Accuracy question

[mcmark]

Hi folks. I have a PRS Alpha that I am getting ready to cut my first HDU sign with. I have cut a sample out of styrofoam from Lowes. Here's my problem. I did the roughing and the finish at different times with the machine shut down in between. I tried just using the machine self zeroing feature, but I ended up with the finish cut shifted about .200" in x and z. How repeatable is the zeroing feature? Or should I use my metal knowledge and zero from a pin that I can put in the material? Can I safely assume that the machine has held its zero between the roughing and finishing or should I re-zero?

Many thanks in advance
Mark
(new shopbotter from Pittsburgh PA)

[jerry_stanek]

If I start a cut that I do a rough and finish I zero x y once and move the spindle to a spot on the table and make a note of where it is then zero both the rough and finish at this spot. I don't re zero the x and y.

[ssflyer]

Hi Mark, and welcome to the forum!

First a couple of questions - when you mention the "machine self-zeroing feature," are you referring to the C3 command that zeroes X & Y off the proximity sensors? You said that the machine was off ~ 0.2" in X and Z - Did you mean X and Y? The C3 command doesn't zero the Z axis. I've found my machine to be pretty much right on the money using the C3 command. You will also have to re-zero the Z axis after a tool change.

I've also noted that even when powered off, my PRSAlpha maintains position, though I'll re-zero anyway, to prevent potential issues.

Did you use the C3 command before the roughing pass, too? That should ensure repeatability.

Another thing I do when carving 3D designs it to Z-zero to the table, instead of the material top. Makes it easier to Z-zero the finish bit when there's not enough material left on top to do it accurately.

Hope some of this helps!

[mcmark]

Thanks guys,

Yes, I mean in x and y with the C3 command. It might have been something that I did as the machine is new to me. 4yrs old, 2hp spindle, fein vac.

I think I'm going to run a little test with some points on the table and verify that it will touch the points accurately after a few C3 zeroings, maybe turning off the machine, etc just to get a better feel for how well it holds.

I plan on running the job with the same tool roughing and finishing (1/4" ball mill). The geometry is not tight and I think it will save me some headaches for my first sign. I'll z zero with the zero plate once off of the top of the material this time and adjust my technique for the next one.

Could it be that the prox switches were dirty as it was a used machine? I had the machine stop once and indicate that it hit a limit switch when it was nowhere near a switch. I've since cleaned them and will check the bolts that the switches pick up to make sure that isn't it.

Thanks
Mark

[dana_swift]

Mark- the prox switches "should" get about 0.002" worst case repeatability if they are set up correctly. Typically they get 0.001" on various shopbots I have checked for both X and Y.

Prox switches are very immune to dirt, so that is not the factor. Check the gap. Also read Brady Watsons recent post (just a couple days ago) on getting maximum performance from the prox switches, excellent advice there.

Start there, if you are not getting results that good, the gap between the prox switches and the bolt heads is the most likely culprit.

D

[mcmark]

Thanks, will do!

[harryball]

From my experience here is how I handle things:

It would take extaordinary circumstances for me to shut the machine down in the middle of a job that required the same zero reference be exact (i.e in the middle of a finishing toolpath). I've left it on overnight before. I have no fear of leaving a system running. If material is a premium I try to schedule things so I can run the entire job the same day.

Once the machine is on my x,y zero is checked. If it looks good to me I use or I re-zero it. In either case I then consider it good for the entire time the machine is running unless something happens. Once I turn the machine off I do not expect the x,y zero to match the next time I power it up therefore I do not span jobs.

Between roughing and finishing passes you should have enough allowance in the tool path to easily absorb a small difference in the re-zero error after the machine was powered down. I don't consider roughing one day and finishing the next day an issue provided I have secured the material to the table so it can't move and parked the gantry. If a small re-zero error in x,y causes a problem between roughing and finishing then your roughing toolpath is likely too tight.

If by some unavoidable reason I must shutdown in the middle of a toolpath (thunderstorm) I do the following:
Stop the machine, write down the line number where the file stopped.
Take the gantry over and drill a small hole in the excess material somewhere with a whole number (i.e. X = 4 Y = 47)
park the gantry
before releasing any vac holddown I tape, tape, clamp, tape
Before powering down put clamps against the gantry to hold it in place. You really have to be careful not to put pressure in such a way the gantry moves when the power is turned off. I find putting opposing pressure on the gantry on every rail works well.

Once, because the material was very expensive, I decided to lock a block of wood onto the table in an empty area. I drilled with bit down into the block and then powered down. I KNOW MY MACHINE. It does not jolt or bump when powered up so little danger the bit would break.

The small hole in the material is a precaution. I always take the bit back to that location and drop it down slowly comparing the difference. If I find a difference I try to minimize it using various methods. I have also ran an entire finishing toolpath again with a slight decrease in ZZero to absorb the difference. And I have also smoothed a small line out of the material by hand as well.

All that is probably extreme for many situations but when I have a $300 sheet of material on the table I tend to get a little paranoid.

In the end, there are many ways to go about it and you just have to think it through and what works for you.


/RB

[jerry_stanek]

If you are using the same bit you could just make one file with both tool paths and as soon as one is done the machine will go back to zero and then start the next path.

[mcmark]

Well, I did a little test tonight. I adjusted the bolts for the prox switches. Ran the C3 routine and put a test hole in a block of wood fastened to the table.

Ran the x and y out about 2 feet and ran C3. x started to zero but then acted like it hit the limit switch, backed up and went to touch off again and it acted like it hit the limit again but up where it backed off to, then backed off and called it zero. 2 feet off in x! Now come to think of it, it did act like it hit the x prox switch earlier in the week when I was giving a trial run of the sign in styrofoam.

I re-installed the wires into the terminal block and re-ran the test. It worked correctly for 3 times...dead in the hole. On the 4th time, it acted just like it hit the limit switch way early again.

I turned it off and figured that I can just zero x and y on a pin to run the sign on Saturday.

Any thoughts as to what may be happening? Bad prox sensor?

Thanks
Mark

[gc3]

the grounding demons have found you.....