Re: Digitizers, Renishaw

>
> I hope you won't mind if I add a few comments to your post and maybe add a
> few comments from other posts. I work for Renishaw's in the UK designing
> some of the electronic interfaces for CMM probes. You will understand if I
> don't reveal too much of the design much of which is or was covered by
> patents. Much of the use of the probes is covered on the web site
> http://www.renishaw.com which is due to be completly revamped soon. Many of
> the user manuals will then be available for download.
>
> >Hi
> >I´m digitizing on a little stepper-driven mill using a Renishaw. (got a
> >broken one from the company I worke for and repaird it). The innerparts a
> >quite simple and there ist no need for high accuracy .
>
> You are at least partly right but I think you will find it is much simpler
> to calibrate if the stylus is on the same axis as the mounting screw.
>
> > So I think it´s no big
> >thing to built one. Basic thing: The Probing Arm has to be able to move in
> >every direction to avoid breaking . They manage it that way: 6 little
>
> Sounds like you have one of the big ones. I just measured a probe I have in
> the workshop and it is only 13mm external diameter. The bigger ones are
> normally used on the big machine tools and are designed to cope with long
> stylus with large balls on the end. You may find that it is rather stiff
> for digitising especially on soft materials.
>
> >Steel-ball (4mm) are fixed in a circle of 30mm diameter. always two with gap
> >of 2mm between them , half buried in electric isolating material. on them
> >lies a 3-arm-Star of 4mm matrial making an electrical connection while it is
> >pressed with a soft spring on them . This is now forming an electrical
> >circuit . Lifting an arm of the star would break the circuit. In the center
> >of the star ist an arm leading straight down through a much bigger hole in
> >the center of the isolating material ,60mm long with a little ball on the
> >end. The star is always centering itself, drawing the ball in each direktion
> >makes one star-arm lift off the balls and breaking the circuit. And that
> >simple thing costs about 1000$- but the idea is great.
>
> The guy who had the 'great' idea just received a knighthood in the new
> years honours list.
>
> >The accuracy is about 0.005 mm.
> >( the dimensions are as kept them in my mind and might vary a little)
> >The Software I use is simple but good, CNCDIGIT. cost about 150$( as far as I
> >remember) remeber
> >Regards
> >Peter
> >
>
> There are two types of probe used for digitising. The simplest is the balls
> and rollers probe as above. This is basically a switch where the contacts
> of the switch are also the physical means of alligning the probe tip.
> Variations of this use piezo sensing or strain gauges to separate the
> contact sensing from the mechanical allignment. Other probes use optical
> sensing, LVT's or LDT's (and probably others). All of these mechanisms
> have a means of deflecting under pressure but returning very accurately to
> the original trigger possition when the pressure is removed. The CMM will
> need to be programmed with a rough track (possibly under hand control) to
> start with and the trigger signal is used to initiate a reading of the
> machines encoders.
>
> The second type uses 3D linear bearings with weak spring centring. Each
> axis has an encoder and there is no switch as such. The CMM follows a
> programmed track across an object and as soon as a deflection in the
> encoder readings is observed it moves the appropriate axis to return the
> encoder to its orignal position. The CMM then reads its axis encoders and
> notes the position. With this system the probe scales do not have to be
> terribly accurate just reasonably linear. There are variations on this
> system but it obviously requires a fair amount of programming in the CMM
> controller.
>
> There is a third type using a laser but you don't even want to know how
> many noughts there are on the price for those. They are great for soft
> fabrics. Try digitising a car seat with a normal probe!
>
> Regards
> Bill Phillips