Workpiece "drifting" alongside X-axis?

[mister_zed]

Hello, everybody!

I have a recurring problem that just occurs sometimes and I do not know what to do about it.

When I route a line along Y axis it happens sometimes that the rubber belts start moving, albeit slowly. Always to the right. The "right" is when I face the keypad, so I believe it would be to the "left" when looking at the design in designer. It does not happen with every carving and when it happens it is usually 30-50 minutes into a job.

Anybody who has seen this kind of problems? What is the remedy to that? Is it bad sensors? I am totally lost with this one. :(

Regards /ZJ

[DocWheeler]

ZJ,

I'm not sure about what you are telling us.
Is it moving when it shouldn't? Routing where it shouldn't?
Or is it just that the speed changes?

[mister_zed]

Hello,

exactly as you said... moving when it should stand perfectly still... the only movement should be the head routing i line perpendicular to the length of the workpiece, along Y axis. Instead it moves along X axis as well, distorting the routed pattern.

[DocWheeler]

Thanks for the clarification, I have never had that problem.
Must be that 50 hz thing you have over there;).

[bergerud]

The first thing I think of (if it is not the 50 Hz thing) is that the brass roller O ring may be rubbing on the front rubber belt. Press the brass roller down with your finger and see if it hits the rubber belt.

[mister_zed]

Thanks, guys!

No, I am quite positive it is not the 50Hz "thing". :-) One machine has this issue, another one does not. Both run on the same 50 Hz. Adn then it does not happen all the time, just sometimes. I never know when.

The machine with issues with X axis moving when it should stand still has rubber belts, the machine with no issue has sandpaper belts.

So it might be the issue of rubber belt rubbing against the Oring. Although one would think that if the X axis should not move in perpendicular carvings then it should remain stopped no matter what. Well, who knows how the machine "thinks"? I know I had issues with the O-ring rubbing the belts before: the person who installed the rubber belts did not place the extra washer underneath the sensor screw and I got extremely "funny" measurement results for quite "normal" boards.

Right now I am toying with the idea of re-drilling the sensor holes in the black metal plate so I move the sensor a bit away from the rubber belts, maybe by 0.05 in or so...

I will look into this in a few days and let you know guys.

In the meantime, if anybody has had issues like mine bfore, please let me know how you fixed this.

[liquidguitars]

The machine with issues with X axis moving when it should stand still has rubber belts, the machine with no issue has sandpaper belts.

I seen this- what i do is cut 1/2" off the rubber belt on the brass roller side.

[DocWheeler]

I'm following what is being said, but I do not understand how the belt
could affect the brass roller when only the Y truck is in motion.
Is down-pressure of the bit bulge the rubber belt to move the sensor?

[liquidguitars]

In my case the belts moved backwards and could not calibrate the unit at all. Taking 1/2" off the belts seemed to do the trick.

I think the AI will control the feed rate and make corrections. when the belts rub it can freak out the machine.

[bergerud]

I'm following what is being said, but I do not understand how the belt
could affect the brass roller when only the Y truck is in motion.
Is down-pressure of the bit bulge the rubber belt to move the sensor?

I think it happens to him with a route operations. I think the x gears wiggle back and forth keeping the path correct. This is not the same as if it was carving and moving in increments and stopping.

[DocWheeler]

Then the belts should not be involved in the problem, right?

[bergerud]

The front belt could be if the belt is rubbing on the brass roller O ring. Is that what you mean?

[DocWheeler]

No, I'm still trying to get my head around the idea that the belts,
that would NOT be moving with a route in the Y direction, would
somehow move the brass roller.

Could a vibration, with low head pressure, cause an oscillation that
would turn the brass roller?

[bergerud]

I think that the x gears are always going back and forth dealing with the backlash for any vector curve, even if the cut looks like it is straight across. Also, it does not always happen. (I have my cover off of my x gears and get to watch the behavior.) I may be wrong.

[mister_zed]

I think it happens to him with a route operations. I think the x gears wiggle back and forth keeping the path correct. This is not the same as if it was carving and moving in increments and stopping.

Well, before routing the perpendicular line the machine must position itself. It was a series of perpendicular lines on same workpiece (2 sided carving with the backside already completed).
The first line was OK. The machine moved 0.4in along the X axis and carved a new line which was OK. And so on and so on...
But suddenly the distances increased and the lines were no longer perpendicular.

I happened to look at the machine when it was repositioning itself for a new line. First, a quick movement along X axis to new position, then the belt movement seemed to stop, but I could see slight and slow movement of the workpiece to the right (I was facing the keyboard, my right). And before plunging down with the carving bit the machine seemed to think unusually long 5-10 seconds, maybe trying to understand why the X belt was moving. :-) I really got the impression that the machine was trying to figure things out.

So for the moment I will stick to the "bulging belt" theory. Yes, I might have a low head pressure, I never did any recalibration on that point. Hmmm... but low pressure would contradict the creation of bulge on the belt.

[bergerud]

There is no "bulging" of the belt. The part of the belt that gets compressed is on top of the roller whereas the brass roller is on the side of the roller. The brass roller, however, gets pushed down by the board and that brings it closer to the side of the rubber belt. Check how close it comes. You could also, as a test, take off the O ring.

[fwharris]

The belts could still bulge with low head pressure so don't rule that out. Also make sure your dampering spring under the brass roller bracket is there and holding the brass roller up.

Do a measure board and see how it acts.

[DocWheeler]

My imagination leads me to think that at least two of three things are combining to cause this.
Loose bolts in the bottom of the left posts, low head pressure, and/or inadequate support on the output side.

Confusing factors are that it did not happen when the board was heavier (carving the first side).
And, that it doesn't seem to skip very much in the X direction.

If the board is bowed it might explain why it works on one side and fails on the other despite a weight difference.

Normally the problem is that the board moves and the brass-roller doesn't which causes a "skip" in the X direction.
This problem seems to be one where the brass-roller moves while the board itself didn't move and the computer moves the belts.

What I was imagining last evening was that the board would oscillate in both the X and Z axis since the rubber belts
might allow that making it roll the brass gear "backward" a little each time and causing the machine to move the belts
to address its supposed X movement.

If possible, check the bottom post-bolts, Check the head pressure, and maybe raise the outfeed roller.

Just my thoughts.