Offset problems (I think!)

[midy]

This video is hopefully going to stand out to somebody what I am doing wrong, as when I finish measuring all the tools and save everything. Then go into MDI and start changing tools - all tools except the reference tool are not where they should be in relation to the part/centre line. They are all off differing amounts - some more than others. Anywhere from 0.04mm to 0.7mm. I'm pretty sure its to do with offsets - I've got a setting wrong somewhere? I've tried manual setting as shown by centroid and get the same results. Have checked and double checked the machine setup and pretty confident the axis are moving correctly/accurately.

[suntravel]

You are sure your axis have 0 backlash and mm/rev is dialed in right?

Mounting the probe with the short 6mm shaft in a chuck is not optimal, I would clamp the shaft in a at least 20mm shaft collet holder.

Uwe

[midy]

Just got back from holiday, so sorry for delayed reply.

Will make a collet holder, thanks.

Yes, mm/rev are all good, followed marty's video for measuring and adjusting accurately. There is 0.014mm backlash on X axis. I have tried putting this figure into backlash comp, still getting the same wacky results with all the different tools. If I check all the tools which are giving incorrect numbers and then MDI back to the ref tool, it is spot on where it should be.

cheers

[Muzzer]

TBH, I'd be pleasantly surprised if you have only 14um backlash on the X axis, looking at the machine construction. If you approach a gauge block or ring gauge with a DTI from opposite directions, then subtract the gauge width and DTI probe tip width, you should get zero difference. That will likely be worse than if you reverse jog the axis and look for first movement. And of course, if you provide a load on the tool, that backlash will worsen.

Having said that, 0.7mm is terrible. If you aren't losing steps, there must be a fairly significant mechanical issue at play. Backlash comp can't make up for mechanical issues - these need to be bottomed out first.

Are you sure those Dickson toolholders seat consistently? Perhaps you could try repeating the same tool measurement having removed and replaced the same tool. Does the process of tightening the holder in itself cause the tool tip to move?

[midy]

Just to clarify, I have set every tool with the auto tool setter (also done manually as per Keith's method) I then go into MDI and go through every tool on a part and they are all out by varying amounts. The only one (which has been taken in and out of the Dickson tool holder) that is spot on is the ref tool, and it is within 0.01mm (backlash permitting)

The lathe itself is a harrison L6 with very little use and the ballscrews and nuts came from a harrison alpha cnc lathe and are of the highest quality, German manufacturer I believe. Again the doner lathe was almost new at the time.

I'm fairly confident in the mechanics and just think i'm over looking something in the software setup?

Also, I have wiped the hard drive and re installed only the necessary software for acorn.

[tblough]

When you MDI, how are you calling up the tool? Are you setting the tool and offset, I.e. T0103?

[midy]

I do it T0101, T0202, T0303, T0404 and so on. Is this what i'm doing wrong?

[suntravel]

There is no problem with offsets in CNC12.

If you are off after probing, it must be the probe or a mechanical issue.

Uwe

[Ken Rychlik]

I just ran this circle on a router, and even though you have a tool zero'd you have to set "part zero" for it to track the tools. Did you do that?

[midy]

I'm on a cnc lathe with this but yes, I make sure the ref tool is part zeroed. Which as I said before, can MDI in and out of and physically remove the ref tool and it is never wrong. It repeats, the others don't...