View Full Version : Making my own sprocket

09-04-2005, 07:48 PM
Well, I've had some time lately to spend on my tadpole project. Electric tricycle, two wheels in front, pedal assist one man vehicle.

I've had it out on the street and got a couple of kilometers on it now. Still no brakes, but what the heck, I'm using the 'short out the motor' braking system, and it works fine at this initial stage of testing.

I haven't gotten it out of low gear yet, and it's giving me an estimated 20 to 25 kph. This is a guess based on my bicycle speedo and the rate at which the neighbor's houses go by kind of comparison. At any rate, I figure the rear sprocket should be about 1.5 times the diameter it is now. That will give me a lower low 'gear' ratio, and will allow me to get to about 40 kph in high gear, which is looking about right to me in terms of the power I have in the motor and the maximum speed I feel to be safe to go at in this contraption.

So here it is. I'm making a new sprocket from aluminum sheet, and I have a question regarding the radius or diameter of the gullets in the sprocket. I'm using bicycle chain which is 1/2 inch pitch, and the rollers on it are .304 average diameter, which is just under 5/16 inch. What might be the proper diameter to make these gullets? I plan to drill out a ring of holes in the al disc, then machine away the outer portion, leaving a bit more than half the holes' diameter on the disc. I will then shape the tips of the teeth as required for a good clean wrap of the chain around this sprocket. I guess I'll have to taper the teeth from each side a bit as well so the chain can engage well even if off-angle a small amount. The sprockets will be lined up perfectly, but I just want to allow for some mis-alignment by doing this.

Any suggestions? I could use a 5/16 drill bit for the ring of holes, and for this I would use a short drill bit so it stays on target. I could always grind a custom bit for the smaller diameter if that's what I should be drilling here. I will be using a jig on the mill to ensure equal spacing for the holes, and to get the root diameter just right. It's mainly the fit of the roller in each gullet I'm wondering about. Thanks for any advice or input on this. Or criticisms, if that's what I should be getting.

I probably could buy a suitable sprocket, but I'm not sure about getting it in aluminum, and I want to keep this as light as I can. I don't really want a steel disc 10 inches in diameter on my back wheel. I would rather make this myself anyway, and I'll get to use my homemade indexing jig. I'll also get to fit the sprocket very concentrically to the hub, since I will bore it to fit before I remove the blank from the jig. I also want to drill a series of decorative holes in this sprocket, and co-ordinate the bolt holes to my hub, so that's another reason for making it myself.

I'm getting pretty anxious to get this thing on the road with a proper speed control and brakes too. So far I've been using the 'touch the wire to the battery post and don't burn my fingers' method of getting a test drive or two in. Pics are coming-

3 Phase Lightbulb
09-04-2005, 08:00 PM
I've made my own sprockets. I wrote some software to calculate the hole positions relative to the center of the sprocket so that makes it easy to drill out the sprocket with only a Bridgeport and a DRO.

I drilled all of the holes, then cut an angle in each tooth by side milling each one:


If you tell me how many teeth you want your 1/2" pitch sprocket to have, I can give you a list of X,Y positions you can use to drill your sprocket on a milling machine w/DRO.


John Stevenson
09-04-2005, 08:10 PM
Go to http://www.wadeproco.com/ and download the demo of Machinists mate it's a 30 day demo.
Click on the sprocket icon and then lookup, slect your chain, size 41, enter number of teeth and type in an output file, something like c:\temp\sprocket.dxf

You now have a choice. You can load up the tooth in a cad program do an array to get all the teeth, print out and glue this to your work.

Or you can take the PCD from the program and set a pair of dividers to 1/2" and step off the centres around the PCD.

John S.

09-04-2005, 08:39 PM
"I'm making a new sprocket from aluminum sheet,"
You mentioned "Aluminum sheet" but not the grade, I'd recommend using 2024-T4 plate (Or Round stock) for the sprocket. Hate to see the machining go to waste on a softer material that would fail prematurely.
Have fun,

09-04-2005, 10:43 PM
Thanks for the replies. I don't have a dro on any axis, so that kind of negates the table of co-ordinates to go by. I'd go nuts trying to hand crank that, and I can't go that far in the y axis anyway. As far as the program to develop a paper overlay, or using dividers, my jig will take care of the hole spacing automatically, and the indexing arrangement I have will allow me to adjust accurately for the pitch diameter. I won't need a pattern to follow. All I need to know is what size to drill the holes.

I've been looking at the sprockets I have on hand, and the largest one ( the one in use right now) looks to have a slightly larger arc at the root than the chain's rollers' diameters, but not by much. It's even a hair larger than a 5/16 drill bit. Bear in mind that this is a used sprocket, and there's bound to be some wear on it.

Because I'm doing this without benefit of cnc or readouts, I will make up a short section of aluminum to test my indexing. It will end up with the same curvature and radius as the full disk will, but for a test I'll only need maybe ten teeth. I'll wrap a section of chain in the teeth and adjust the pitch diameter as needed to get a good result. I have an easy way in mind to profile the teeth, so that isn't a concern. It will be tedious because there's 60 teeth, and I'll have to flip the disc over and profile opposite sides of tooth, so that's 120 operations, but again none requiring table movement, and actually np spindle movement either, except for rotation. I won't be cranking the table in either x or y to facilitate the drilling, so I'll be tightening the gibs to eliminate play, and the accuracy of my initial holes will be depending on the stiffness of the drill bit at the tip, and the evenness of the indexing setup. I'll probably do a round of pilot holes using a very short bit, then go round again with the 5/16 bit. ( I think I've settled on that size as being pretty much optimum for this chain ). I'm going to index three notches at once to even out any variation in my indexing 'wheel', so I should be getting a decent sprocket out of it.

As far as the grade of material, I don't know what it is, though I do know it's not the softest stuff around. It cuts cleanly and crisply, without chips welding to the cutter like some of the other sheet I had. Thanks for the warning on that anyway, Ed, and thanks John for suggesting that program. I think I'll download it and check it out anyway. I'll probably learn something from it. Adrian, I hope my sprocket turns out as nice looking as yours did. Thanks for posting the pic.

09-05-2005, 12:48 AM
Ok, just a short update. I cut a test piece with about 8 teeth on it and just filed the sides of the teeth for clearance. I laid a section of chain on it and it fits nicely. If I hold down the first and last link with my fingers, I get a barely perceptible amount of play in the chain between these points. Each roller fits it's tooth without snagging and without leaving the previous link loose at all. The geometry seems to be right, so that's what I'm going with.

Now for the snag- I wondered a bit more about the alloy I was planning to use. Just for kicks I put a 90 degree bend in a scrap piece of this 1/8 thick aluminum. It looks like I got an inside radius of bend of about 1/8 inch with only minor markings on the outside to indicate stress from that tight bend. I don't think a stronger alloy will do this. I know that 6061 T6 won't do this bend without at least some surface tearing. I guess I have too soft a piece to be making my sprocket out of. Would 2024 T4 make this bend without tearing- I don't know but somehow I doubt it. And I doubt that my bender would bend more than soft aluminum of this thickness and length of bend. I use a 6 ton hydraulic jack in the bender, and the test bend was about 6 inches long. And I got a very tight bend on this stuff. Crap, I guess I can't use this- need to get to metal shop .

[This message has been edited by darryl (edited 09-05-2005).]

[This message has been edited by darryl (edited 09-05-2005).]

09-05-2005, 01:10 AM
No, 2024-T4 will fracture, a poor choice for bending. 6061-T6 will bend well across the grain, with a large bend radius, but will likely fracture along the grain. For bending, 5052-H32 is best, but a poor choice for a sprocket. Any way around the need for bending?

09-05-2005, 02:30 AM
Ed, I have no need to bend, I just did that as a way of testing the sheet material I was going to use. I would like to speed along and make this thing, but I'm going to wait until I get to a metal shop and ask for the alloy you mentioned. I don't know if that will be available, but if they can supply something more suitable than what I have, then I'll be happy. Thanks for the feedback.

I've added a collage of pics to my previous post to show what's been going on. I reduced the size of the pics so they wouldn't run off the page and piss everyone off. Then I assembled them vertically so it came out as one pic.
From the top- at the left you see a piece of mdf board that has a center hole in it and turns on a pivot point fixed to the table. That part you don't see in any of these pics. The mdf has been milled to an exact size on it's outer diameter so a loop of chain can be press fitted onto it. The chain rests on a lip at the bottom of the board. To the right and farther back is a shiny steel pin, just to the left of the threaded rod. That pin is the pivot point for the ratchet. Looks like a bit of hardware there, but I'm just using pieces that have been made up for some other use. In this case, it was a quick and easy way to get a pivot point for the ratchet. Sitting on the mdf board is my test piece with a length of chain wrapped on it.

Next pic is a close-up of that piece. Then you see the ratchet disengaged, then engaged, then a bit closer pic of the setup. In use, I just turn the mdf disc and engage the ratchet, drill the hole in the workpiece (which would normally be attached to the mdf disc), then turn the disc for the next hole, etc.

Last pic shows the only pic that turned out semi-ok before the battery went dead. This is the tadpole. You can just see the left front wheel, and back from there. You have to imagine the motor located under the seat because you can't see it, only the wires that come from it. Better pics later, I promise. There's the rear swingarm, the struts with four valve springs for suspension, and hard to see but between the springs is a white plastic assembly which is actually the adjustable shock absorber.

A considerable amount of sheet metal work remains to be done, but I figured I better get a pic or two out so anyone who's interested can see what I've been up to. The pedals are not on yet, but there's the joystick right in the middle, sticking up where it's ready to rip my nards off if I hit something. There's a change going to happen here once I get the brakes on it and hooked to a sort of handlebar with levers.

I'm planning a formed lexan fairing for the front and that will extend part way back, and that's just about be the extent of the bodywork. Some sheet metal fills in yet, a rear fender goes on, and I want to put a zoomy design in the sides by cutting away parts of the sheet aluminum. Maybe a lightning bolt, maybe something else. The highest part in the picture is the back rest- that part gets tapered towards the top and makes an area either side for brake, signal, and clearance lights. I'm going led if I can, but I really want to make this thing visible from the rear especially, and also from the front. Led headlights, reflectors, whatever works that I dream up while eating ice cream and pinapple slices. (going for that right now actually)

As it is now, I can lift it myself onto that worktable, but not with the battery in it. The battery, a typical sized 12 v but deep cycle, weighs about two-thirds of what this contraption weighs as you see it. More later if I don't get barred for being ON topic. http://bbs.homeshopmachinist.net//smile.gif

[This message has been edited by darryl (edited 09-05-2005).]

09-05-2005, 02:40 AM
Just re-read your post. I'm guessing you're judging the material suitability by seeing if it will bend without fracturing. Believe it or not, if the material fractures while bending, this is a better choice of material for a flat sprocket. It indicates to me you're using at least 6061-T6 or better. You need a tough grade of aluminum for a sprocket. If it bends without fracturing, the material is too soft for a sprocket, i.e. the gummy material that "welds to the cutter", probably 5052-H32.

(My edit)

Just saw your update, NICE WORK! Sounds like your material choice will work. Keep us up-dated.
Have fun,

[This message has been edited by egpace (edited 09-05-2005).]

09-05-2005, 03:06 AM
without reading a majority of the replies, i'll throw in my two cents:

i've made a pile of high-strength sprockets for bmx bikes out of 6061-T6 initially, then 7075-T6. 6061 is definitely more than enough for your application, and it's also pretty much the most common alloy out there in terms of aluminum.

i used a cnc machine to profile my teeth, but i think drilling holes at the right spacing will work, but you'll probably have to do some sort of angle on the teeth if you want it to run correctly. i got my pitch diameter measurements out of an old version of machery's handbook under "transmission roller chain." 5/16" works exactly for the roller diameter on bicycle chain and i can get you numbers for pitch diameter (circle which goes through center of all the rollers) if you'd like, provided you say how many teeth you need.

i've been running some form of my sprockets for about 2.5 years now without a single problem and there's about ten more out there somewhere in circulation getting horrible abuse by a bunch of ratty bmx kids who like to slide down handrails on their chain and stuff, so i know they stand up to abuse.


there's a pic of my latest one, which has been running strong for over a year now. running a tiny gear (27 tooth to be precise) is actually harder on the sprocket, due to the fact that there's less teeth to distribute the force of pedalling amongst. this also equates to chains wearing out faster, but the weight savings and smaller likelyhood of smashing it on something while riding and folding it over is a greater tradeoff.

[This message has been edited by psomero (edited 09-05-2005).]

09-05-2005, 03:13 AM
Apropos of your future with this magnificant machine may I suggest you think about steel for your sprocket.
My basis for this suggestion is experience a long time ago with a Morgan 3 wheeler. with a twin cylinder motorcycle engine on the front.Driving thru shaft to the gearbox and then chain to the rear wheel. 100 miles per hour was often achieved. Going round corners was a little hairy as the passenger had to counter the lean a la motorcycle outfits. Racing was always fun.Provide seats for the obligatory passenger. You wont cut much ice with the female of the species. difficult to get girl and her 'new look'frock and high heels in the space available.
I think the motor we had was a Vincent but these days a Harley V twin would seem about right. The blown Chevvy V8s are a bit heavy.
Disc brakes would be the go.
best of luck,

09-05-2005, 03:15 AM
3 phase,
Nice work.

09-05-2005, 08:49 PM
Ed, that was my intention with the bend test, to see if the material was soft or not. If the bend would have been tougher, or cracked, then I would have assumed a stronger alloy of some type, more suitable for a sprocket than the soft stuff. As it turns out, I have a piece of 6061 plate in the right thickness, so I'm going to use that. I did ask someone working a cnc shop next to us about the 2024 and he said they don't stock it, but they do have a 5000 series and 6061. The 5000 series is too soft for a sprocket. I'll be keeping my jig, so if I need another one made from better material, I still have the option. It will be interesting to see if the wear on the teeth shows up any grain orientation in the 6061.

Psomero, I can't tell that your sprocket is homemade, looks like it would have come with the bike. It looks to me like 5/16 is the right size for the drill, so thanks for the input on that point. I will be profiling the teeth and also with a jig so I'll be getting each one shaped the same on both sides, and to proper profile. If I run into a problem with the method I intend to use, I'll post a note with one of those 'dismay' faces behind it. http://bbs.homeshopmachinist.net//smile.gif This will be a 60 tooth sprocket and I do have the pitch diameter figured out and adjusted to on the mill. Also I did a test piece as per the pics. All that remains is to fasten the blank to the jig and have at it. I'll have a working sprocket in my hands tonite. Haven't decided on a pattern of holes to lighten it yet, but I do have a very limited area where the mounting holes have to go, so I'll do the required first, then worry about other holes later. At least I get to bore it to fit the hub, so it will be automatically centered, with no chance of it going off kilter.

A big difference between bicycle and motor driven like mine is that the torque stress on the chain is greater on the bicycle because of the relatively small size of the driven sprocket. I will be having a 60 tooth rear sprocket being driven by a 16 tooth. Granted, ny vehicle weighs more than a bicycle, but if you factor in the actual weight including rider, it's less than double the weight that a bicycle with rider would be. Add to that the revered ratio between drive sprockets and driven sprockets, and the tension load on my chain will be considerably less than if it were a typical bicycle. Add the other factor of more teeth on my driven sprocket, and the tooth load will be much less as well. My drive sprocket is steel and hardened, so I won't be worrying about that. I have a second one anyway.

I did consider steel for the driven sprocket, but for the above reason, and the fact that my power is very limited I don't think aluminum is going to wimp out. My math shows that I'll have a maximum surge power into the motor of less than 3 hp, and in steady state use at my full speed I'll be using less than 1 hp, more like 1/2 hp. That's input power, not output which is likely to be about 35 to 40 % less than that. For what my average power is going to be, I'm also confident that the plastic jackshaft pulley will stand up. I'll show some more detailed pics later.

john hobdeclipe
09-05-2005, 10:17 PM
That's nice looking work! Remind me someday to tell you about trying to pedal a guy's attempt at a human powered airplane.

Consider using stock front chainrings (that's bicyclese for front sprocket) from a bicycle shop or internet bike parts house. You could make an adapter to fit commonly available 130mm (Shimano) or 135mm (Campagnolo) bolt hole circle rings, and you have nearly instant access to chainrings from 39 teeth up to 61 teeth. Look here:


Rings in the 39 to 54 tooth range should be available at most any bike shop. Anything over 54 teeth is a bit unusual, but obviously available.

It's up to you to determine if these would hold up to your horsepower requirements. bear in mind that a person on a bicycle rarely develops over about 1/2 HP, and racing cyclists may push 1 HP and occasionally more, but not for hours on end.

09-06-2005, 08:51 AM
<font face="Verdana, Arial" size="2">Originally posted by darryl:
Psomero, I can't tell that your sprocket is homemade, looks like it would have come with the bike.</font>

that fact helped them move very easily at $20 bucks apiece when similar pieces were going for atleast twice that retail...

09-06-2005, 10:00 AM
Forgive me if I'm wrong but in your original post you stated a biger rear sprocket will give you a higher top end speed?

Shouldn't that be smaller rear sprocket?

09-06-2005, 11:15 PM
The larger sprocket alone would give me 2/3 the speed of the original sprocket, but I still have the ratio change available on the motor pulley. That gives about 2.5 to 1, so in 'high gear' I would be getting 2.5 x 2/3, or about 1.7 times my initial test speed. I don't expect to get that much, considering wind resistance, etc, but I will still ultimately be getting a faster speed, even with the larger rear sprocket.

I did make the sprocket last night, and it took me longer to figure out a zoomy looking pattern of holes for it than it did to drill out for the teeth.

09-06-2005, 11:45 PM
Just a note on 2024-T4 vs 6061-T651. The 2024 is a lot stronger (about 50% greater ultimate tensile) and also work hardens more than the 6061. It is commonly used for aircraft parts and you might find some at your local aircraft repair shop. It's also nicer to machine.

3 Phase Lightbulb
09-07-2005, 12:13 AM
High performance sprockets are made out of 7075-T6. Nothing else seems to last as long in the worst possible conditions (constant bombardment of sand into your sprockets). Anything harder and you chew your chain up, anything softer and they just grind away too fast. I run 7075-T6 sprockets on all my bikes.




09-07-2005, 02:25 AM
Good idea, Evan. Next time I'm near our airport, I'll poke around the hangars a bit and see what materials might be laying about.
Adrian, what do you do with those after they get like that, sell them as saw blades? http://bbs.homeshopmachinist.net//smile.gif
Yeah, I'd like to use a stronger alloy for projects like this, but in this town they look at you funny when you ask for 7075 or 2024 or whatever else that they don't stock. Along with that funny look comes this line- 'ooh, that's going to be pricey'.

Here's another set of pics. The top one is the sprocket nearly completed. Some work remains on the holes, maybe some milling between holes, I don't know yet.
Next you get to see a tooth or two actually in focus, and larger than life. Try to ignore the machining marks, I'm trying to. I know the chain will polish that down and I'll forget how bad it looks magnified.
After the trauma of that last pic has eased, check out the pivot point I used to spin the mdf indexing disc on. You can see the post sticking up and the two setscrews that are used to force this piece up against the bottom of the T-slot, fixing it in place. Try to ignore the little divots taken out of the table. Oops.
Under that you see the sprocket being pushed towards the pivot point, and this is how I profiled the teeth. I just swung the sprocket a bit to the right, keeping some pressure against the pivot, and the cutter took out the excess meat on one side of the tooth. I have a stop bolted to the table so I can't turn the blank too far, which would have allowed the cutter to eat the gullet up.

[This message has been edited by darryl (edited 09-07-2005).]

09-07-2005, 03:04 AM
looks good!