Need hlep with backlash...please

[SpeedMill]

Was just getting ready to order the plans for a 4x4 Hybrid and convert it to a 4x8 with rack and pinion with Ahren's setup. I though that this was the best way to avoid backlash. I am trying to learn everything that I can before I build so I do not make any stupid mistakes.

My problem today is a discussion about backlash at this link. http://www.mechmate.com/forums/showt...5149#post35149. In this thread it seems that backlash does not matter. Can some of the experts here tell me if it does or does not???

Here on the Zone I thought I learned that,

1. Backlash will cause iregular parts
2. Backlash can make rough jittery cuts on the part
3. Backlash will change the total distance traveled with multiple changes in direction
4. Eliminating backlash should be a goal when designing the build

Are those assumtions false? Does backlash matter? Can someone point me to a thread or book or web page that clearly tells me what backlash does?

TIA

[CMJacked]

manual machinist deal with backlash on a move by move basis. Loading your axis is the only was to deal with it. Rack an pinion is a gear to gear mesh that will by design have backlash, loss of movement will have to be programed for, keep all your moves in the same direction you touched off from.

[jsheerin]

Backlash does matter. Your point number 3 is very debatable however which from a quick scan is what they were going on about on the mechmate forum. Basically, backlash will cause you to be out of position by up to the amount of backlash at any point.

R&P can be made without backlash (spring loading the pinion, using a split pinion, etc.), as can screw drives (anti backlash nuts, spring loaded nuts, etc.).

[CarveOne]

Originally Posted by jsheerin Backlash does matter. Your point number 3 is very debatable however which from a quick scan is what they were going on about on the mechmate forum. Basically, backlash will cause you to be out of position by up to the amount of backlash at any point. R&P can be made without backlash (spring loading the pinion, using a split pinion, etc.), as can screw drives (anti backlash nuts, spring loaded nuts, etc.).

jsheerin is correct in that a properly tensioned pinion gear keeps the pinion gear teeth in firm contact with the rack teeth at all times, at least until excessive load forces cause the tension to be overcome. Severe overloading can cause the teeth to slip over each other. The pinion gear teeth must not bottom out in the rack teeth valleys or there will not be full engagement of the teeth, leading to backlash problems no matter the amount of spring tension. This can happen when the pinion gear and rack becomes well worn.

As long as you are aware of how this system works you should be able to deal with backlash if it should ever be a problem.

CarveOne

[ger21]

Originally Posted by jsheerin Backlash does matter. Your point number 3 is very debatable however ........

It's not debatable, it's completely false. Which is what everyone on that MechMate thread said, but the OP didn't believe.

Ideally, you don't want any backlash. With a rack and pinion drive, spring loading the pinion into the rack should minimize it enough so that it's not an issue.

[jsheerin]

Originally Posted by ger21 It's not debatable, it's completely false. Which is what everyone on that MechMate thread said, but the OP didn't believe.

The only reason I said it was debatable is that backlash could change the total distance traveled in a given move, ie you don't really know exactly where you will end up. You could be anywhere within the backlash region in theory. So if you wanted to get really nit-picky you could make that claim, but the part about it accumulating is definitely not true, and that was really the core of their argument I believe (I didn't really read it, as the first post that popped up said they had banned someone in the thread, so I didn't see the point in wasting the time to read it all).

[CMJacked]

spring loading the pinion, using a split pinion

Could you direct me to a place I can get more information on that type of system? I'm currently running a Bett's 148" swing 96' between centers with two carriages under CNC control, The Monkey tail and X axis is ball screw but Z is rack and pinion. The Lathe will handle 200 tons between centers unsupported. I would be very interested in a system to help tighten up Z axis. I program for the backlash and really do not have a problem that cant be overcome by maintaining directional load in Z axis. When returning to a start position you have to move past and back to position to load the axis again. No big deal in this situation. But when using the Milling attachment and retracting out of blind holes sets it up for some dwell while the tool is in motion. I know this is a bit off topic but well, there it is..

[jsheerin]

Try here:
http://www.atlantadrives.com/zerobacklash.htm

[SpeedMill]

So basically 1,2 & 4 is correct. #3 really has me confused because if I thought that if I made a pocket & used the back and forth method (raster?) and I had 5 thous. backlash I would lose 5 thou every time I changed direction because the motors have to make up for it.

From what I am hearing here and what they are fighting about on Mechmate it sounds like all I need to do is add 5 thous. to the size of my part to cover the backlash. Am I right?

[ger21]

Originally Posted by SpeedMill From what I am hearing here and what they are fighting about on Mechmate it sounds like all I need to do is add 5 thous. to the size of my part to cover the backlash. Am I right?

No, they may already end up .005 larger, due to the backlash. You'd probably want to make the part .005 smaller.

The proper thing to do is remove the backlash. Next best thing is to use backlash compensation in the software, but it may not always work as you'd expect.

[SpeedMill]

No, they may already end up .005 larger, due to the backlash. You'd probably want to make the part .005 smaller.

So if we have 5 thous. BL then a part can be either 5 thous. larger or 5 thous. smaller for a total of 10 thous. but not more?

Not trying to be a SA just trying to get a grip on what BL causes. I DO believe that BL is bad but how bad is it really? Is too much time & money being spend on eliminating BL if it really does not matter that much?

[Gerald_D]

If you are cutting reindeer lawn ornaments, or kitchen cabinets, 0.005" really does not matter.