Originally Posted by cnczoner  Good ideas. I just got back home, so I'll check some of these tomorrow.
The quill definitely has some play (backlash), but I haven't measured it. The X-axis has 0.004" backlash, and the Y-axis has zero.
I'm using collets, but I don't think the tool is pulling out -- I usually set it so the end of the flute is just at the edge of the collet and when taking it out it later I don't remember seeing anything odd. I would've probably noticed. I will try to reproduce the chatter to see. When it was pulling into the material, I was slotting 6061-T6 with a 0.5"-dia 2-flute HSS endmill, 1500 rpm, 0.1" DOC, and it pulled down another 0.1" or more.
I've never used the fine feed (this mill was converted when I got it, and I've never used a mill before this), so I'm not sure how it works. But I'm looking at the crappy multiple-time photocopied exploded parts diagram now and I can see the many gears. I do plan to go into the head sometime soon, as I want to see if the bearings are easily replaceable with higher-speed types, as I want to double the motor speed. Perhaps I should look into changing the way the quill is CNC'ed.
I'll be back when I run the tests.
Thanks,
-Neil. |
1500 RPM is mightly slow for a 1/2" cutter in 6061. You should be more like twice that.
If your Z drive is going through gears, you've got a problem. When you plunge, you'll be taking up any backlash in one direction. Once you start cutting, the tool will be trying to pull the tool down further, which will take up the backlash in the opposite direction, so you'll have a very hard time getting Z accuracy and repeatability to any better then the amount of backlash. 0.1" of slop in a quill drive is a lot, but certainly not unheard of - it was simply never intended to be used that way.
Also, be sure you're not putting the endmill *too* deep into the collet - that can reduce clamping force, depending on the fit.
Regards,
Ray L.