
. I wanted the bolting to be tight both during the epoxy`setting and
then again during use,
so before bolting the epoxy'd crank in place,
I made a shim out of plastic, so that after the epoxy was removed,
the shim would be removed and bolted product would still have a gap
there
. the spindle was greased to ensure the epoxy wouldn't stick to that,
because removal of the crank is required for installation of new
chainwheels,
and might also make bearing maintenance much easier .
7.4.15: todo"file
chainwheel`flanges and epoxy
. apply chain to check allignment;
but, I can't really see anything, because I need to reinstall the
smallest chainwheel
. while doing that,
I padded each bolt of the innermost chainwheel with 2 washers
to extend it farther inward and ensure a perfect allignment with the
innermost chainwheel of the wheel`sprocket
(this combination is the lowest gear available, and the one most useful
to such a slow-moving and heavy vehicle .
todo:
. the smallest chainwheel needs to have it`s inner projections shaved a
bit:
there is a slight wobble radially,
and on the shallow side, the wheel rubs against the hub
. that is not risky surgery; because,
the projections extend quite a bit
farther inward than is needed for a solid bolt grip .
todo:
. the front derailer needs to be realligned
to accommodate the new
chainwheel allignment .
todo:
. the goopy epoxy should be applied next,
to toughen the putty.epoxy job, and freeze the shelf bolts in place .
7.3.1: direct-drive
. investigate and eventually shelve plans for braking
via direct
transmission .
7.4.27:
nautilus-cammed linear system
is max efficient
. I was noticing that what makes rotary pedaling inefficient is that
tangency is lost;
ie, when near the top or bottom of the stroke,
the pedal force is pointing down when the tangent would from the side
. hydraulics can allow pushing in any direction,
so that the driver`s right leg acted normally, while the left leg did
the work from the side ..
. there are picts showing linear pedal
systems
and the idea behind that is simpler, cheaper,
and lighter than what hydraulics could offer
. with nautilus weightlifting cam designs,
you could do even better than linear,
because legs have better leverage when the joints are not bent much,
so that's when you want the cam to have shorter diameter,
where it puts the leverage into high gear .
-- that should maximize in ways that are obvious but unavailable due to
patent complications ?
tech:
. there's elliptical chainwheels that are like nautilus,
but they are trying to make up for not having linear drive,
ie, since the force is downward, they shape the chainwheel so that
most of the work gets done when the pedal crank is near the side
. a cammed linear system would push a rack across a pinion that was
cammed,
p90 = pinion at 90deg;
. p0: left.knee`angle is accute; so p0 diameter great,
when left.knee`angle = 45, then pinion diameter is decreasing
at a rate that takes into accout the other leg`s pull power efficiency
curve
. when the left.knee`angle=0, p`angle =180,
because after both legs are done, p`angle has to do a full 360 (2*180)
. p180 is approached p`diameter is rapidly increasing,
so that the diameters are greatest at p0 and p180,
but the rate of change differs
. perhaps the optimal way even when considering weight,
is to give each leg it's own rack&pinion so that
the left leg is getting a very short pinion`diameter all the way to
p180,
while the right has a rapidly increasing pinion`diameter near p135,
so that while left is short at p180,
the right leg is maxing the pinion`diameter then .
7.5.17: transmission
fabrication
. make your own transmission parts with bolts and epoxy:
. if the rear cogs are hard metal like the pedal axle,
then ceramic`saw some triangular bites around the inside,
so that using the bites to place a circle of bolts
will hold the cog in place
. oxy-welder can be a good metalforming tool even if welds are easily
contaminated .
