Rent Mill & Lathe DVDs at Smartflix | Great aluminum & steel prices at OnlineMetals
BACKLASH AND LEAD ACCURACY
What is backlash
and why does it matter? What about lead accuracy? How does it affect me
and my machine? Or in other words... "In search of 'perfection'"
Also: Ballscrew and ACME Leadscrew Accuracy
Charts
- PREFACE
- INTRODUCTION
- BACKLASH IN A MANUAL WORLD
- CNC TO
THE RESCUE?
- EXPECTATION
- A
PERFECT CIRCLE
- OTHER DESIGN CONSIDERATIONS AND THE
PROS AND CONS
- CONCLUSION
- RESOURCES
CNC TO THE RESCUE?
What about a CNC (or computer(ized)
numerical(ly) control(led)) machine? Aren't the backlash problems solved
by CNC?
Not really, and the end result gets worse!
And here's why:
To keep things simple, we won't discuss open
and closed loop systems here. Let's just assume that we have a computer
that tells the motor how to turn to move the table. There won't be any
feedback to the computer to make adjustments for error, etc.. or "backlash
compensation" techniques in the controller software of G-Code programming
- and this makes up our CNC machine:
1.
COMPUTER > MOTOR > TABLE/MACHINE
Now let's use an example that will better
show how backlash affects CNC systems.
2.
Computer says move 1" in the X direction and the > Motor turns the
corresponding turns to move the > Table in the X direction by 1"
Now before I throw in the term "lead
accuracy" or "lead error" let's do one more thing. Let's also make the
table mentioned above to move BACK to its original position (we'll call
this X0 - btw, we started at X0).
3.
Computer says move 1" in the X direction and the > Motor turns the
corresponding number of rotations to move the > Table in the linear
X direction by 1" (this should bring us to X1)
then the
Computer says move -1" in the X direction and the > Motor turns the
corresponding number of rotations to move the > Table in the X
direction by -1" (this should bring us back to X0)
With examples 2 and 3 we can talk about BOTH
backlash and that term "lead accuracy" or "lead error".
So we've converted a machine to CNC with the
screws and nuts that came with it. We don't know anything about where it
came from, how accurate it is, etc.. and you run the machine. What
happens?
Sticking with the backlash example as shown
in 3, you would think that the table would go back to its original
position of X0 right? Nope. remember, there's backlash! When the screw was
reversed in order to go the opposite direction, it had to spin around
enough to positively engage the nut before the table would move. Assuming
that the screw is 100% accurate (i.e. no lead errors), the machine should
return back to some position in between X0 and X1 (hopefully you're closer
to X0 than X1 otherwise your machine probably does have serious issues!).
Let's say you fix this by getting an
"anti-backlash" nut that delivers 0 backlash and you can get back
to X0. Great!
But wait! There's more!
Now let's bring in "lead accuracy" aka "lead
error" into the picture.
You noticed something else strange. In
example 2 you told the table to move 1" to X1 but your trusty DRO tells
you otherwise. Let's say the actual move was to X0.996" (i.e.
you're just shy of 1" by four thousandths of an inch - 0.004").
You've just experienced "lead accuracy" aka
"lead error". This is the amount of allowable axial error of a given rated
screw whether they're ballscrews or leadscrews (we'll talk about this in a
bit as well).
Thinking that you can solve your problems,
you go out and buy any ballscrew/ballnut out of a catalog because you've
heard that "they're more accurate".
You retrofit your machine to accommodate the
new ballscrew and ballnut and realize that you still have backlash
AND errors in travel. Why? And imagine when to your amazement, you
experience positional errors that are even more obscure and unexpected
that what you had previously seen. In more complex moves, you basically
add up lead errors with backlash and you can get some interesting results
at different points..
|