Unexplained asymmetry in milled part<resolved, Head tram is way off>

[suntravel]

For me this sounds like the problem is in front of the machine...

Do you have zero backlash preloaded ball screws on X and Y axis?

In your report no compensation is dialed in.

Easy to get off center with worn out ACME screws on the axis....

Uwe

[apm]

when using "jog to home" postion.

power up

1.) jog the machine to the home position. for a mill Typically max negative in X, max positive for Y and max positive for Z
review axis naming and direction conventions here https://centroidcncforum.com/viewtopic. ... 437#p28437
to make sure we are all on the same page.

2.) then use the part setup menu to define the part zero position (g54). which matches the fixture to the intercon program.

Isn't that what I did here where you can see I am 0,0,0 at g54?

For me this sounds like the problem is in front of the machine...

Do you have zero backlash preloaded ball screws on X and Y axis?

In your report no compensation is dialed in.

Easy to get off center with worn out ACME screws on the axis....

Uwe

Its an old prototrak machine I am pretty sure they used zero backlash ball screws on there. I know they are indeed ball screws. I don't really think I have a backlash issue either as when I do repeating depth cuts I don't see any evidence of steps emerging which would suggest with time I am walking off in a direction. This part in particular I was working my way down step by step with the depth repeat 0.050" at a time to a final depth of 3/4".

I think my next plan is to run the machine tonight back over the part, generate and post another report. If that fails I have a stupidly simple test part that I plan to use. Once again I will set 0,0,0 on maybe G55 and put the G55 in the top of the intercon program, drill the two holes as though they were part of my fixture, and then just load a program that runs a contour around it and see how this part lines up.

I would like to think that I must be missing something stupidly simple here, but I am really feeling pretty stuck as it all seems on the surface to be correct.

[apm]

Well I admit I made an embarrassing mistake. After convincing myself I was programming everything exactly correctly there was only one last spot that the problem could possibly be in. I checked the tram on the mill and the head was way off.

Of course I've just been doing a single toolpath to cut my way around so it's nowhere near as obvious is it typically would be if I were making multiple passes on the part. It actually would be handy to figure out how to better program this up in the future. In a perfect world I think going rather deep it might be nice to do a step over cut just so that there's a little clearance for the end mill had that have been done I would have certainly seen the tram error.

Oh well such are the joys of learning something new. Sometimes your lack of confidence in one area which makes you less likely to check the other spot. Making matters worse it's been 20 years since I've run a Bridgeport. Almost all of my work in the last 20 years has been on a Cincinnati Mill that never requires one to look at the tram. I forgot just how flimsy a Bridgeport is. Hopefully I didn't waste too much of anyone's time on this issue. Thanks for the help

[suntravel]

Thanks for reporting back and I am glad you figured it out.

There is no geometrical perfect machine and also no perfect ball screws.

That is one reason Centroid build backlash and screw compensation in the CNC.

I suggest after setting up tram and play of the gibs (if the machine is with gibs) to dial both in.

Uwe