Re: [CAD_CAM_EDM_DRO] DRO intermitent fails

Sven Peter wrote:

> Hello friends,
> Can give me somebody a hint:
> Today I finished the reconstruction of a #2 TOS- mill.(checoslowakia)
> Aligning the Mitsubishi-DRO-ruler I over traveled it.
> First I got no readouts. After opening it and cleaning the ruler with
> contact cleaner it worked only in the test by hand.
> Later I had to readjust the spring and then it started to work
> in one direction but sometimes in the other direction it fails to
> count the travel. I suspect that there is still some oil on the
> glass ruler or the reader doesn't travel right.
> The amazing thing is that the readings do not fail on the same spot
> and not always and there is an area of 25 cm where the ruler works
> perfect.
>
> I saw that the nests where the balls of the connecting rod sit, are
> worn.
> In the first moment I thought it could be fixed with light center point
> hits in order to close the nests. (or some Epoxy)

There are so many designs for these linear readout spars, I'm not
sure what variation you have. But, assuming it is a glass scale
optical encoder, they often use a scheme where the readout head's
analyzer grating is separated from the main scale by little strips
of teflon. If these get messed up, the analyzer grating can rock
when the direction changes, instead of sliding straight both ways.
Also, if the analyzer grating is allowed to twist, instead of sliding
straight parallel to the scale, it will throw off the quadrature
90 degree relationship between the A and B signals.

The tricks to get the wire, bail, link or whatever that drives the read
head back and forth without bowing (which effectively causes
backlash) are many, and these usually take the brunt of overtravel
accidents. When messed up, they can cause all sorts of alignment
errors between the analyzer and scale gratings.

I haven't run across this ball link scheme, I can kind of guess what
they are trying to do, and it sounds better than the music wire
designs I've seen in Teledyne/Gurley spars, for instance.

Anyway, if there is any rocking or twisting of the analyzer grating
when you change direction, that needs to be fixed.

Jon