Sieg X1 CNC backlash compensation test

[kreutz]

X1 micro-mill CNC converted Backlash Compensation - YouTube

[wildwestpat]

Sorry but that video appears to only be checking the repeatability of a single axis to repeat a move and is not demonstrating backlash removal. Your video is very good at showing the repeatability but you are approaching the set position from the same direction each time.

Backlash compensation is to accommodate the error found when approaching a reference point from either a positive or negative direction hence the table is zeroed to a DTI moved positive "A" inches then reversed "2A" and then reversed again back to the original position where the test indicator will verify that the backlash has been compensated for or show the backlash. BUT and it is a very big but this may not work well for circle cutting where the direction of cut varies with the angular placement with respect to the two axis being driven to generate the circle.

Regards - Pat

[kreutz]

Sorry but that video appears to only be checking the repeatability of a single axis to repeat a move and is not demonstrating backlash removal. Your video is very good at showing the repeatability but you are approaching the set position from the same direction each time.

Backlash compensation is to accommodate the error found when approaching a reference point from either a positive or negative direction hence the table is zeroed to a DTI moved positive "A" inches then reversed "2A" and then reversed again back to the original position where the test indicator will verify that the backlash has been compensated for or show the backlash. BUT and it is a very big but this may not work well for circle cutting where the direction of cut varies with the angular placement with respect to the two axis being driven to generate the circle.

Regards - Pat

Backlash compensation has already been set and enabled in Mach3 as shown on the initial video frames. What I'm showing is that it will come back to the exact position (within 0.0001", as shown on the dial test indicator) several times after traveling back and forth 1" which include two changes of direction and so involves Mach3's backlash compensation. Backlash on this axis is only 0.000144" (average over 20 measurements) and I don't really need to compensate it for most of the jobs. Circular cuts and pockets in Alu and Steel are perfect even without backlash compensation enabled.

Thanks for your comments

Best regards,

kreutz

[wildwestpat]

Hi kreutz

Sorry to be persistent and we will have to agree to disagree as in IMHO you have achieved good repeatability of the same move under CNC and this looks particularly good using the standard screw, brass nut and bearings. A fantastic achievement with this mill using the oem parts.

Hope the following explains why IMO you are measuring repeatability and not backlash. Backlash is experienced in cnc between directed and achieved motion as this kicks in when the direction is reversed and the expected motion of the table with respect to spindle takes place. You need to measure the desired motion looks to be one inch and check this is what has been achieved. There is often confusion of repeatability backlash (lost motion) and scaling as each has an individual impact on the way errors creep into the final work pieces.

Normally to achieve 0.0001 accuracy most machines get converted to ballscrews with double nuts and angular contact support bearings replace the oem screw, brass nut and simple bearing support of the screw.

Like your other videos very inspirational.

Regards - Pat

[kreutz]

Hi kreutz

.............. Backlash is experienced in cnc between directed and achieved motion as this kicks in when the direction is reversed and the expected motion of the table with respect to spindle takes place. ..............

Regards - Pat

Thanks Pat;

I agree, my point is that without backlash compensation several movements in a row, changing direction at the end of each one, will provoke to end up in a different position from the starting point due to losing steps (because of the backlash) on each direction change. Which in my opinion proves that backlash compensation is working (in this video) and the backlash figure set is the right one (0.000144"). I used the 0.0001" dial test indicator in order to make sure there was no appreciable offset. That is the reason I titled the video "backlash compensation test" instead of "backlash measurement".

BTW, the original (oem) nuts are cast iron. I don't know how long they will keep the current backlash adjustment. The gibs were changed to brass gibs and that alone made a big difference.

Best regards,

kreutz