I have looked through the other posts and cannot find anything on this ... but is there a way to synchronize slaved axises?

For example, on my machine it takes 3052 pulses for the x axis to travel 1".  It takes 3054 pulses for the slaved A axis to travel the same distance.  Not a big difference, however with the axises slaved together, at the end of a 96" travel, the A axis lags behind the X axis by about 1/16" - which is a problem.

I have changed the "steps per" settings on the A axis, but it has no effect on the travel distance.  It appears that if the axises are slaved, then both axises are sent exactly the same number of pulses (ie what is under the "X" axis setting)

Although I am sure this is a mechanical issue - one pinion gear must be slightly closer to the gear rack than the other - I am sure it would be much easier to fix using software, if that is possible.

Help! 

That will have to fixed with the mechanics. If the steps per are off, the difference will be accumulative. If everything is the same for each, they will have the same amount of steps per.

Brett

Brett,

I suspect that what is/has happened is that one pinion gear is slightly larger than the other - or is meshing into the gear rack at a slightly different depth.  From what I can calculate the difference is about .001"

I was hoping that this could be adjusted in software, but apparentlly it cannot.

Well, it could be fixed in the software to a degree. But, they will both have to be perfect. All the steps per is is how many pulses Mach will have to send to move one unit. It doesn't matter if it takes 2 or 200,000 steps. But, certain numbers will not divide evenly into the motor or encoder steps leaving slight differences.  If you gear diameters are the same and the racks are the same pitch I'm guessing you have more backlash on one side than the other.

Brett

The gear rack and pinion gears are "identical" - at least they are suppose to be.

What I find when I adjust the "steps per" is that ONLY the X axis can be adjusted.  When I modify the X axis' steps per, the amount of travel does adjust.  When I modify the A axis' steps per, nothing happens.  The field changes - but there is no difference in the length of travel.  I have modified it as much as 10 steps (which on my system translates to about 1/4" over 96" - but the actual position has not changed.

I have been working on the mechanical side of the machine.

Changing the pinion gear has had no effect.
Changing the gear motors has had no effect.

I have "repeatability" - ie I can move from 0 to 96 ten times - and I always end up at exactly the same stop.  So, I do not believe that I am "losing steps".

What I am seeing is that sending 292,800 pulses to the X axis moves it 96".  The same number of pulses to the A axis moves it only 95-29/32".


Any other suggestions?

Joe, I have two questions: You say the 'racks' and 'pinions' are identical - are the racks made up of segments? And if everything is equal, how did you arrive at a difference of 2 pulses to travel the 1 inch between X and A?

You can probably see where I'm headed - I suspect that there is a mechanical difference between the two. The rack segments are positioned to allow one rack to be short (or long). The depth of engagement of the pinion in the rack will not make any difference from the drive side of the rack teeth - if the racks are identical in length no matter how long they are. Lathe carriage racks, as you probably know are pinned after careful measurement and most I have ever seen had varying gaps between long segments.

Could that be it?

Bill C.

I am doing some more measuring and experimenting.  I am working on measuring the X and A axis right next to the gear rack.  A quick measurement give me a reasonably accurate reading - ie the same amount of movement on the A & X axises when asked to move 96". 

My machine has a rather long Z axis (40" total - with a travel of 0-26")  What I may be seeing is a very small variance in the flattness of the X axis - that is being amplified on the length of the Z axis.

I will do some more measuring & testing to see what I can find out.

Try measuring the rack with dowel pins sized to fit between the teeth down to the pitch line. Measure a continuous piece then measure across any gaps. You know....

Bill C.

Also, racks can be a problem. They must be identical and best if from the same manufacturer. Very small deviations in manufacture can be disastrous in that type of gearing.

Each axis consists of 2 pieces - each about 71" long.  The racks are mounted to a steel bar which in turn is mounted to the CNC frame.  I removed the mounted racks and nested them together.  They align perfectly.  So, I don't think that the racks are a probelm.  The gantry currently rides on 8 guide wheels - 4 on each side.  I am going to remove the two center wheels and see if that affects how it rides.   I am wondering if the extra wheels may be affecting the ride.  Hopefully it will make a difference.