Ed Nisley's Blog: Shop notes, electronics, firmware, machinery, 3D printing, laser cuttery, and curiosities. Contents: 100% human thinking, 0% AI slop.
This cable guide / pulley may work better than the one described there, because it puts the cable a bit closer to the original location.
To recap, the problem is that the cable bends around the small finger at about 8 o’clock on the derailleur arm. After a few zillion shifts, the concentration of stress at that point breaks the cable, strand by strand, until it snaps at the most inconvenient moment.
The small brass disk (about 0.43″ dia) has a groove machined around the perimeter that’s roughly the size of the shifter cable. The hole (Number 8 or 9 drill) is a slip fit for the 5 mm bolts, but it’s off-center enough that the cable passes roughly where it would without the disk.
A notch in the side of the disk rests on the finger, guiding the cable over the finger without (I hope) bending it at that point.
The cable just wraps around the screw under the original stainless-steel washer, which pretty much crushes the poor thing flat.
Shift at Large Chainring
Here’s another look with the derailleur pretty much over the large chainring. You can see the disk and groove in action.
This was another quick-and-dirty lathe project, with everything done to eyeballometric accuracy. If it works better than the previous half-assed effort, I might actually get around to making a third one and recording the dimensions.
I collected some loose cells and pulled some cells from the packs to see how they compared individually.
These are discharging at 500 mA, rather than 1 A, mostly because there were fewer tests and I could run ’em overnight. Other than the Tenergy RTU cell, they’re all old and wearing out…
Single Cell Comparison – Aug 2009
The green line is a new Tenergy RTU 2.3 Ah cell; it has a higher voltage, but still isn’t delivering anything close to its rating even at a load only slightly higer than C/5. I have three packs of those that will be cycling through the amateur radios on the bikes, but I don’t like the relatively low capacity. I’ll run these eight cells through the fast charger and do some rundown tests to see if they improve; I have my doubts.
The black line comes from an old batteries.com 2.5 Ah cell. It has the highest capacity of the group, but a rather low voltage. I’ll start cycling those through the blinky lights on the bikes.
The red line is a Tenergy 2.6 Ah cell. I think these are a year or two old, so they’re not faring well at all. OK voltage, but very low capacity. I think the batteries.com cells will work better in the lights, as they have 50% more capacity at a slightly lower voltage.
The blue line is an ancient Lenmar 2.0 Ah cell. As a fraction of its rated capacity, it’s doing OK, but the low voltage is a dealbreaker. Scrap.
Given the poor results from the old & new Tenergy cells, I’m not sure quite what to do. The advertised ratings are obviously optimistic, shipping charges pretty much wipe out any incentive to sample a batch of new cells, and cells get reformulated often enough that old tests you find on the web (this one included!) are useless.
I’ve been using NiMH AA cells to power the amateur radio HTs on our bikes for the last several years, using homebrew 6- and 8-cell packs like this one. In addition, I cycle a handful of loose cells through the LED blinky headlights we use as rear markers.
I don’t lavish much care on the packs, although they generally get recharged before they’re completely flat… if only because the radios automatically enter a low-power mode that takes some fiddling to cancel. They’re charged on a homebrew C/10 charger, typically overnight, and are uniformly warm to the touch when I take them off the charger. Slow charging is reputedly bad for the cells, although everybody seems to agree that fast charging isn’t much better; I have a 4C charger that really puts the screws to 4 cells at once.
Over time the cells wear out and I’ve recently started figuring out which packs & cells to replace. I’m using a West Mountain Radio CBA II for the tests, running on our Token Windows Laptop. The X-axis divisions are its idea of how to do it; Gnuplot does a better job, but you get the general idea and exact numbers aren’t really important here.
Here’s a screen shot with all the discharge tests in one convenient lump. You’ll surely want to click on it for a legible legend…
Pack Comparison – August 2009
Some observations…
I’m using a 1 A (roughly C/2) discharge rate, because the radios draw about that much during transmit, although they run at 30-100 mA during receive. Battery capacity is inversely related to discharge rate and the usual highly over-optimistic advertised cell capacity is usually based on (at most) a C/5 or a much lower rate.
The shortest curves, the orange & black ones under 0.74 Ah, are two ancient 8-cell packs made from batteries.com cells. The cells actually have decent capacity, but the discharge voltage is much lower than it should be.
The black curve to the far right near 2.47 Ah is a freshly charged set of Tenergy 2.6 Ah cells that had been oops discharged completely flat. Other than this run, the Tenergy cells have been a major disappointment: the 6-cell packs near the bottom are running less than half their rated capacity and the 8-cell pack in blue isn’t much better.
The green & red traces out there to the right at 2.23 Ah are Duracell 2.65 Ah cells that are holding up remarkably well. Recent reviews indicate that Duracell (or whoever owns them these days) reformulated the chemistry early in 2009 and the new cells are crap. These cells are colored black-and-green, which seems to be different than the new ones.
The cluster of traces around 1.73 Ah are three 8-cell packs made from two dozen shiny-new Tenergy Ready-To-Use 2.3 Ah cells. I’m unimpressed so far, although they are still in their first dozen cycles. There’s obviously one weak cell in pack A that causes the abrupt fall-off in the two shortest times, but they’re all pretty much the same.
Given that we have three bikes and I want a backup pack for each bike, that works out to
8 cells/pack x 2 packs/bike x 3 bikes = 48 cells
I’d like to think that spending four bucks per cell bought you better cells, but the Duracell reformulation puts the kibosh on that notion. In any event, you can see this gets spendy pretty quickly…
I’ll run the best of the old cells in the blinky headlights, which run at a 50% duty cycle of 400 mA or so.
We spent four days biking along the Pine Creek Valley rail trail with a Rails-to-Trails Conservancy group ride on our Tour Easy recumbent bikes. Because a crushed-stone path creates a lot of noise that the fairings direct right into our ears and because we weren’t going very fast, we left the fairings at home. As a result, the bikes were wonderfully quiet.
Some years ago, Mary sewed up “bubble wraps” to store our fairings on those rare occasions when they’re not on the bikes. She had some red flannel left over from another project and a hank of cheery Christmas-themed edging, so they turned out to be rather conspicuous.
The trick is to get the size right when the fairing is rolled up. With the fairing in its natural bubble shape, the wrap is rather limp, so you need pockets on both ends to hold the wrap in place. The toes are, she admits, an affectation, but didn’t take much figuring to get right. The width is just slightly more than the fairing’s flat width; you find that by rolling it up and measuring the roll.
She actually made a paper template first to sort out all the curves, then transferred that to the flannel for final cutting.
Tuck in the fairing’s head & toes, roll it up toes first, tie the (attached) strap in a neat bow, and it’s done!
We have three fairings and they roll up together, each in its own wrap, into one tidy, albeit rather heavy, package.
We’ve been using Cateye Astrale “computers” on our bikes for decades, mostly to get the cadence function. After all this time, we pretty much know how fast to pedal, but old habits die hard.
The cadence sensor counts pedal revolutions per minute, which requires a magnet on the crank arm. They provide a small plastic-encased magnet with a sticky-tape strip that’s worked fine on our previous crank arms.
Our daughter’s Tour Easy arrived with fancy curved pedal crank arms that put the cadence sensor magnet much too far from the frame. You really want the magnet & sensor close to the bottom bracket so that it doesn’t get kicked and doesn’t snag anything as you pedal, but that just wasn’t going to work out here.
A turd of JB Weld epoxy putty solved the problem: mix up a generous blob, shape it into a pedestal, glom the magnet atop it, adjust so the magnet is parallel to and properly spaced from the sensor, then smooth the contours a bit.
Add the cable tie for extra security; you don’t want to lose the magnet by the side of the road!
The black electrical tape is mildly ugly, but serves the purpose of keeping the cable from flapping in the breeze. The adhesive lasts about a year, then it’s time for routine maintenance anyway.
Although recumbent bikes use ordinary bicycle components, they tend to have somewhat different frame geometries (to put it mildly). Our Tour Easy ‘bents seem to put a particular strain on the front derailleur cables, perhaps because the cable enters from a different angle than the derailleur designers expected. The little finger that’s supposed to guide the cable actually concentrates all the bending force at one spot… precisely where the cable breaks.
If you look carefully, you’ll see a little brass disk (between the derailleur body and cable) that cradles the cable. I made that for the previous derailleur, but this one has Yet Another Geometry. I know there’s a difference between “high pull” and “low pull” front derailleurs, and perhaps this is the wrong one for this application, but there seems no algorithmic way to sort this stuff out.
Cable guide pulley
The solution is to make Yet Another Cable Guide Pulley, with a groove around the perimeter, an off-center hole, and a notch to clear the finger. It’s not exactly a pulley, but I’m not sure what else to call it. Maybe just a cable guide?
This was a quick-and-dirty manual lathe project, two days before leaving on a trip: turn down some brass stock, put a groove around the perimeter, part it off, drill a hole, and cut the notch. Not a trace of CNC to be found: all done by guess and by gosh, marked out with Sharpies on the actual part in real time running.
The general notion is that the cable rides the groove smoothly throughout the derailleur’s entire travel range and, thus, doesn’t bend around the finger. This changes the shift geometry just slightly, but, fortunately, long-wheelbase ‘bents have a sufficiently relaxed chainline that indexed front shifting isn’t much of a problem even with slightly misplaced positions. Besides, that’s why SRAM grip-shifter have all those clicky stops, right?
(The shifting is actually a bit goobered, with the outer chainring shift a bit too close to the middle. When we get back, I’ll re-do this with somewhat more attention to detail.)
Pulley in action
Here’s what it looks like in action. I’ve had good success with this sort of thing over the years, so I think this one will work just fine, too. It simply takes one broken cable on each new derailleur to spin up enough enthusiasm for making Yet Another Pulley…
The front ball joint on the mirror on Mary’s helmet loosened enough that the mirror blew out of position every time we got up to a decent traveling speed. I’ve repaired these mirrors several times before; they’re plastic and tend to fracture / wear out / break at inconvenient moments.
The first pic shows the mirror (the black surface is reflecting the dark floor joists overhead) with an old blob of epoxy that repaired a break in the outer socket. The socket originally had stylin’ curves joining it to the mirror, which proved to be weak spots that required epoxy fortification.
This time the socket split axially on the side away from the mirror, which released the pressure on the ball socket that seats into it. I found a chunk of brass tube that fit snugly over the socket, then carved some clearance for the existing epoxy blob. The key feature is that the tube remains a ring, rather than a C-shaped sheet. to maintain pressure around the socket.
Clamping the reinforcement ring
Here are the various bits, with the reinforcing ring clamped in place. I coated the socket exterior with JB Weld epoxy, slipped the ring in place, and tapped it down with a brass hammer to seat flush with the front face of the socket. That left gaps between the socket opening and the tube that I eased more epoxy into with an awl. A bit more epoxy around the exterior smoothed over that ragged edge.
The strut at the bottom of the picture ends in a ball joint held by a socket that slips into the mirror socket. The loose brass ring above the mirror is some shim stock that I added some years ago to take up slop between the ball socket and the mirror socket and tighten the ball joint. I suppose that pressure eventually split the outer socket, but so it goes.
Repaired mirror joint
The clamp squished the outer socket enough to snug it around the ball socket, so when I reassembled the mirror it was fine. To be sure, I dunked the ball in my lifetime supply of Brownell’s Powdered Rosin for a bit more non-slip stickiness.
I have a box full of defunct bike helmet mirrors, dating back to those old wire-frame square mirrors that clamped onto the original Bell helmets. The newer plastic ones just don’t last; we ride our bikes a lot and even fancy engineering plastic isn’t nearly durable enough. A few bits of metal here and there would dramatically improve the results!
I’m going to build some durable wire-frame mirrors, but … this will keep us on the road for a while. I suppose I should make a preemptive repair on my helmet mirror while I’m thinking of it…
The Smell of Molten Projects in the Morning 