Today I got the actual seat to be used on it mounted:
The seat is not totally finished, as it is still being adjusted and then will need the padding attached and covered against weather. (We don't get much rain here, but we do get some, and it sucks to sit in a squishy puddle your whole ride).
The basic seat is 1" plywood cut into two "L" shapes, curved to support the back/etc as many recumbent seats are, with 1/4" plywood soaked in a hot shower/tub until fairly pliable, then screwed down to the pair of L-supports from one end to the other in alternating pairs of 2" long drywall/wood screws I'd had from a partly-full box saved from a pile of stuff in the alley a long time back (probably scrapped from someone's home-improvement or repair project).
The first screw would go in closest to the end I started from (the top, just below the headrest space), threaded in about 3/4 of the way, then I would begin the next screw until it's head was pulling the plywood down a bit farther than the previous one, and alternate them that way until the thin plywood was curving and seating fully onto the thick plywood supports. (I did both sides at the same time in pairs like this, so it was actually the previous step times two each time).
At two points on deeper curves, as I was nearly done at the deepest point, the thin plywood cracked, meaning I had not actually soaked it long enough (only a few hours--next time I will do it for a whole day or two, at least). Alternately, it could be that I curved it too quickly, and that if I went slower, instead of taking about 30 minutes to screw it down, I might take long breaks between sets of screws, and let it soak more during that time, to help it reshape itself to the curve without cracking. By the time I was totally done, the two cracks had gone all the way thru the wood, so it's actually 3 pieces on there now. It still holds the shape correctly due to the thick supports, but it will not be as stiff as it should be, and may flex a lot more than I intended during pedalling, wasting energy and possibly not being as comfortable. I can fix the cracks by bonding more thin plywood to the backside, either by glue or by screws and washers, but I may just leave it until I make a new seat, which will happen after I've discovered the flaws in this one.
I got the basic shape from looking at many many images of existing seats and bikes on Google's image search, then adapted it to my own body's fit as best as this first version can. To adapt it, I first marked one with basic measurements of my own body and cut it out with the jigsaw, then fit it against me in a seated position (while on the couch, with the support in the groove between cushions), and marking areas that protruded too much. Then I cut, ground, and sanded down those areas as I found them, repeating the process until it fit as well as I had patience for, over a couple of days. I then used that one as a pattern to trace and cut the other one from. I did not do a very exacting job of it on the backside, as I expect that to change as I decide what to mount on it, etc., and will probably end up cutting and drilling notching it here and there to fit things on it, such as the rear lighting assembly.
The cushions you see on it are there to let me test the fit with padding, as it's not shaped to be used bare, although it is certainly comfortable enough even like that compared to any regular bike saddle. :-)
You might think "that seat is going to increase the air resistance and slow you down", but the truth is that my body is wider than that seat (not by much), so it's actually going to be at worst the same as if I were riding upright in the Columbia's Comfort Bike Saddle position. But at best, it will be a definite improvement, since I am actually laying back a bit, so the wind should flow over me better (probably about the same as with a partly-crouched riding stance on a regular bike saddle, such as you could do on dropdown "racing-style" handlebars).
Currently the seat is temporarily held on by a pair of radiator hose clamps at the rear, one on each side, thru a hole in the seat supports just above the white ex-fork where it is joined to the seat stays. The clamps go around both the ex-fork and the stay on each side, to stiffen the rear of the seat in place. At the front, there are two wooden blocks as spacers between the top tube and the seat supports, with a long bolt thru all the wood just under the bottom edge of the top tube, preventing the seat from moving at the front end.
Since this arrangement likely won't support much in the way of torque against it, compared to the length of the back of the seat, I'm still working out a setup under it like that used on the temporary saddle a few posts back, using a modified seatpost clamp for the front end. There will possibly be a second clamp on the back supports, holding it to a new tube that would be welded to the meeting point of seattube and stays, but first I need to find a way to bend that tube to curve around the back of the seat from the bottom. This tube is a more ideal solution, as it gives a much better support, and would eventually allow me to create a rear suspension system that would be problematic at best to do now. Also the tube would be a mounting hardpoint for rear accessories that would be better than the back of the seat.
This is what it looks like from the rider's perspective looking down at the handlebars (though because of field-of-view limits in the camera, the pic was actually taken from behind the seat).
You can see the ex-Honda-gas-scooter handlebar controls, and the Shimano Deore XT thumbshifters from a bag of parts I got via a posting on the local Freecycle.org mailing lists along with some frames and other stuff. The cabling from everything is not hooked up yet, and the still-connected gripshifters I decided to abandon because of their feel are hanging loose. Even if I had spare gripshifters like the ones on my Columbia, I wouldn't really be able to effectively use them with the Honda controls on there anyway, unless I modified both the gripshifters and the Honda controls (especially the twist-grip throttle) quite a bit.
Note the mirror *below* the left handlebar grip--it is there for the same reason I did that on the Columbia--if I have to turn the bike over on it's handlebars to fix a tire or something, it will prevent me having to take the mirror off first, or damaging it. It isn't quite inboard enough to be kept safe in a crash or skid, though, at the moment. The mirror came from the same bag of parts as the thumbshifters.
The handlebars still need to be raised about 7 inches, and the stem flipped rearward, first to enable my knees to clear them during the top of the stroke during turns, and second to make them easier to hold during rides while fully seated back against the seat, especially during wide turns. To do this I have to build a riser out of another stem plus a bit of bike tubing, then insert this stem into it, or use a clamp-on stem I have from the same parts bag as the mirror and the thumbshifters.
This shows the clearance on the 24" rear tire between the seat supports, and to the seat body.
There is *almost* enough room there to use a 26" wheel instead, but I might have to sand away about 1/8" or so of the body at the deepest part of the curve (base of the spine area) in order to ensure it clears all the time and never rubs. It may also be difficult to insert or remove the 26" wheel into the dropouts without loosening the seat from the frame at it's rear supports.
With the motor removed until I get a longer belt, I've been testing out alternative chainlines for the pedals.
This one looks crazy, and doesn't work very well either, especially under high torque loads, because the top derailer impacts the frame with it's forward roller. If not for that, it would work ok, I suppose, but is pretty complicated for what should be a simple chainline.
The catch is that due to the motor chainline *having* to be inside the space of the dropouts, and thus on the inside part of the rear receiver axle, the pedal's freewheel on that receiver axle ends up slightly outside the straight-back chainline the pedal chainring would like. Because of the way the frames fit together, the dropouts of the Magna frame, which hold the motor's middle-stage torque conversion/freewheel device, end up in the way of the direct topside pedal chainline, so it has to be run under or over that at some point.
Over it means it must run inside the frame, and thus takes away room I could be using for a storage or battery compartment, and has to clear the frame so close to the rear part of the chainline that it causes the pedal freewheel to skip under high torque, and sometimes to come off under bumps.
Under it is better, but must still clear the frame at the midpoint, and must have something to hold tension in the chain and align the chain between the two out-of-line sprockets (the pedal chainring and the receiver freewheel). The top derailer in the above pic is there to do the first, but it's spring-action causes the problems I have with it. So if I build a simple roller (for instance, out of a roller skate wheel with bearings, on a hanger from the frame above), then use the derailer as it is on the bottom, it is probably the simplest chainline I can get without modifying the Magna frame.
At some point, when I have found enough things that require frame modification, I'll build a new frame for the whole bike, taking each of those things into account. It will be a much simpler frame, not being from the aborted idea of a no-weld recumbent by bolting two bike frames together. :-) As an aside, if it were not for the motor and the torque/speed conversion that has to happen in two stages due to the amount of change needed, I would not even need to *have* the dropouts/etc from the Magna, and could remove them entirely. In the future version, I will have a different arrangement of this conversion, and thus will have a cleaner chainline, too.
I mentioned before that this has a non-1:1 steering ratio due to the different length of the pivot bar at each end of the steering tie rod.
You can see here what effect that has on moving the bars vs fork/wheel movement. The bars don't have to move as far to turn the wheel, which means I have less knee/body interference to deal with at sharper turns. Each pic shows the maximum right/left steering possible as it exists right now. I don't think I need anything more than this, based on my experience with the Columbia upright bike (and other bikes over the years). If I do, I can modify the pivot bar length or angle/position and the length of the tie rod to clear the headstocks and allow for a greater steering angle range.
I still haven't done anything about the slop in the tie rod pivot points, nor their weakness in a situation such as the one that bent one in the first test ride, as I have not come up with a fix I can do with components I already have, except for one that would leave it difficult to adjust the length of the tie rod if I need to later. This would involve welding thick steel tabs along the sides of the eye stem to the tie rod end, so that it cannot flex sideways (and thus cannot bend). This would require modification to the pivot bar design at both ends, so that these tabs would still clear the headstocks enough to allow the steering range I currently have.
I've also been thinking about the lighting system and how I'd go about making the casings for them. Since I have the Honda scooter for parts anyway, I might as well use it's DOT-approved reflectors, lenses, and casings for all the lighting at the rear.
It's also on a handy rack, which I can mount on the back of the seat to hang a backpack from if I need to. Currently that rack is just hanging on the headrest supports, and the turn signal/markers on it's booms are thus pointing up and forward, whereas they will actually be pointed rearward. The taillight is also just sitting there, perched on the rack frame, but it is in about the position and angle it will be at when I'm done. That puts it about at the same height as an average car's driver's eyeline, making me very visible to the majority of those on the road, even if I'm in a regular lane such as a turn lane in a line of waiting traffic. I have on very rare occasions not been seen as well as I'd expect in those situations, even with my existing lighting and signals, because those are currently at about grille or bumper height on the average car, being just above and behind my 26" rear wheel on the Columbia. Putting the light up higher should fix that.
The headlight will remain a CCFL, since that has worked very well for me so far. This time it will be a nearly square unit, since I don't have the handlebars up front to protect it's ends. Again it will be a former scanner slide/film/etc adapter, in this case the same one I show in disassembly at my first Instructable (about the bike lighting system), at http://www.instructables.com/id/Bicycle_Safety_Lighting_and_Turn_Signals_From_Mos/
It's just perched there on the top of the fork for the pic. When done, it will have marker/signal lights on it's side edges, just as my Columbia unit does. The difference is that I will use bicycle pedal reflector covers over those, so that I have amber reflectors as well as amber lights for those corners. I *might* instead decide to simply put the LEDs inside the front Honda scooter signal/marker lamp units, which look pretty much like the rear ones, and are also on metal booms I could attach to the fork. They also have integrated reflectors in them, also DOT-approved, as the bike pedal reflectors are (but are meant for this purpose, as the pedal reflectors aren't, exactly).
I've also had an idea for making a simple and very cheap partial fairing for the bike, including enough of a canopy to keep drizzly rain off of me. Since it won't work to just bend over forwards with a raincoat on to help keep rain off of my legs and such, since the seat lays me back, with my legs quite forward, then it's definitely something I'll want now and then. Most of the time there's not much wind involved with the rain, so it tends to fall almost straight down in just a constant drizzle, when we have it. When it's really windy, I try not to be out in it, as I can't safely control even a normal bike very well in gusty conditions with motor vehicle traffic slipping all over the oily wet road around me, driving even faster than they normally do (for reasons I still can't comprehend).
If I cut the flat cylindrical sections from 2 and 3 liter soda bottles, I can cut and tie or pop-rivet them together to create the shapes of the fairings. Where necessary, I can make frames for the shapes and attachments to the bike from wooden or PVC scrap I've collected. It won't be as smooth or work as well as a single-piece fairing, and certainly not as stiff as polycarbonate or something, but it *will* be very easy to repair when damaged, and materials other than zip ties and/or rivets are freely available as trash or scrap.
I might even be able to melt the sections together, but I'm not sure how well that will hold, unless I build a press to do it with. I have some heating elements from various things including fusers from scrapped copiers and laser printers, and I might be able to build a press from those. It will take some doing to test the theory out, so I will almost certainly make the first version ziptied or riveted together, and try smoothing options later.
Tuesday, February 10, 2009
Today I got the actual seat to be used on it mounted: