54 Tooth

     So, keeping all this in mind, I decided to see if I could save some money and do it myself.  There are a number of ways to make a bigger ring, but a fairly easy one is to use a big farm equipment sprocket with suitable tooth count as a pattern.  I went to TSC and bought a 54T sprocket with ½" pitch for about $18, and proceeded to try to use it as a pattern.  I made an alignment collar to center the sprocket and my piece of $5 scrap .090" 7075 and after that, it's just a matter of using a 5/16" bit and using the steel sprocket as a drill bushing.  The holes were drilled out fine, but when I cut off the excess, I cut off too much metal and the teeth were far too short.  At least I know what I did wrong and I still have the sprocket for more chainrings.  In the meantime, I had a chance to grab an available 60T Vuelta ring, so I used that instead.  I might still finish it up and see if the teeth still might work for me.

70 Tooth

     For attempt number two, I found two resources for big ring patterns.  The IHPVA has a bunch of big sprocket patterns, plus instructions for use, that can be printed out and pasted to the AL plate, but I found that some of them had center-dots in the tooth pattern that were difficult to see.  Not a big deal, but something that needed to be paid attention to.  The other resource I found was Giles Pucket's (likely the same guy who wrote the IHPVA page) sprocket calculator.  If Moz's site is down, here's a local copy of it.  Anyway, it's an awesome little application and worked perfectly for printing out 4 pieces of paper to tape together and use for a pattern.  I got a scrap piece of .125" 7075 this time, followed the instructions, and glued the paper on.  Center punched the holes, drilled them out with a ¼" bit first, then used the 5/16" on the tooth holes, and a 1½" holesaw and ½" bit for some lightening holes.  Next, I'll cut off the excess, take out the center section, start filing the teeth and I'm not sure if I'm looking forward to thinning out the teeth.  Even a .090" piece needs to have the teeth thinned so I have an extra .032" to take off all the way round.  At the very least though, the sprocket should be stronger than normal.  A little over an hour got me to this point.

16 Dec 2006

     I managed to get a little more time on the 70T chainring.  I got the centre cut out, most of the holes de-burred, and some of the teeth filed.  I have been very inefficient in my proceedure for making this chainring so far.  Lots of going back and forth, wasted time and poor techniques.  It's my first one though, so just so long as I get it finished and its useable, I'm happy.  While filing the teeth is a little tedious, once you figure out a technique that works for you, it's actually fairly fast.  The slow thing will be filing the edge of the chainring thinner all the way around so the chain sits deep enough in the teeth.  It's mounted to a 130mm BCD crank that is 175mm, and will be cut down to 140mm.  Not exactly sure what I'm going to use this baby on, but I think it'll be some flavour of ice racing trike for the moment.

      I haven't decided whether I'll just knock down every 5th tooth to act as shift enhancers for a smaller chainring, or whether I'll take every second tooth off and have the sprocket become a 1" pitch 35T.  If I go with half teeth, I'll have to make sure chain alignment is spot on so it doesn't try to derail itself all the time.  About 1 hour 45 minutes total time so far.

15 Jan 2007

     The big ring was finally finished almost 3 weeks ago, but I have just been busy and didn't update it. Another hour of tooth filing, deburring and painting got the final result to the right.  An old 52T chianring is sitting on top of it to compare sizes.  I shortened every 5th tooth on the big ring to act as shift gates as I'll likely run a 70/5?/36 triple and want to make sure it downshifts ok.  I might need to modify the derailleur cage to work with a 34T drop however.  While that was hanging up drying, I started on some 62T and 40T oval rings and my technique and processes were MUCH improved over my first attempt at the 70T ring.  Total time to make the ring was just under 3 hours.  Now I need to build something to put it on :)

Oval Rings

     I came to the realization/decision that if I am going to go to the effort of making my own custom rings, there's no point making round rings when I believe that ovals are more efficient.  That being the case, I'm going to concentrate on finding out what is the optimum percentage of ovality and what is the optimum amount of offset.  Actually, I guess a more accurate term for them is "elliptical" because that infers a symmetrical shape.  An "oval" is not necessarily symmetrical in both axis, but I like the word because it's easier to type :)  

62/40 Oval

21 Jan 2007

     While finishing up the 70T, I was messing around with the sprocket calculator and realized it also makes oval rings.  So, after doing some 'net research and seeing what the "pros" use for how out of round they make their rings, it seems they do the big rings about 10% and smaller rings about 20%.  The low rings are used when more torque is needed (ie hills) and generally get spun at low rpms, so the higher oval percentage makes it a little easier to get past the "flat" spot in the pedal stroke.  For a little explanation, the whole idea behind the oval cranks (in a nutshell) is that, primarily, they help you get through the "flat spot" in your pedal stroke and secondly, they give you a little more "oomph" when you are in the pedal position where you can put the most power down.  They do this by aligning the larger dimension with the point you are about 90º from full leg extension/leg contraction, and by aligning the smaller dimension with the point your one leg is fully extended and the other is fully contracted.  So, on a bent, half way through the stroke when your feet are perpendicular to the chain line, the tallest part of the oval will be pulling on the chain and that gives you a temporary higher gear.  When your feet are lined up with the chain line, the smallest part of the oval will be pulling on the chain and that gives you a temporary lower gear.  The premise is, that while pedalling proper circles, you are limited in your top gear by how big of a gear you can push in your pedalling "flat spot".  So, by reducing the gear size when you are at your lowest torque, it helps you get through it easier and presents you with a larger gear than you could normally push when you can produce your most torque.  Got it? :)  If not, they just look cool :)

     So, I decided to step it up a little and went to a 62T/15% ring and a 40T/25% ring.  The nice thing about the sprocket calculator is that it automagically aligns the oval rings for you, but you can go in and tweak the alignment yourself if you want.  Also, because the 20% oval on the 40T ring makes it a smaller diameter than usual, I had to use a 110BCD pattern, so I'll have to make another adapter to mount it to the 130mm BCD crank in the picture.  I can see how there will be a lot of permutations and combinations by combining different tooth counts with different oval percentages.  Anyway, after doing the 70T ring, I found that I was wasting a lot of time and not being very efficient at it.  With the one under my belt, these ones went a lot faster because I knew what to do in which order.  So, in the top picture, I printed out the pattern, glued it on, drilled out the tooth holes, cut off the excess and started to pre-drill some mounting holes, all in only an hour.  Much better than before :)  Oh yeah, and that ketchup-looking stuff on the one 40T ring is not kethcup, it's my blood.  After trimming off the excess, the teeth are very sharp and will latch on to inattentive fingers in a heartbeat.

     The next time I got at them, I got the center holes drilled out on 3 of the rings, and the excess material removed from the 62T ring.  That only took another hour.  The 40T rings are small enough that I don't think I'll need to remove any material - although in hindsight, I should have made a 42 or 44T instead.  Next time . . . :)  So, in the mean time, they are sitting as is while I finish up a couple quick trike projects in time for the ice races in Feb in Brantford.  After the frame is done, I'll finish the rings, then, if there's time, I'll send the frame out for powder coating while I finish file and paint the chainrings and adapter plate.  I figure another hour to finish the chainrings and the adapter.  The crank pictured is shortened now to 140mm.

07 Feb 2007

     In the rush to get the chainrings done for the Brantford ice races I only managed to get the 62T painted.  Both rings mounted up great to the old cranks I shortened to 140mm.  The rings worked great and shifted so smooth, I think it shifts even better than a commercial round set.  I was really amazed.  People commented that they felt a little different when they first rode it, but after only a couple minutes, I'd ask them about how they felt, and they really couldn't say anything bad about them.  Someone posted on the net that the derailleur flopped around a lot with oval chainrings.  ??  The RD didn't move at all on either ring so a typical example of someone who knows nothing, shooting down something they've never tried.  I still want to try a pair on my Speed if the FD post is tall enough.  My daughter used the trike at the ice races and she had no problem with the chainrings aside from a crappy FD shift lever that was mis-behaving.  The only two downsides are:

1. Due to the short chainline on the FWD delta, the chain would hit the big ring when you were using the 40T ring and 6th gear.  Mounting the rings to a RWD bent will solve that problem completely, and
2. I think I cut too much meat out of the chainring and it is a little more flimsy than I would like.  It seems like it would be too easy to bend if you hit something or were a grinder.  I will use fewer lightening holes on the next one if I use .090" 7075 on it.  The 70T seems much stiffer but I used a different hole pattern and it was also .125" thick.  I don't think it will pose a serious problem though.

     I'm going to shorten another set of 48/38/28 cranks to 140 and replace the 62/40, double and use the big ovals on a different project.  Their test was a success.

62/54/38 10% Ovals

26 Aug 2007

     I decided to make a new set of ovals for the Speed.  I made some 62/54/38 10% ovals out of.090" 7075 AL.  I also made them with markings for 7 sets of mounting holes so I could adjust the offset in about 10º increments.  I decided to go with 10% ovality because 20% and more is too much of a speed change and was getting hard on the knees.  15% might also have been fine, but I think I'll opt for a little moderation this time and work on the amount the rings are retarded in relation to the cranks rather than going for more ovality.  I marked out 35 holes for mounting the rings to the crank spider, but only drilled out 3 sets right now that equate to about 40, 30 and 20º of offset.  I'll set it at 30º initially which equates to about 25º on the trike and see how that works.  I'll get around to painting them later.

59/48/32 10% @ 5º

1 May 2008

     The initial set of 53/39 rings that I put on the Raptor 44 were an interim measure until I could get bigger rings made.  I did the calculations and we decided that a 59/48/32 with an 11/32 cassette would meet the gear requirments while still providing the option of going to an 11/34 should the need arise.  Based on my experience with the ovals on the Speed, I was happy with 10% ovality as the optimum compromise, and set them at ~5º offset.  I made them out of .090 7075 and total fabrication for them was about 4 hours total.  The biggest job was laying out and drilling all the lightening holes :).  The rider is very happy with them and I think 10% and 5º will be my standard from now on, barring some extreme bent design that means I need to modify the offset slightly.

62/48/34 10% @ 5º HT

4 July 2008

     I had a stock Biopace double on the Raptor but I was running out of top end gears.  I wanted something in the low 60's for top gear, a mid ring that was in between the 52/39 I had been using, and a granny ring that would get me into the low 20 inch range for some steep hills I'll be riding.  I also decided that I couldn't see why I needed all the teeth on the two biggest rings.  On a minor level, they might look a little cooler with 1" pitch, a little more importantly they might shift a little better (not that they weren't shifting fine before), but the most important thing was that I would only have to file HALF as many teeth :)  Woohoo!  I'm not worried about the chain slipping because you can see that there is at least 50% chainwrap, if not more.

     So, I used some .090" 7075 again, took the 62/40 ovals off the 140mm cranks I'd shortened back in Jan 07 and put the 2 new rings on them.  I wanted to have the new rings ready for the Waterford MI races that were on June 21 - 22, and I succeeded in that.  The granny ring will have all its teeth because that ring will take the highest loading of the three due to hill use and any shifting will be down onto it so reduced tooth counts won't help the shifting there.  I still have to finish up the granny but seeing as how the FD is already maxxed out in capacity, I'll be chopping the bottom of the cage and adding an inch or so to each side so the chain will clear it when I'm using the granny.  I have about 200km on the two half-toothed rings now, not to mention the racing for two days, and they worked most excellently :)  I'll paint 'em flat black, mount the third, and then stick a fork in 'em.  Oh yeah, the "HT" = "half-tooth".