a diy rowing machine

don't ask how me why, but it seemed necessary to build a rowing machine myself. I had a rough look at what commercial rowing machines do, and then came up with this version. You can always click the pictures on the right for a higher resolution version for all your zooming pleasure.

The basic frame is made of L-shaped profiles which each are made of cheap timber strips measuring 19x44mm. This is sturdier than it sounds, but if I had to do it again I would opt for one size heftier. The second, upper frame is essentially the same just shorter. The rest is quite obvious except for the fact that there was some intricate, puzzle-like ordering problem when putting it all together. unfortunately I can't remember the details, so you will have to try yourself ;)

These pictures obviously show the full machine from the front, back and side. This may come in handy when trying to put the rest into context, and also demonstrates that the whole thing stores upright.

Seat, rails and footrests are actual sculling accessories, thanks ebay. I think that's way cooler than the dinky crap you get on commercial machines and also gives it some extra flair.

The handle is a bit less authentic, as is the rest from here onwards :P. But notice the great velox handlebar tape!

The rope from the handle goes through the right section of this contraption, then through the mechanism in the next picture, through the middle part of this, through the next picture again, and then over the left part of this picture to the bottom side of the machine. Initially I had smaller pulleys here like the one on the bottom side, but they don't have bearings (cheap shit), and melt after about 1 minute of use. They are quite hard to replace as well, so I came up with this instead. Since pulleys with bearings are quite expensive I decided to use just bearings, which are dirt cheap (anyone care to explain why cheap bearings + cheap pulleys make expensive pulleys?). The middle section is only that wide because there were that many bearings in the pack, but as it turns out that is quite a good feature as it allows the rope to move over the bearings and thereby reduces twist. Larger washer with smaller washers on each side separate the three sections. You need to be carefull with the size of the smaller washers, the three sections spin in different directions!

The core of the rowing machine is a standard bicycle rear wheel. The rope from the bearing-thing goes around the freehub once, back to the bearings and around the freehub again (same direction of course). Obviously it needs to go the right way around, freewheel and all. The problem is that the rope keeps falling off the freehub, gets entangled with itself or other stupid things, so I cut three pieces of perspex as guides and glued them between MDF sheets cut to horseshoe-shaped bits. That already worked much better, but now the edge of the perspex was rubbing against the rope and quickly eating away at it. so I had to take the whole bugger apart again and bend (heat+swearing) the end of the perspex splightly, which is hard to see in the picture because it bends right towards the camera. Two turns seem to be about right, one is definitely too little friction and the rope keeps slipping. It still slips when you jerk on the handle, but then that can be called a feature, after all you can't jerk on the oars of a real boat either.

as you can see the rope is still quite frayed from the earlier trials, but it doesn't seem to be getting any worse now...

once all of this is done, the rope passes to the bottom side of the rower where it goes through a pulley (the small one), and to the other (seat) end of the rower...

where it is attached to a shock cord. To get enough shock cord into play to allow enough pull, yet not too variable pull, two more pulleys are used. So you get two to almost three rower length of shock cord. Ignore the red stuff, it just keeps the seat in place for storage.

Now comes the tricky part: how to brake the bicycle wheel? I am using an eddy current brake made of 2 sets of 41 neodymium magnets measuring 8mm dia x 5mm each. these are superglued into holes drilled into MDF. I picked the strongest magnets I could find (N52), and these little buggers are surprisingly strong, it is for example quite difficult to pry them of the fridge with your fingers alone. If you ever felt like you were at the wrong end of some divine curse (think sisyphus or prometheus), try putting 82 neodymium magnets covered in superglue into the right places! I opted for a "through" configuration where the same poles (N/S) are all going into the same direction. I am not sure whether that is the best (and have not managed to do that entirely) but it seems reasonable to keep the field together through the rim. the mdf plates with the magnets are held by 4 threaded bolts and can be adjusted with nuts as seen in the picture. the adjustment is quite fiddly, a millimeter here or there makes the difference between stopping the wheel almost immediately to virtually no resistance. together with some flex in the frame that's fun to get right.

finally, the whole thing needs to be attached to a computer, of course. a small arduino does the job of measuring wheel speed and heart rate and reporting both through usb, a small perl script on the computer computes (as they do) elapsed time, average speed etc. it also let's you row against your own ghost from the last recorded run! of course no UI, just a line of changing numbers in an xterm!

all in all good fun for the whole family, but perhaps not the best rower in the world. also quite a lot of work, certainly not worth doing to save money on a real rower... If you build one as well, I'd like to know what it looks like: robertle@semistable.com