On Wednesday, April 11, 2012 09:25:43 AM Rolf Bredemeier did opine: > Hi, > > at time i am trying to convert a little desktop-lathe (WABECO) to CNC. > In the next day i will build my own spindel encoder. I need it only for > threading. > The following is valid for the generic 7x10-12-14-19 lathes which I believe your WABECO may be.
> The machine will be driven by steppers. Waht is the best choice for slot > count for the encoder wheel? (I will mill it.) > > I think, i can make wheels from 1 to 64 slots. > > What is your guess? > > > Greetings from Germany! > I was somewhat restricted in choice of bit size to mill with, not to mention that space is pretty restricted if it is to allow the change gear cover to be re-installed. So I chose to mount the opto devices in a row, hanging down so the slots of the opto's straddled the wheel. Because I was trying to achieve good quadrature at it's output with 1/16" mill, I played with the number of slots until that looked good, and wound up with 39 slots, with one extended inward slightly to function as the index. The opto's, being mounted on a common pcb, then held the center one on the circle that was the index pulse, and the 2 outer ones then ran on the slightly larger slot circle. A pix of its output sans the index as my scope is only a dual trace is there too. Code to carve that wheel, and eagle/pcb2gcode files to make the PCB for the opto devices, which were Honeywell H0A2001's can be had from my web page, in the not terribly well marked "Genes-os9-stf/eagle" sublink. The file 'genes-encoder.ngc' will carve the wheel, in my case out of .040" thick brass of a door kickplate but any opaque, sufficiently rigid material can be used, and includes a quite precise fit on the back end of the spindle where I trapped it between the 2 bearing preload adjustment nuts. The finished wheel is disk4.JPG. As a scan, not a camera shot. It actually looks better in the halscope view. The file 'PCB-Pallet.JPG' gives a somewhat out of focus idea of what the pallet I made to hold the pieces of pcb on the mills table looks like. Unseen is a fin about .050" high on the bottom that exactly fits the tables slots, assuring proper X alignment if its removed and subsequently remounted on the mills table. The bit of brass tubing inlet on the left front and connected by the tab of metal under the left front screw serve as the target for the 'tholefinder.ngc' file, which combined with a non- damaging probe made from the wire of a sewing machine needle threader, mounted in the spindle and running at about 600 rpm while tholefinder is running, serves to locate the pallet and setup the G55 and G56 co-ordinates used by the board etching files to make the top and bottom of the boards come out in registration. I drill the thru-holes a little over halfway through the board and they meet perfectly in the middle. The (t/b)edautoz.ngc files also contain a fudge factor variable so you can control the etching depth, and these file also use the probing function G38.2 to find the exact height of the board, and do this for every time a tool change is called for, lowering the tool itself, non rotating, to touch the board. Slow enough that no marks can be seen on the board. I did have a bit of judder in the Z motions my mill made, about a thou, if I didn't keep the post well lubed with Mobil Vactra, giving the post a shot above the gibs every time I loaded the next file in the sequence. If this stuff is any good to you Rolf, be my guest. I am in this same process, unk at the moment whether or not I have to put ball screws in it, backlash, using the existing Z screw drive, looks as if it might be a problem, but I've only cut air so far. Its in the .020" range, and isn't terribly well defined, needs half nut fine tuning. Badly. I have a part I want to make, and how well it can make 10 in a row will determine if I spend the $$ on a ball screw. I need to take some more pix now that I have the motors mounted and running things. Maybe I can do that today, while I'm waiting on a new spindle controller board. I had an attack of dumbass and blew the one I just about had working into a pile of slag, forgot its line voltage hot when I went to clip my scopes ground lead and check a waveform. Dumb... One of the things that would help this lathe quite a bit would be to ditch the compound slider, and mount a much stiffer tool holder directly on the cross slide, letting LinuxCNC handle the compound sliders job. This would allow the tool itself to be much better centered on the carriage and reduce the tipping under cutting forces that are a real PIMA the way it is now. Since that will need a much taller holder, is there a QC holder that would hold at centerline height if bolted directly to the cross-slide? > Rolf Cheers, Gene -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) My web page: <http://coyoteden.dyndns-free.com:85/gene> Bachelor: A guy who is footloose and fiancee-free. ------------------------------------------------------------------------------ Better than sec? Nothing is better than sec when it comes to monitoring Big Data applications. Try Boundary one-second resolution app monitoring today. Free. http://p.sf.net/sfu/Boundary-dev2dev _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
