Giraffe wheel modification

I took a bmx spider (my brand choice was Profile, but Redline works fine too), and milled down the arms to just show the 74mm bolt circle holes. Then I had the spider drilled out and tapped for a 1.37" standard track hub, and mounted the spider with the lockring. I added a chainring of choice (28tooth), a bicycle crankset cut & drilled at 135mm with 74mm BCD, and a good bicycle chain. The chainrings are replaceable, and the spider is part of the hub. Rides MUCH better than stock, and absolutely no chain slop.


Cool. Did you have any issues with keeping the chainline in alignment? It looks like that modification may have moved the cog outwards a bit.

Looks interesting. I’m concerned that you’d have the same problems I’ve had using even very high quality track hub and parts. On the running mounts, the sprocket tends to unscrew with all the back-pedaling force.

However, the replaceability of the bottom sprocket is pretty neat. Does it come easily off the wheel without unscrewing the spider?

If I ever build a girraffe, I’m going to try this for the hub;


No more worrying about the cog backing off.


Good idea. Any problems with roundness?

Since the “sprocket” can be changed with bolts there is no reason to retain the track threading. Why not weld the spider to the hub? Oh, right, lacing up the spokes…

The ISO brake mod looks cool, too. Since the sprocket rides the bolts instead of the hub threads there is more risk of out-of-round slop appearing. The nice thing is that the ISO fittings are stressed to take the load. Disk brakes apply forces many times greater than is possible with pedals.

I’ve been an advocate on these boards of track threading, but for critical use where there are lots of shock loads these bolted setups are probably superior. On the mtb links those bikes are obviously used in downhill situations where the wheel tends to chatter over the rocks. You rarely see that kind of load in normal street or track fixed gear riding. And I’m beginning to appreciate the way unicycle gear is abused with shock loads too. Not from pedaling, but from the pedal hitting the ground in UPDs. I imagine a giraffe’s pedals get whacked even harder than a regular unicycle’s pedals, so the drive train gets some pretty severe hammering every now and then. This too is rare in road and track fixed gear riding.

I wonder if the sprocket would be available in a size that would be large enough for a giraffe something like a twentysix, being that it is made for a bike it may not be available any larger than a seventeen.

I built a giraffe wheel using a Suzue ProMax Hub and a Euro-Asia Imports sprocket, both of which are near the top in terms of track components quality. I hoped to prove my theory that using higher-quality components would eliminate the slip and make using them viable. However, our tests at Unatics showed that the hard back-pedal pressure on the running mount was too much. The wheel would develop some slippage between the sprocket and hub. Although I did have a different wheel with a similar components (surprisingly, of lower quality) pass our tests last year, I believe that the fine-threaded track hub design just doesn’t work well when the pedalling force is matched forwards and backwards, like on a tall unicycle. This conclusion is not based on hard falls or shocks or vibration, but simply on that mount.

Here’s a pic:

Yeah, it’s clear now that track stuff is probably not right for giraffe use. Giraffes use a really low 1:1 ratio that you would never see on a road or track fixed, and the sudden stomp onto the pedal during the mount is unlike anything you could do with your foot clipped to the pedal. With your foot clipped in it’s only muscular contractions and smooth inertial effects. The stomping motion adds an impuse of stopping the leg mass.

It wouldn’t be easy to quantify the differences, but clearly there are qualitiative differences. Regardless, if the empirical evidence is that even good quality track gear won’t hold up then some other solution is warranted.

On a completely unrelated side note that isn’t worth its own thread, I rode a 29" unicycle down to the canoe club yesterday to go paddling. It’s about 3 miles to the river on an occasionally steep, paved bike trail. I walked part of the trip, partly because the trail was narrow and crowded, and partly because I couldn’t find a lamp pole or sign to use to get on (I’m still unable to reliably freemount). I tried to ride back but couldn’t. My legs were useless so I took a cab. Anyhow, I’m very proud of going 4 miles in my first “epic” voyage on a unicycle. It was also the 29’ers maiden voyage. It’s not as fast or as “cool” as the Coker but it’s easier to control on sidewalks and faster than the 24", so on the urban trip to the river it’s just perfect.

Obligatory giraffe comment: This ride would have been dangerous and difficult to do on a giraffe. As it was I hit my head on a few low-hanging branches. Also, the hight of a giraffe would put the rider above both safety rails on the Key Bridge. Not a good place to have an UPD.

  1. Chain alignment was solved by adjusting the cones and spacers of the hub to place the chainring in alignment with the crank sprocket. Also, the bottom bracket itself can be adjusted with spacers to fine-tune the alignment.

  2. Spider was drilled on a lathe, so it is perfectly round, and of good quality, so the bolt holes are accurate. These rings are perfectly round.

  3. On my 2nd wheel, I used a sealed bearing bmx hub (no lockring) and locktited & tack-welded the spider onto the threads. Yes, it would be difficult to replace all the spokes, but if one breaks, the odds are good that it will pull out somewhere between the spider arms.

  4. Upper ring is 30, lower is 28. If you need to change the gear ratio or repace a worn or broken chainring, standard chainring bolts are used in the spider, which is 5.5mm thick.

  5. Additional weight was saved with a 3/32 chain. I myself have had no back-pedal loosening of the spider, and I’d be happy to lend it to Harper for his “input”.

  6. I did consider drilling a standard chainring to fit a standard ISO disc hub, but I thought that had more room for roundness error. If someone can do this, that is clearly better, but you’d need spare drilled chainrings ready to go.

Quality components make for a better quality ride. Quality round chainrings and cogs are a good thing. You now have the worlds best Torker giraffe.

The wheel with the BMX hub and the spider tack welded on should hold up. I would feel a lot better on a giraffe with a welded on cog than a track style cog with a lockring. I volunteer Harper to test out both wheels.

Yes, the disk brake mount would be more tricky to machine. Getting all of the holes perfectly concentric with the cog would be difficult. And then the bolts need to fit snugly so there is no slop. Tricky.

The disk brake hub modification does have the advantage that the cog can be removed for replacing spokes or changing the gearing.

For reference, here’s the how to guide at for the fixed gear disk hub conversion. Mojoe’s links made reference to that how to guide, but didn’t actually provide a link to get there.

Could you use the disc brake hub modification on one side and this modification on the other side to make a DM style twin chain giraffe?

You’d need one left side drive crank and one right side drive crank at the top (presumably off a tandem).


It would be easier to use the spider mod with a flip-flop hub. I believe there are fixed-fixed versions of the flip-flow around, or at least there used to be 20 years ago. Tack welding could be used with fixed-free or a free-free flip-flop hubs.

How about drilling and tapping a hole or two through the sprocket and hub and use a bolt (or bolts) to prevent the cog from spinning off? It makes a lot of sense with the spider mod as there really isn’t a need to change the spider. Track racers occasionally change sprockets between wheels so anything other than a lock ring is impractical, but a giraffe rider could get away with it if there is enough metal to make it work.

My impression while looking at the setup that I built (see above) is that the sprocket and flange are too far apart to make an effective bond. In addition, the flange is aluminum and the sprocket is cromo. I considered this but ruefully rejected it.

Agreed. I was thinking of drilling down low, directly into the hub body where the lockring is. Don’t know if there is much metal there, though. The bearings are in the way. Also, I wouldn’t want to do this with a nice hub and sprocket.

The only function of the lockring is to keep the cog from unscrewing, but the shallow threading makes that quite difficult without huge axial forces. The screws wouldn’t need to be very big because they would be very efficient at keeping the cog from unscrewing.

Dave, have you tried red locktite? Seems extreme but then again so is a mechanically-inspired fall from a giraffe.

On the wheel with less-expensive components (Suzue basic, Sem sprocket, Surly lockring), I used red loctite on both sprocket and lockring. I got up on the giraffe and stopped the wheel in place against the porch, and bounced on the front pedal to tighten the sprocket. Then I put on the lockring and stood on the wrench (about 200 ft-pounds) to tighten it (risky but I was careful). This approach seems to have worked because we tested it and the rider has used it for months now.

On the second wheel with more expensive components (Suzue ProMax hub, EAI sprocket, and Surly lockring) I used grease on the sprocket and red loctite on the lockring. I used basically the same approach to tighten things. We have slippage.

The easy conclusion to come to is “Duh - use red loctite on both and you’re all set; not only that but you don’t need to use more expensive components”. However, that really doesn’t address the basic design considerations:

  1. Have a strong hub where the axle will not bend under vigorous use. The 3/8" axle is not sufficient.

  2. Be able to change sprockets at will with a minimum amount of fuss. The red loctite and risky tightening procedure are not acceptable for a good solution.

  3. Be able to service the wheel (i.e., change spokes, even rebuild the wheel) like a normal hub. This eliminates the use of red loctite, welding, and similar approaches.

  4. Have a high-reliability, secure connection. A solution that depends on loctite isn’t a solution, it’s a patch. More like using advanced chewing gum.

  5. Make use of high-quality components to ensure a smooth, non-binding pedal action. We’re ok there with the track solution; everyone that has tried either wheel has commented on how extremely smooth it is.

So a stock Torker or similar basically handles only 4. The track hub solution handles 5, and marginally 4, but nothing else. unisk8r’s solution handles 2, 4, and 5. The standard Sem solution handles 4 and 5 (based on one I just delivered). I am working on a solution that handles 1, 2, 3, 4, and 5, but I’m not free to discuss it yet.

The other nice thing about unisk8r’s solution is that it allows one easily to match the top and bottom gears, since they are drawn from the same parts pool. It is also easy to “gear down” the unicycle. The disadvantage of that is that they are drawn from the same parts pool, so it would be harder to “gear up” the unicycle. The wheel seen in my photo uses the smallest possible chainwheel on top with the largest possible track sprocket on the bottom, which achieves very close to 1:1. From there it is easy to gear up, but not possible to gear down.

I have had requests to gear up, but none to gear down. However, unicycling is quite dynamic these days. Exciting, isn’t it?

With the standard 74mm bolt circle diameter, it will fit chainrings as small as 24 and as large as 34+. That should solve most “gearing up” ratios.

I still think that a solid bmx chainring (for either 1/8" or 3/32" chain) bolted to an ISO disc hub is the best solution, but I myself have no way of drilling the 6 holes concentrically. Anybody?

I want to repeat how clever and innovative I think your use of the spider on the wheel is, unisk8r.

Thanks Dave!

One other item of note, on my 2nd wheel where I welded on the spider, I used a sealed bearing bmx hub with a 14mm chro-moly axle, drilled out for an 8mm axle bolt instead of a nut. The bmx’ers have some nice stuff these days…if they ever start using disc brakes, I think we’d benefit greatly.

Pete Perron