PDA

View Full Version : Help! Calibration Creep



johnlew
01-14-2007, 07:49 PM
Help and suggestions greatly appreciated. I am running Mach3 on a 24" x 36" router table with Xylotex driver and 269 oz stepper motors. I originally calibrated the table with Mach to 1" being spot on. I then discovered that I have "creep" in my movement. Over a 12" movement the creep was about 3/4" so the screen was showing y at 11" but was really out 11 3/4".

I am trying to recalibrate. When I start out stepping 1" it is about 1/64" under. By the time I step out to about 4" I am about 1/16" over. But when I home it returns to zero right on the money.

I can understand how I can calibrate if I am consistantly over/under. But when it starts out under and ends up over I need help.

Assistance is greatly appreciated.

GaryCorlew
01-14-2007, 08:09 PM
what are you using to move the table? ballscrews, acmescrews? threaded rod? belts?

johnlew
01-14-2007, 08:38 PM
Gary,

The table is a DynaCNC table with acme screws. The steppers are connected via timing gears on the motors and screws. They have a 24 tpi belt that drives the screws. One thing to mention is that the timing gears on the motors is smaller than those on the screws.

Thanks for the help.

InspirationTool
01-14-2007, 08:46 PM
Gary,

The table is a DynaCNC table with acme screws. The steppers are connected via timing gears on the motors and screws. They have a 24 tpi belt that drives the screws. One thing to mention is that the timing gears on the motors is smaller than those on the screws.

Thanks for the help.

Well, if Mach is set up for 24 rotations = 1 inch, that would do it.

If you can accurately measure movement, you can just change the steps per inch setting by the ratio of actual to expected movement.

So, say you have 1500 steps per inch in Mach. You tell it to move 12 inches, but you get 13. Multiply 1500 x (12/13) to get the correct amount.

You can also count the teeth on the timing gears to get the correct ratio, assuming the original steps per inch were right for a 1:1 ratio.

-Jeff

GaryCorlew
01-14-2007, 08:57 PM
is your power supply away from the driver board? you may be getting noise from the power supply? what it have for a spindle? are the wires that send power to the spindle away fom the wires that send the step/dir signals to the steppers? is this happening with the spindle off?

GaryCorlew
01-15-2007, 09:54 PM
Have you made any progress sorting it out?

erase42
01-15-2007, 10:42 PM
Is this only happening on one axis, or all? If only one axis, does it the problem follow a certain driver, or does it remain no matter what channel you hook it to. You need to determine if it is electrical or mechanical in order to eliminate half of the guessing. They are open loop steppers correct? There are no encoders or feedback of any type?

johnlew
01-16-2007, 08:23 AM
is your power supply away from the driver board? you may be getting noise from the power supply? what it have for a spindle? are the wires that send power to the spindle away fom the wires that send the step/dir signals to the steppers? is this happening with the spindle off?

You gave food for thought. The power cord for the spindle is cable tied to the stepper cables. Never thought of interference when I secured them. I will seclude the stepper wires and try that. The spindle is a Porter Cable 1 3/4 hp. The controller was assembled by Xylotex and has the power supply inside.

I tried to recalibrate yesterday with the same results.

Thanks

johnlew
01-16-2007, 01:12 PM
Is this only happening on one axis, or all? If only one axis, does it the problem follow a certain driver, or does it remain no matter what channel you hook it to. You need to determine if it is electrical or mechanical in order to eliminate half of the guessing. They are open loop steppers correct? There are no encoders or feedback of any type?

It is happening predominantly on the y-axis. The deviation on the x-axis is acceptable. They y-axis is driven by two steppers. I tried running it with the A-axis being a slave to Y, and also tried wiring the motors from the y source so they have no latency in signal.

I am trying to sort things out and am at my wits end. I will continue to test. If anyone has had experience for a awhile with these maybe you can answer a question: I have this setup in an unheated/uncooled garage. Will temperatures from say 45 deg. F to 90 deg. F effect the calibration?

Thanks.

PowerCNC4u
01-16-2007, 02:23 PM
[QUOTE=johnlew;242373]Help and suggestions greatly appreciated. I am running Mach3 on a 24" x 36" router table with Xylotex driver and 269 oz stepper motors. QUOTE]

Are you using the Xylotex switching power supply? It sounds like a problem with the power supply not being able to provide the correct power to a 4 motor driven machine. I know DynaCNC uses an unregulated power supply to power 3 or 4 motors. I also know that Xylotex sells a switching power supply that is ok for 3 axis / motor machines, but I don't know if it has enough power for the 4th motor. You could be losing steps because of a lack of power. It may also may be a lose connection on the wiring. Did you solder the ends, crimp the ends, or crimp the ends onto a connector somehow? Considering the cables are moving all the time on a cnc the best way to deliver good current is to crimp them without using solder. the solder will break or crack with the movement.

Sid