[zeed]

I am having some strange problems with my second build,I cant calibrate all 3 axis correctly.
I can make it to go exatly 30mm,making circles perfect,repeatability tests all ok but when I make for example 150mm circles it goes 151mm.
OK,then recalibrated all axis,but then goes perfectly to 150mm but when tested the 30mm test it makes 29mm...
I have no backlash that I can mesure.
I am using 12mm x 1,75mm pitch
threaded rods,directly coupled to my 2Nm stepper motors.
Using this drivers:
2/4 phase stepping stepper motor driver 70V/6A - eBay (item 300528146572 end time Apr-20-11 07:56:00 PDT)

and 50V power supply.

[atwooddon]

Zeed, what control software are you using? Mach3, EMC2, etc?

Don

[zeed]

Mach3...

[atwooddon]

Are you using the Mach3 calibration function?

It is interesting that you get a calibration error going both ways. Usually, if you calibrate over a short distance then check longer distances, any errors in the short distance calibration will compound resulting in something like your 151mm instead of 150mm expected. However, when you calibrate over longer distances, then measure shorter distances, any errors in the longer calibration are minimized at shorter distances.

Don

[zeed]

Well,things just got worst...
When calibrated on 30mm distance,it gives me 146mm at 150 distance.
when calibrated on 150mm distance it gives 31mm on 30mm distance.
I used mach3 calibration function.
Hard to explain with my bad english....so made some video and uploading to youtube now.

edit:video

[zeed]

Well,tried today again ti fix this problem,but no luck...
If it works fine on long distances then small distances are problem and vice versa..

Is it possible that threaded rods are not linear with pitch?

[atwooddon]

Well,tried today again ti fix this problem,but no luck...
If it works fine on long distances then small distances are problem and vice versa..

Is it possible that threaded rods are not linear with pitch?

It's possible but would be very unusual. Have your tried moving to a different part of your table to check the calibration? For example, if you are checking in one corner of the table, move to the other corner and retry your test. That might indicate if the threads are not consistent.

Also, are you sure you are not losing steps during any of this testing? That would explain the longer calibration to shorter distance check errors.

One other thing, does the same error occur on all 3 axis or just one?

I must admit this is a strange problem. One thing I have run into is the calibration function in Mach3 was problematic over longer distances. So, I calibrated manually by adjusting the steps per unit in the motor tuning section of Mach 3. If you have not tried calibrating the steps per unit in the General Config --> Motor tuning section, you might try that.

Don

[zeed]

It's possible but would be very unusual. Have your tried moving to a different part of your table to check the calibration? For example, if you are checking in one corner of the table, move to the other corner and retry your test. That might indicate if the threads are not consistent.


Don

BINGO !!
That is it,on one part of the table(most used) it is impossible to calibrate it correctly,but on the other side it is almost perfect.

It is time to put some new screws on the beast ,thinking of some screws with 3mm pitch.
Dont know how to call it in english but here is photo:


Thank you atwooddon for your advices and help.:cheers:

[atwooddon]

Yippee, Yahoo

Glad you got it worked out.
Don