I almost said this had no effect because I switched my springs left to right but now that I'm really thinking about it, The foam may be skewing in the opposite direction now.

Along with the C16 Guide kit, Mark and Trish sent me some new springs, I'll take the ones off the machine and check them all to get the two closest matches.

The C16 Guide kit makes this somewhat moot however, it guides the foam perfectly. You just have to make sure the two edges of the foam are reasonably parallel. Assembly of the kit is very straight forward and easy, if Mark wants me to I'll post a separate topic showing some pictures (he might want me to wait until they have them listed in the store)

Now I am still losing steps on the X-axis, even with the guides not engaged. Mark also sent me a replacement X-pulley since my original plastic one had split. When I put it on the machine I could not keep the belt from skipping no matter what. So I went back to my slightly larger temporary pulley and while the belt wasn't skipping I was loosing steps. After some suggestions about checking how freely things rolled I decieded to put the correct pulley back on, because if I could find out why the belt kept on skipping then I might find any other problem.
Well, I've had some success with that. I took the X-Stepper off the machine and checked the X-belt tensioners. I noticed that the large fender washers on those had rub marks around there edges. My steppers are tapped for 1/4-20 screws instead of the original nuts and bolts that came with the P3. Since I can get those tighter than the original hardware I theorized that I had tightened them down to much and compressed the nylon spacer on the inside so it made it harder for them to turn. While checking this I also noticed that if I just spun the roller it seemed to roll fine, but If I put any sideways pressure on it so the black roller pushed against either fender washer then it became much harder to spin.
So I cleaned all the roller parts (they had some black gunk inside them) and used some spray silicon on them and now they roll very free even with the side force applied. I put all this back in the machine with the correct drive pulley and guess what, the belt doesn't skip anymore!

So I'm getting closer, I'm still losing steps on the X. It's very consistant, about 1/32" (really the smallest I can measure with any accuracy) for every 40" travel back and forth.

After I get the Springs all checked out I'll see if that helped the X step loss, and I'll run the foam with the guides removed to see if the skew is better/gone.

 

Gefahren, I replaced the plastic tension rollers with bearings. Maybe something to do while you have the thing apart. I got the idea from (electric car) Neil.

viewtopic.php?f=250&t=2859&hilit=plastic&start=75#p30732

 

Ok, I've replaced the tension rollers with some flange bearings I had laying around. It however is not making a difference.

I recalculated the amount of x travel lost at about 1/64" per travel 40" down the X and back. 1/64" may not sound like alot, but it's cumulative, so if I travel that distance 8 times then i'm 1/8" off.

It does not seem to matter what speed I am doing the jog at, it appears to lose the same amount wether I'm jogging at 400 ipm or 40ipm.

Just for the record, I'm using Mach3 (I just made a fresh profile but that didn't help either) and a Smoothstepper going to the Easy-CNC driver board. I have found an old parallel port cable and will try it without the smoothstepper tommorrow.

Any other suggestions??



edit - I just tried it with the microstepping set to 1/2 instead of 1/4 and it had a greater error (not double). Tommorrow I will also try a new driver chip(since I have a spare) But I have really low confidence that it will help any.

 

Gefahren, loosing, or gaining, distance could have three reasons, off the top of my head. One is, obviously, lost steps. Two, the material is slipping. Three, CALIBRATION. I calibrated my PP using a dial indicator. I find that using a caliper is not precise enough. If you have a calibration distance of, lets say 4" (like I did), and you have a mistake of only 1 or 2 thou on a 40" travel the error is 1/64" or more. I used a 2x3x4 block to make my calibrations. BTW, I found the x axis to be the most difficult axis to calibrate do to the nature of the PP.

 

hmm, I don't have a dial indicator, just a digital caliper.

Maybe someone could explain it (I'll check the calibration as best I can either way) but how could a bad calibration cause it to cumulativly lose steps?

for example, lets say the calibration is way off, 1" in Mach only equals 3/4" in the real world. If I command a move of 40" then the machine will only move 30", then if I command it to move back 40" it will still only move 30", So I'm back at the same point I started from.

What I'm seeing is after the first move the bit is off a little bit (I have it peck a hole in the foam at the zero position). Then it's off a little more with the next move and so forth until after 8 moves it is sitting approximatly a full 1/8" short of where it started. If I keep commanding the back and forth travels the error continues to grow.

As per your example, if you had a calibration error of 2 thousanths on a 4" travel, then that would be more than 1/64" over the 40" travel(which is still pretty darn close), but it wouldn't be cumulative since the error would occur in both directions.

 

Did you try to cut a small circle, like 1" diameter, preferably in wood? It will show if you have backlash, which could be the problem with cumulative errors. I did a very shallow pass in MDF and it showed when I had backlash. The hole was not round. I still have a slight problem but I'm learning to live with it. :girl_cray2: