High torque drive belt for z axis?

Before I discovered Stock Drive Products I had a very hard time sourcing reasonably priced mxl pulleys and timing belt. By random luck, though, I found a bunch of plastic 3mm pitch HTD pulleys at a local surplus store. They had 8mm bores *and* just the right diameter to fit between the studding and the smooth rod, along with a nifty optical encoder edge.

Clearly the universe wanted me to use HTD pulleys. Clearly mxl is for chumps.

So, I obtained some HTD belt, and yesterday (finally!) belted up my z axis. The HTD belt seems to works okay-ish mostly, but I immediately noticed that it skipped when the splice point hits at pulley -- it's clearly too stiff, and can't wrap around the radius properly, causing some slippage.

[I also noticed as soon as I took the splicer off that I'd left a left a full turn in the loop, but that didn't seem to effect the drive.]

Q1 - has anyone experimented with HTD belt for z axis?

Q2 - has anyone experimented with using a channel or constraint attached to the M8 rod in order to force the spliced portion to stay in contact with the pulley?

Q3 -- any suggestions?

Cheers,
L.

My 2 mm and MXL belts do the same thing; despite what I thought was a pretty careful splice, the stiffness, and perhaps a not quite exact distance between the last two spliced teeth causes each pulley to skip a single tooth when the splice travels by. Since it's one tooth in about 800, I haven't worried about it too much, I just check that the bed is level after a few pounds of extruded filament.

I see on the Mendel design, they've moved to a continuous belt with an idler pulley to tension it, getting rid of the funny toothed splice. Years of experience with bicycle chain indicates that fixing a skipping chain is not trivial, especially when it's under a heavy load. Just 0.5% elongation on a bike chain is enough to cause it to skip under load.

One thing that does help though, is to reduce the friction on the Z axis. On my printed Darwin, I've left out the backlash springs and the upper M8 nuts on the corner brackets, and let gravity do the backlash compensation. That reduces the load on the belt, and reduces the skipping quite a bit.

Wade

"One thing that does help though, is to reduce the friction on the Z axis. On my printed Darwin, I've left out the backlash springs and the upper M8 nuts on the corner brackets, and let gravity do the backlash compensation. That reduces the load on the belt, and reduces the skipping quite a bit."

I'll take a look at that. I literally got the belt on, spun it twice, and had to leave to go home and sleep, so I'm not sure how bad the situation is, but doing it by hand, it looked like it was slipping a fair number of teeth, not just one.

L