[dcigir]

Hello to all, i run into the problem and will try to explain it and hopefully someone can get me to the right solution.

I made a CNC table. To remove some first time user issues this is my third table. So this time I decided to make large table, work area 1150mm x 1100mm. Because of the mechanical construction and weight of the gantry I decided to go with 3:1 ration on X and Y+(Slave A) axis.

Machine is assembled and in working condition, but the problem I got is with Axis calibration. On my Z axis i used Digital micrometer and measured travel distance of the axis, puit it in Mach 3 axis calibration wizard and now I have Z axis calibrated and dead nut in 0.00 decimal.

But the I try to make the same thing on the X and Y axis, and when I measure it, and put it in the Mach 3 wizard, at first it seams that is everything OK. When I move for example X axis from 0 to 700mm, axis moves 708mm.

The I thought I have wrong measure and give it another try. Result is the same. Then I use analog micrometer and try to do calibration with this new one, but the problem is the same. Right know I have X and Y axis step per unit set at 150, because only with this I can get approximately travel of 700mm when I move it 700mm. My digital micrometer has travel only of 12.7mm so I'm unable to measure it for long distance travel.
The only thing that is different from my previous built is that I have now 3:1 ration on this axis but I can't figure it out why I can't calibrate this to be dead nut.

Is there somebody to help me? I hope that I have explained with what I have problem.

Thanks
Denis

[Littlemango]

I have noticed the wizard seems to not compensate for reduction....at least with my rotary. It has 2:1 reduction and in order to get the wzard to work correctly i had to split the error in half and tell mach 3 that number. For instance if I asked for 360 and it went 364 i would have to tell it it went 362. If it had gone 356 i had to tell it 358 to have it not over-correct. To be honest i havent even analyzed exactly why that is and would have thought i would have to tell it double the error but i suspect its the reduction ratio. In your case i would try the tactic and account for the fact that you have 3:1 reduction when you give it the "distance travelled". So tell it the target travel plus or minus 1/3rd of the error is how far it went , depending which side of your target distance it lands.

[jalessi]

Denis,

Calibrate the axis with the longest possible measuring standard such as a 48" machinist ruler.

See attached link:

https://www.ebay.com/sch/m.html?_odk...ruler&_sacat=0

[dcigir]

Thank you, will try this weekend.

[RCaffin]

You CAN calibrate Mach3 using the encoder count, the reduction ratio and the ball-screw pitch.
Then you can check it against a machinist's rule over a long distance. Works fine for me.
Ah - watch out for backlash. If you have some, then it can be extremely tricky.

Cheers
Roger