ER, Collets and tool accuracy

[booski]

Now, I'm not a machinist by any means, I've done basic lathe turning and milling but that's about it.

With regards to CNC machinery, I've had to learn on the fly.
Something that's always seemed to have bugged me or plagued me is inaccurate (or so I thought) tooling. It's been a massive headache where you've put a tool in the spindle, make a cut only to find that the tool isn't concentric and theres massive amounts of run out.
I've spent hours and hours pouring over my machine trying to figure out what the hell was going on. I've gotten through many, many, MANY engraving V bits thinking that they were naff and cheap and poorly ground.

I was nearly ready to pull my hair out until recently whilst doing my most recent project I discovered something. I knew that ER systems need to be tight but without a torque wrench, I had to go by feel so I just did it up to what I believed to be 'tight enough' because I was worrying about the collet twisting in the spindle.

As it turns out, I wasn't doing it tight enough! My most recent project as it's been developing has had many cuts made for a differing variation of the same circuit on the same board, this mean a lot of tool changes. The boards, for once, have been coming out very consistent in that they actually look bloody good. So what you may be thinking have I changed for this to happen?

How tight the ER nut is done up, that's all. All this massive headache that I've been suffering for years now, actually, has all been down to how tight a nut is. With all of my most recent cuts, the nut has been done up as tight as reasonably possible with 2 short spanners but the accuracy achieve has been incredible. I never thought that it would make that great of a difference because after all, it's one precision ground surface against another precision ground surface, what could go wrong?!

[daedelus]

Hey, great stuff. I love it when a plan comes together.

In the video you posted before I noticed a lot of scratches consistent with sandpapering the burrs (I know because I have a few boards which have similar scratches, as well as a few without).

My question is... did you have a lot of burring before you fixed the nut problem, and do you have many burrs now?

In other words, was the run out the cause of the burrs or is it dull tool bits?

[booski]

In the video, the first couple of attempts on the left side of the PCB where made using a 0.1mm pyramid engraving bit rather than a V bit. This is because the pyramid bits are stronger at 0.1mm than V bits but because there's no real cutting edge, they do tend to burr quite a bit. That's easily sorted with a little scouring pad, I don't like sand paper because you're removing material off an already thin layer.

If you look at the end of the video where I am using a V engraving bit and where the picture is better, you can see that the cut is quite clean.

I do get a little burring, it's hard to avoid when your trying to remove such a thin layer of a foil like material but it doesn't help that the spindle isn't 100% perpendicular to the machine axis but that can't be helped to much with a machine made out of plastic!