I ve got a puma 300M, We got the machine in 2000, and we bought it new.
The c-axis has always had some kind of backlash while performing cutting in g112 cycle. It s like it over cuts, and then turns the c- axis back a bit to compansate.. Our older Moriseike lathes that are 15 years old performes much better. The company that sold the machine says it doesnt perform better, The personal from Fanuc thought it was strange but he couldnt find anything wrong with the servo and control. Does any one else have a simulair machine to compare with? I guess the roundness of a milled hole is about 0.3-0.2mm (0.008"). BUt it depends on where the hole is located on the part.

I've worked on Daewoo machines for the UK agent for 11 years before setting up on my own, never had that problem. They are good workhorses for production. One of my customers does a lot of C axis work very accurately. I suggest you go back to your Daewoo agent with the problem. One thought though, has the encoder belt been changed lately? The Daewoo's have a slightly different pitch to the standard belt that you can buy locally. Might be causing the encoder problems. If you want to send me your code for this part I may be able to test it on a Puma at a customers to see if I get the same effect. Post to www.alanbexton.co.uk or E.Mail alan@alanbexton.co.uk.

Alan B

There is nothing wrong with the code. It works correct on our older Mori Seiki machines, but they have a seperate motor to rotate c-axis. Our Puma controls the c-axis with the spindel motor. The problem accours at all c-axis milling in g112 mode. SOmetimes it s OK but most of the time it is not.. Higher feedrates and cutting on a bigger x-axis dimensions seems to work OK, but all parts aren/t designed like that..

You mis-understood me I wasn't suggesting the code was faulty, I wanted to run it on a Puma here to replicate the error so I could see it for myself. Then I may be able to offer a solution. Sweden is a bit to far just to pop in and look.
Alan B

You mis-understood me I wasn't suggesting the code was faulty, I wanted to run it on a Puma here to replicate the error so I could see it for myself. Then I may be able to offer a solution. Sweden is a bit to far just to pop in and look.
Alan B


The error accours on most c-axis milling contours, I am sure that they would know if their machine had the same problem.

I guess it could be some problem with the break(m88) that is supposed to reduce the backlash while milling. I have tried different braking force, but the problem still accours... I am starting to think that the design of the c-axis is poor.

The machine should perform better than you describe. Has your machine got a seperate encoder on the spindle? If not it is probably using a CZ sensor on the spindle itself. There would be 2 half casings between the spindle pulley and the headstock. This system is very accurate so if this is the system used you definitely have a machine fault somewhere in the eletronics and probably in the Fanuc side. If you have the encoder then that would allow a little error but still not the amount you describe. Check the encoder belt and plug for tightness. The brake system is possibly two pads on a disc on your machine. Check they are not binding on, and check they are not worn down. I can not think of any other areas to check at the moment but I will continue to think about the problem. I would be interested in the solution if / when you find it.
Alan B

The resolution on the c-axis is 0.01 deg. In c-axis mode without any clamp it is possible to move the c-axis a little bit by hand, this seems to be the same distance as the cuck moves during milling. I could messure it if it would be of any use. I have tried different clamping force on the brake before, without any better results. I will check the belts tomorrow, mabe the problem is there.. I dont think there are a electronic problem in the Fanuc side, they have been here and checked the machine.

It is strange that there doesnt appears a position error alarm when the machine performes milling, is it possible that the problem known to the machine builders and that they have change the parameter ?

The whole poblem is really sad, the machine and toolholders for milling is really rigid, I even got special designed face milling units from Sandvik with capto built in, and then the machine really cant use the milling capabilities at 100% ...

Today I messured the backlash during c-axis mode.. Messuring at a dia of 260mm the indikator moves about 1mm, this is done by hand without the M88 clamp. I ve inspected the belt drive, and one of the belts was not in good condition, but I dont think one out of 8 belts should cos this problem. I ve talked to the mechanic at work and we are going to change to new belts and try to tighten the belt drive a bit, he thougth that the belt tension were OK, but what does he know???

Sorry I've not been on the forum lately. The belts do not want to be too tight or the rear bearing will fail. Do you have an encoder on the headstock near the belts? If not is there a casing between the pulley and headstock with 2 cables going in? The only other thing is possibly the positional window parameter is to big it should be about 20. I think you should go back to your agent and get to try your job on one of their machines to see if it the same. They should be quite happy to do this and at least you will know wether it is just your machine or all of them.
Alan B