Re: Re: What do we want?

In a message dated 10/19/1999 1:16:06 PM Hawaiian Standard Time,
Ian@... writes:

<< In theory, a chain drive should be less accurate than a cable as movement
round the sprockets is not continuous. That is, the links of the chain go
round the sprocket as a series of straight lines rather than a pure curve.
Whether the resulting (small) error is significant in your application would
have to be a part of your considerations.

>>

Ian,
I think the error is called "chordal rise". Worst case would be for a
sprocket with a hypothetical tooth count of three. Cog belt drives do not
have this error to any significant degree and are used almost exclusively for
this type of drive. Think of a flat belt with smooth round pulleys and then
cut notches for projections cast onto the belt. Most higher end machines use
HPT metric belts rather than the older x series belts. I know there could be
some variation in belt or sprocket pitch but in practice, I have not seen it.
If you go to some of the larger drive products outlets, they sell cog belt
strips up to 10 or 12+ feet long. This can be used for a nice setup as belts
with Kevlar or steel cords have close to zero stretch. Another plus is that
the belts require no maintenance if kept clean. Chains will require oiling
etc. Though not as bad as leadscrews, keeping them clean will require some
ingenuity.
Peter