PDA

View Full Version : backlash or loose nut on a chester 20v mill



lukecnc2012
09-08-2015, 05:31 PM
Good afternoon all.
I've been working on getting my chester 20v mill to be as accurate as possible with it at the moment being .2 out.
I've used a DTI to check for backlash and gone x -10 the back to zero and it is great using g0 command.if I move the machine on feed to say -5 then go back and forward quick then back to zero the machine is showing 0.22 deviation.ive taken the machine apart today ad I'm going to convert it to ball screws if I can find the right screws etc.Just before I go ahead and buy the kit I wanted to just ask to make sure this will help eliminate the issue.

Any help much appreciated.

Luke

Clive S
09-08-2015, 07:05 PM
It would defiantly make a difference using ball screws as apposed to trapezoidal screws. ..Clive

lukecnc2012
09-08-2015, 07:13 PM
i will hopefully be placing an order tomorrow with damencnc for a new X and Y axis screw kit.Then hopefully i should be able to get some reasonable repeat ability.as at the moment i cant get it anywhere near 0.15

lukecnc2012
13-09-2015, 12:37 PM
Hi all i received my ball screw kit from Damen cnc last week and got it finally installed today.I set the steps etc and adjusted for any backlash if hardly any.but upon doing a circular test in aluminum at 20mm diameter i am still getting big difference in size 20.24mm in X and 20.29 in Y. i had the same problem with the standard screws which is why i changed as i though they were the issue but it looks more now down to me setting it up.Any advice much appreciated.

komatias
13-09-2015, 12:56 PM
Luke,

0.05mm is probably the best you will get without spending megabucks on the actual bearings, preloaded ballscrews with very heavy preload and loads of tweeking and an autolubrication system. Just for your information, if you were to ask 0.05mm tolerances on a drawing sent for RFQ on a milled part, you may be shocked at the price difference it will have compared to a 0.1mm tolerance.


You will also be losing motion in the actual friction of the machine. Have you installed and autolubrication system yet?

When you did the circle test, did you do roughing and then a finishing pass or two? Small endmills may introduce some issues too so you get around that with a finishing pass that is easier on both cutter and machine.

Finally, how did you calibrate the x and y steps/unit? DRO? measurements or based on the 200 steps per revolution and screw's pitch.

Another thing to note is that you probably have C7 accuracy ballscrews if I read Damien's site correctly. These cannot be preloaded evenly throughout their travel. They are rolled and have an error allowance of 0.05mm/300mm of travel. The price goes up exponentially with accuracy of course.

So your best bet is to fiddle with the backlash compensation of the CNC program you are using (MAch/EMC) and be happy with what you have. You will definitely be hitting the limits of the machines accuracy too. I had my Chester Superlux re-scraped it was so out.

Feel free to drop me a line if you need more detailed help.

George

lukecnc2012
13-09-2015, 01:05 PM
Hi George.
Thank you for your quick reply.im using mach3.i completely understand in regards to accuracy come at a cost which in the coming future I will be looking for a new machine.im yet to fit the oiler which I've got boxed up still I hope to have this fitted this week.i ran some parts a few months ago with an accuracy within .1mm which I found fine to do but upon running it again today it's not even coming in close.ill take your advice and keep adjusting the backlash to see if I can get it any better.

Thanks again.
Luke

komatias
13-09-2015, 01:18 PM
I assume you are using a motion control board too right?

lukecnc2012
13-09-2015, 01:28 PM
yes thats correct