Problem- backlash

[stevieboy]

Hi Guys,

After reading a recent post on here I think I may have a similar problem. It's a BMC 20 with ultimax 3, 1993 Y.O.M. I think. I've had my doubts as to the accuracy of this machine for a while but it's never been too much of a problem as the things I've made on it have never needed to be spot on. I had to clock a part in on a bore the other day and noticed something worrying.

When I tried to zero the clock in by moving the X axis there was an audible clunk when I moved the hand wheel form side to side. As I move the hand wheel back and forward the DTI needle jumps by 0.02mm but it jumps in the direction that I'm making the axis move. It does this in the Y axis as well. So I tried a test after reading the post about backlash.

I move the DTI needle up to a block on the bed and zero the clock, then move the axis in by 0.3mm on the clock dial then zero the axis part setup. Then I move the axis back out until the clock dial reads zero and check what the part setup reading is. I did this for both X and Y axis and the part setup reading was 0.272 for the X and 0.246 for the Y.

I could do a more accurate test with a DTI that was 0.001mm increments but would have to buy one. Thing is, am I looking at it in the right way. I'm planning a strip down of all the guarding and chip enclosure in the near future to fix all the leaks and cracked welds. If i can fix any other problems while I'm at it then all the better. So what do I need to do.

I don't know a lot about these machines and I learn as I go along so replies in laymans terms please, otherwise I'll get lost. It's the clunking that worries me, does something need to tightend up, or is it just worn parts. It has done 28,000 hours.

Any help.

[bloke]

Hello, Steve.
First, you'll need to check the ballscrew isn't floating in the bearings and the locknut is tight.
If you still have the clunking, it is probably overcomped in the X-axis minus map. Go to the toolchanger diagnostics page and press enter 642 enter.
Select the X-axis minus map, press enter backlash comp and enter the amount of you kick as a minus figure. save, exit and re-check.

[stevieboy]

Ok, cheers bloke.

I'll check it in the week and post again if I have any problems. So are you going to Mach2010?

[stevieboy]

Ok, I managed to spend an hour messing with it. I took the way covers off and had a poke around and found what the clunking was. Where the ballscrew nut is there is what appears to be a flexi plate coupling of some kind with four bolts and locknuts holding it together. The bolts had come loose, on both axis but alot more on the X, fixed it.

I didn't do anything with the ballscrew nuts as I don't know what I'm doing. Do I have to slacken the little grub screws in the flats and then turn the nut with a spanner? How tight should they be? How do I check if they're floating?

I found the lead screw maps screen and that's where I gave up. I can find the maps for all axis both positive and negative. The X maps go from 0" to 24" and the backlash amounts are in millionths of an inch but the figures go up by a different increment for each one. Does this mean that I have to measure each inch of movement from the machine 0. Also how do I measure a millionth of an inch. The figures went -20, -40, -60, -40, -100 and so on.

Any more help?

[bloke]

Once in the leadscrew map page, go to the minus map. A softkey should appear that says "enter backlash compensation". Select this and it will allow you to add/subtract compensation from the whole map in one go.
If for example you had 0.0003" of backlash, simply enter 300 after pressing the "enter backlash compensation" key and it will add 300 to every map position in the minus map and thus comp out 0.0003" of backlash.
You could also just add 300 to every position manually if you wanted. 100 millionths of an inch is 0.0001" - if you do everything in units of 100, it keeps it nice and simple.
Lock the nuts on the screws up tight as the preload is ground into the bearings already.

[stevieboy]

Ok, I managed to find time to have another look. I typed in the figure I needed to in the enter backlash amount and it made it worse. Trouble is now I can't remember if i typed it in as a plus or minus figure. Anyway, after doing a bit of messing about trying to get it back to where it was I gave up. I've only altered the X axis minus map, everything else is untouched.

So in an ideal world where there was no backlash should the reverse map be the same figures as the forward map but obviously in reverse. If the figures in the forward map are all minus then should all the figurse in the reverse map be positive.

I tried getting the old figures back but the new ones keep coming up again. If I set them all to default they go to zero and the kick on the needle dissapears. But if I reload them the new ones appear not the old ones. Not sure what I've done to cause this.

I know I can sort it out, just need to know what I'm doing, or more importantly why. If I had a better understanding of the whole thing it would help. I know the maps are to compensate for any tolerences in all the moving parts, ball screws, bearings etc but what it the difference between the forward and reverse maps.

I checked the bearings and they're all tight, no kick on the needle at all when I put it on the end of the screw.

Any more help.

[bloke]

In that ideal world, the plus map would be the same as the minus map. What you need to do is:
Load the map you have.
Make the minus map the same as the plus (so the positions on the screen have the same values entered in the minus as the plus I.E. if the 14" position reads 180 in the plus map, make the minus map read 180 also).
When you've done that, enter the backlash compensation in the minus map. A positive figure will reduce backlash, a negative one will increase it.
If you find you have 0.001" of backlash, enter +1000 into the compensation.
If you find you have 0.0005" of backlash, enter +500.
If you find you have 0.0006" of over-compensation, enter -600.
You'll get the hang of it!

[Rally]

BMC 20's are notorious for thrust bearing problems the 2 we had both had new installed on on several occasions.
Something to add!!
PS: bloke! Great Info! Glad to see people willing to help!!

[stevieboy]

All sorted now, cheers bloke.

On with the next job.

[faja]

Bloke, thanks for a lot of very useful information!
I have one question to you:

In case of backlash compensation in leadscrew map should we always compensate only minus direction map?

When axis calibrates in minus direction (initial movement to LS) I compensates MINUS map by POSITIVE backlash value.
When axis calibrates in plus direction I makes changes in PLUS map by NEGATIVE value of backlash.

Is it important to look on direction of axis calibration?

[bloke]

Add the compensation to the minus map for X & Y and add it to the plus map for Z.