Thread: Help on Z

Hi Bob,

Just have a read of your code. You can see the z depth it's telling the mill to cut. If it's all below the surface of your wood / plastic then I'd say you're losing steps.

Brad.

Hi Brad
I was wondering whether the problem may have been accumulated error, because there are so many little ups and downs. In the range of up 0.0002 and then down 0.0004 etc. The machine itself cannot possibly be that accurate so I thought that it was a possibility.

Bob

Hi Bob,
I looked at the code up to 5,000 lines and it looks sound. Feed rate of 900 and plunge of 300 is conservative if that is mm/min so that shouldn't be a problem. It must be coming from your rapids but that doesn't make sense as the rapid on Z is upwards or positve so if it is missing steps then the cutter would get lower not higher.

Do you have a pulley or helical coupling slipping on that axis? Is everything else tight on that axis? Check back to Y axis bearings.

Originally Posted by rodm

Do you have a pulley or helical coupling slipping on that axis?

Another option bob and could be it possibly I think I have a spare rolling around will check tonight.

Those are all good thoughts and I have had a couple more.
It could be that the chuck is not holding the bit securely or even that the spindle is not being held in place by the clamps.
The clamps idea is very easy to check, just needs a mark on the spindle casing and another mark on the clamps. run the program and see if they move.
With a slipping chuck, I would need to measure the distance from the tool tip to a point on the spindle face. Wait for the thing to go wrong and then measure again.

Rod's idea of the slipping coupling is a bit harder to check. I had previously thought of this and I checked that all the grub screws were tight on the Z axis. They were (are). The coupling nearest to the motor is easily checked. I put a mark on the top of the coupling aligned with the flat on the motor shaft. They didn't move. The coupling to the shaft is a lot harder to check as little of it is visible. ... Just had a thought though, a mark on the shaft of the tool bit aligned with the previous mark on the top coupling should check the tool bit, the coupling to the motor and also the collet's holding power all at the same time. At least, it would narrow the search area down if the problem was in that area.

Rod, your instruction to check back to the Y axis bearing is a little difficult for me to understand. How can the Y axis affect the Z axis. Well, I know it affects it laterally of course but I don't understand how it could affect spurious positive Z axis height increases unless you think the bearings have come completely adrift from their attachment to the gantry sides?

Sean also suggested that it could be because I didn't skim the surface of the table after aligning it. It is true that I didn't do that yet, because I didn't want to mess up the nice smooth surface of the table. However, the table top is not so far out that it would account for a 5mm space twixt the cutter and the table. I will surface the table top but it is already quite accurate.

Bob

Hi Bob,
Arrr the English language - I said back to and not the Y axis itself. In other words all the Z axis connections.

You know your machine better than any of us and all we can do is imagine what your problem is. Quite often an idea thrown up sparks you to think laterally.

To help understand the problem if the cutter is getting higher then what is happening is occuring on the downward or negative travel of Z axis. As you have conservative plunge rates 300mm/min (??) and it is soft material it is less likely that it is missed steps and more likely mechanical. If it is missed steps then you have bigger problems such a poor signal quality or faulty BOB or grossly under rated stepper motors. I don't think this is the case but it should not be elimated if all else fails.

So my mind says mechanical and first test would be to grab the Z axis firmly and lift. push. pull to see if there is any obvious movement. It is a new machine so there is a slight chance you didn't tighten something after alligning it. Then check the drive chain on the Z axis to eliminate any slippage in the drive chain.

Hi Bob,
Me again.

If you want to check for missed steps then write some code (example below) and run it from Mach3 with your spindle air cutting. Mark the position of the Z axis or use a gauge block for the starting height to compare it after running the code.
G0z-50
G0z0
G0z-50
G0z0
and repeat the sequence 50 times more.
Note it is a rapid movement so it should show up any problem.

and me again ......
If you have a dual shaft motor then wack a pulley on the back shaft and park the spindle nose on a block of wood on your table. Try to turn the pulley and you will feel if there is any slippage in the drive chain.

Thanks for all that Rod
The only reason I am writing now is to castigate everybody for not noticing the absolutely ridiculous thing that I said in my last message.

IE Just had a thought though, a mark on the shaft of the tool bit aligned with the previous mark on the top coupling should check the tool bit, the coupling to the motor and also the collet's holding power all at the same time.

Just think about how really stupid that was and now think to yourselves ... "Oh dear. How did I miss something so blatantly stupid?"

Yeah it really needs a mark on the font fence and left rear wheel of your car for that idea to work.

We didn't miss it we all just thought hmmmm Bobs having a moment

Originally Posted by rodm

Yeah it really needs a mark on the font fence and left rear wheel of your car for that idea to work.

At least the front fence will stay still while I measure it.

Originally Posted by WillyInBris

We didn't miss it we all just thought hmmmm Bobs having a moment

A moment? My moments seem to last for years.

Ok, I have tried some of the suggestions:

1) Grab it and shake the crap out of it to see if it moves. No movement visible at all

2) Block the spindle and turn. Nothing

3) Coupling slipping. Not that an ascertain

4) Mark tool bit and coupling. Hit self on head for stupid idea.

5) Write and test short program. Did so and got gained steps.

6) Decided to redo the motor steps on the z axis and it seemed to be about 0.3mm out.

7) After resetting this and rerunning the program it appears to be fixed.

I don't understand. Even if the number of steps was out by 25 mm, why would that cause gained steps?

I will rerun the test program a few times but make it about 10,000 lines

Bob

At the end of about two and a half hours the Z axis had gained about 63 mm.

This is not quite as bad as before given that the previous gain of 5 mm was gained over a period of about 5 minutes.