More CrazyBikeness

Today was the first of four days off in a row; ideal time (and weather) for doing as much work as I can on this project, perhaps even getting it to a road-testable stage.

I finally figured a way to connect the two frames strongly, but I cannot do it with the parts I have available without welding a few things. :-( I had really hoped to be able to do so without that, but I'd have to either go buy things or I'd have to wait until I found them, and neither of those is a good alternative, since I already have everything I need to weld the parts together. It will still be able to disassemble relatively quickly into two halves, which was part of my no-weld goal, though the welding means it will be more difficult for someone else to replicate, if they don't have a welder handy (though I imagine if they could build this, they could come up with a DIY welder easily enough).

The first thing needed was to add a long seatpost fixed to the front edge of the headstock tube on the Roadmaster frame. To do that, a filler had to be added between the bearing cups so the frame is taking the load, not the cups--those cups are meant to take only a vertical load, and don't have enough inserted metal into the frame to actually hold together, and would probably just rip out of the ends of the tube with these loads over time. The best place to get a filler was some more bike frame tube, in a diameter just a shade larger than the existing headstock tube. Part of the front of the old wrecked Schwinn 24" that I used the rear triangle of for the Original Recumbent (OR) has elliptical tubing with the dimensions I needed, so I clamped it down and cut a short length from it, split down the center across the short axis.

This will fit on the headstock tube so that the new tube section has it's long axis protruding beyond the bearing cup circumference, so the seatpost can be welded to the new tubing without doing anything to the cups, and potentially causing problems later with steering.

After cutting out the section above, I verified it's fit and ground off the paint around the weld area:

Then just to be sure before welding, I set all the parts in place flat on the ground, to check that the frames would then sit where I needed them to, with the right spacing down between the various sprockets and belt pulleys, to prevent interference during a ride:


Then I tackwelded the tube half onto the headstock, then clamped the seatpost to the modified headstock's front edge, ensured alignment was as perfect as I could make it, and tackwelded the post on. I rechecked the fit, and it all lined up as intended the first time. I fully welded the parts together, then reassembled as below:

The idea is that this seatpost goes down at least as far into the seattube of the Magna frame as it is welded onto the headstock of the Roadmaster frame, and clamped into place by the seatpost clamp built into the Magna frame. I may add a second clamp between where the seatstays and the toptube connect to the seattube if I feel any issues during testing, but I don't think it will be needed.

Now, the steering was something I had been toying with various ideas for, and the chain-sprocket steering was what I wanted to do just to see how well it would work. But because of spacing and body-length issues, as well as strength of joins, etc, I didn't want to move the Roadmaster headstock back too far from the Magna seattube. That meant I wouldn't have room for a full-circumference sprocket on the bottom of the headstock, which is the only place it would easily line up with the top of the Magna headstock. This would severely limit my possible steering angle, by the amount of sprocket I could leave intact (the chain must remain fully on sprocket at all times or it will become detensioned and come off, leaving me with no steering!)

So I decided a rod-steering would be better, and looked around for various things to use for the rod. I have a number of things I can use, and one of the ones I decided against (but was nearly perfect in every way except being too soft a metal) is pictured above.

An aluminum stem from the old 70's Schwinn 10-speed I would have used for CrazyBike1.0's front frame will probably become the front steerer tie point, as it is at perfect right-angles where the clampscrew goes thru the bar clamp, and all I need do is add a bit of tubing around the screw itself for whatever steering rod I use to ride against, rather than the threads of the screw. Even a rollerskate bearing will work, if needed, as it's height and diameter are right for going between the open ends of the clamp.

Installing the stem at right-angles to the frame/wheel should give me enough torque to steer with. If not, installing it normally and clamping tubing into the frame that then has the steering-rod connection/bearing at it's farthest end will definitely give me whatever torque I need, depending on the length of tube I use (as short as possible to get the desired result).

Mounting at the handlebar end will probably be done via a clamp on the bars themselves, whichever clamp I find works best for the arrangement I end up with in the front, that won't interfere with the usual bar stuff (brakes, shifters, etc).


The last major structural decision is how to brace the rear portion of the Magna frame against sidesway and vertical movement vs. the Roadmaster's center portion. My original quick-made idea was to use a block of wood and clamps, but that will not last very long, as the wood crushes between the frames during use, and if the wood splintered it would probably cause the frame to come apart. Very bad on a ride, especially with the load on this frame from the motor and whatever batteries I end up with.

Below is the frame without any rear support, and just the seat-mockup to show about where it might go (actually farther forward and down, in front of the seattube), with a set of bars inserted on the steering headstock in the middle:

Next shows a black U-fork set into place approximately where it would be welded to the Roadmaster frame (at the center of the upside-down U, to the meeting of seattube and toptube), and bolted to the very rear of the Magna frame, just above the dropouts.

The version I will probably go with is this one, however, because this white fork is actually lighter than the black one by a few ounces, even though it is longer. I think it's probably better steel, too, given that it's just as hard for me to bend it (to spread the ends for meeting up with the Magna frame's triangle) as it is to do for the black frame, which is heavier and thus thicker metal to get the same strength.

This one will be mounted to a much more secure place on the Roadmaster frame--the unused caliper-brake bridge on the seatstays. It curves over the whole chainring area, which looks nice as well as being helpful in keeping me from contacting those (I can easily hang coroplast sheets from it to cover them, for instance). Again it will be bolted to the Magna frame just above the dropouts, and the triangle it forms from those two points to the Roadmaster's brake bridge should be sufficient stiffening to keep the chainlines from shifting (across the gap from Magna to Roadmaster frame).

Since there are two chainlines, that's pretty important. One is for the motor, and the other for pedals.


Another new thing is I am considering changing the device used to get beltpower converted to chainpower, so that it uses a very small chainring (14Tooth and 1/4" chain) instead of the 18Tooth 1/2" chain it currently would use, because it also means I can go from a 50Tooth chainring to an 80Tooth chainring with a slightly smaller diameter. That gives me a much better reduction of 5.7:1 instead of about 2.7:1, getting me closer to 90RPM (max) at the main input to the rear drivetrain to match my average pedal cadence. I'd need 7:1 at 36V or 4.7:1 at 24V to do it; this puts me somewhere between those, which I can deal with a lot better than 2.7:1 (which only really works at around 12V, which the motor does not get enough torque from to be useful, I expect, but haven't fully tested).

To do this, I will have to cut the hub used for the device in half (but not the axle), and use only the freewheel half. On the left side will be some nuts and spacer tubes, then in the middle will be the hexagonally-bodied 14T sprocket that came from the exercycle's "resistance wheel/fan". It conveniently already has a roller-bearing inside it that exactly fits the bike-rear-axle I'm using in this device (remember what I said about coincidences in a previous post?).

Since I don't want to weld to the sprocket or it's body directly (might need it intact for something else later), then I will slip over the 3/4" hexagon a 1/4" drive socket with a 3/4" hex head. The 1/4" drive hole means it fits perfectly over the axle, too, just outside of rubbing on it. That will be welded to the remaining half of the hub's center tube, to transfer the rotation of the freewheeled belt-pulley to the sprocket. (Years ago, I bought a barely-liftable giant bag of assorted sockets from Deseret Industries thrift store for a few dollars, so I have a *lot* of some sizes, certainly enough to use for things like this).

The bearing for the left side against the frame itself will be the cup and ring normally used inside the hub, but modified down to only the bare minimum, so the sprocket will end up in as perfect a chainline as I can achieve to the rear section. It's going to take some fiddling around to figure out exactly how to make it work, but I already have most of the idea in place, and will have the rest before I start cutting into the hub. What bothers me is that I already *have* a hub I cut in half around a year ago for one of the first motorization experiments, that came off the same white bike as the front fork to be used as the rear frame-join on this CrazyBike, but I can't find either half *anywhere* now.

Time to reorganize so all the bike parts are in sorted-out boxes of just one kind of part. :-)