Cartesian Coordinates Accuracy + Gears Question

C

CarlosA

H-M Supporter - Gold Member
H-M Supporter Gold Member
Joined
Nov 12, 2015
Messages
160
For some reason, every time I lay out a bolt pattern on my small import mill at least one set of holes is off by .008". The holes adjacent to them are off by .002" and measuring the remaining 4 possibilities on a 6 hole pattern are spot on.

I am building a spur gear differential for a robot project, are these minor issues going to adversely affect the spur gear mesh?

I did 3 test patterns in 6061 being very careful to always turn the hand wheels clockwise (passing the mark then zeroing in on it to relieve backlash) - spot drilled for center, drilled undersize in one shot, then reamed to a precision .250" hole.

Any suggestions or feedback on my method, or should I just run it?

More details for anyone interested:

12 tooth 24 pitch pinion for each axle
12 tooth idler gear that basically connects the two sides
.50" centers on everything
6 holes, 4 of which will carry idlers, the remaining two just for holding the structure together
.250" rods for the idlers to ride on
idlers made of delrin
pinions machined into the axle shafts in 6061
side plates will have a ledge that rides in some larger bearings
the axles/pinions will ride on a thin delrin sleeve inside the side plates
one side plate will have 60 external gear teeth to mesh with the motor pinion (this is why i`m sweating it, lots of gear teeth to cut into it after doing the major work above)

Edited to add: I was thinking since i`m only using 4 of the 6 holes for idler gears I could pick the best four and eliminate a lot of the error. But still, i`d like to know what i`m doing wrong!
 
Last edited:
Your method sounds correct. Maybe a lead screw or hand wheel dial issue? Maybe try locking the table locks while drilling.

Yes it might affect the mesh if the gear runs out, but I would think that the shaft should align the gear on center and the screws should just be holding it in place.
 
Couple of observations. How are you establishing your error in positioning? Do you have a CMM?

On your method of backlash comp, I have always done basically the opposite. I have no "slack" while dialing to the new position, and stop when I get there. Only if I overshoot it do I back up enough to clear the lash and approach in a consistent direction again.

If you're feeling good, you can work out a small prick punch on the error side of the pin hole that will at least for a little while "move" the centerline of that hole.
 
Tony Wells said:
On your method of backlash comp, I have always done basically the opposite. I have no "slack" while dialing to the new position, and stop when I get there. Only if I overshoot it do I back up enough to clear the lash and approach in a consistent direction again.
Click to expand...

I always do that too, mostly because that's how I tune my guitars.
 
CarlosA said:
I did 3 test patterns in 6061 being very careful to always turn the hand wheels clockwise (passing the mark then zeroing in on it to relieve backlash) - spot drilled for center, drilled undersize in one shot, then reamed to a precision .250" hole.
Click to expand...

Carlos,

Not to gang up on you here, but I am with Tony and Charles. To manage backlash I always turn one direction. If I overshoot I cannot just backup a little, I need to back way up and then approach again....at least on my sloppy mill.

Perhaps that's what you meant?

-brino
 
Agreed.
Backlash management.

Daryl
MN
 
Thanks for all the responses. I will try always turning one way on another test piece. But here is why I do that ... when laying out a bolt pattern I start from zero - some of the coordinates are negative from zero, some are positive. So when going from zero to hole 1 I don't account for backlash - when going back to zero from 1 I should adjust for backlash, right? Then when going to hole 2 in the "positive" direction from zero, again account for backlash?
 
Carlos, no matter on the positive or negative numbers.....always be consistent on the direction of approach. That way the only true variation you will see is lead screw error, but not backlash. That's a different issue.
 
You must log in or register to reply here.
Back
Top