We haven't gotten to run comprehensive driving tests to see just how well it performs - that'll be sometime this week after programming gets a little time with it. The gears are definitely more difficult to backdrive than a standard spur gear setup, although ours are quad thread worms - which really makes them halfway towards being helical gears (part of why they are still backdrivable).
As far as efficiency is concerned - after some research and calculations its not nearly as bad as you might think, although there are a lot of factors and variables that weigh in. By going with worm gears we are able to get our complete reduction in a single gearing, compared to typical spur gear setups that generally need a pair of gears to get into a reasonable speed range. Also the quad thread makes it significantly more efficient than a single or dual thread worm because you don't have as much sliding friction against the gear face, and the comparatively low reduction for a worm drive also helps. I ran some rough numbers and came out with ~86% efficiency for a setup like we've got. Considering a pair of 95% efficient spur gears would add together to equal about 90% efficient we aren't that far off from more standard solutions.
Also of note for comparison are the planetary CIM gearheads that came in the kit this year.
In another thread Joe Johnson estimated them to be about 72% efficient ( 85% per stage ). So if we get anywhere near my calculated efficiency, we'll be at or above what a lot of teams are running this year.
As long as we keep them lubed well enough we shouldn't notice too significant of a loss to the worms. We'll have more real-world experience to report on this by the end of the week, however.
- Jeff