Engraving /Micromachining tool holding advice needed.

[Flynth]

I recently acquired a desktop cnc machine and I was both amazed by its positioning accuracy and disappointed with its spindle runout. I'm interested in using it for micromachining of soft materials (brass, aluminium, copper on pcb's) and engraving. Unfortunately the "spindle" is simply a dc motor with a 5mm output shaft and a 5mm to er11 coupler. By careful manipulation it is possible to lower few thou of runout to few tenths (0.015mm) in the er taper, but then the collet settles into up to few thou of runout after few minutes of work.

So I decided to do some experimentation. First I made a tiny boring bar and a holder for it (pictured)


Then I turned the OD of 9mm brass stock to 8mm and polished it to make holding it repeatable. I ended up with a taper under 4 tenths (0.01mm). I then used a Morse 3 collet (runout around one tenth) in my mini lathe and bored very precisely 5mm in one end and 1/8th (3.175) in the other for a fairly tight (manual) push fit.

You can imagine my surprise when I measured 4 tenths(0.1mm) of runout. Still not as good as planned.

So instead I bored out the 5mm hole and the screw to hold the shaft without preparing the OD and I chucked the whole motor with my "adapter" installed into my big lathe chuck. Then I used a bench psu to power the motor and bored the hole in tiny increments like that. (I attempted boring directly on the cnc before, but it wasn't rigid enough).

II ended up with one good one roughly acceptable holders. The 1/8th has under a tenth of runout (0.01mm) 10mm from the holder and 4mm holder has about 3 times that.


At last I have a usable 1/8th holder for pcb milling etc.


So I'll probably have to remake the 4mm holder.

I'm wondering if there is a way to make one without having to use the motor to do the turning. I imagine my initial failure was because the 5mm hole isn't adequately concentric. Then if I use the spindle itself for boring that compensates for it (the screw ensuring same position each time). Perhaps using an expandable lap and lapping it would be the solution? What do you think?

I can make a tiny expandable lap from aluminium, but I thought to ask first if anyone can comment from experience on what precision I can expect from a two part expandable lap.

I'have some ideas for how to make such a tiny lap, but all involve custom d bits etc. So before I spend days on it I decided to ask here if anyone has any advice to offer or links to any content about such tiny diy laps etc. Buying one unfortunately is uneconomical, because the only manufacturer I'm aware of is in US, and shipping alone would make it not worth it. However if there is a manufacturer of such tiny laps in Eu I would also be very keen to find that out.

Also, I'm considering making a complete miniature spindle with belt drive RC motor for that cnc. This is a different project however and I have to wait for parts to arrive from China before I can start. So I'm trying to improve what I have first.

On a side note, it is amazing that shipping small parts from China can be so cheap while equivalent distance the other way - US is at least two orders of magnitude more expensive. 8

 

[Steve-F]

I might have a small spindle, ....will have to look.

 

[Lo-Fi]

You're unlikely to get great results with the bearings in that cheap DC motor, which is probably the limiting factor. If you want something accurate and smooth running, it's not hard to construct a tiny spindle with a couple of magneto bearings and belt drive. Or really go for it and run a straight shank ER collet holder in graphite air bearings. You can buy the holder commercially and blocks of graphite are cheap and easy enough to get hold of.

For what it's worth: The Chinese gov essentially subsidizes export shipping, making Chinese stuff even more competitive...

 

[Flynth]

You're unlikely to get great results with the bearings in that cheap DC motor, which is probably the limiting factor. If you want something accurate and smooth running, it's not hard to construct a tiny spindle with a couple of magneto bearings and belt drive. Or really go for it and run a straight shank ER collet holder in graphite air bearings. You can buy the holder commercially and blocks of graphite are cheap and easy enough to get hold of.

Building one is in my plans. I already sourced standard roller bearings that appear to be good enough, but I have to wait probably about a month for the straight shank ER adapter. So what I'm doing with this motor is just temporary. Still, I'm trying to make the best of it.

I might have a small spindle, ....will have to look.

Thanks. I'm definitely interested how it is constructed. If you can take some photos and perhaps measure runout that would be great.

I see belt-drive mini spindles occasionally in online videos, but I haven't seen one where runout numbers, max speed etc would be mentioned as well as construction details.

 

[Lo-Fi]

It's a cool project!

The trick with home build spindles seems to be to machine the critical parts to final dimension in their own bearings, then treat is as a non serviceable cartridge that cannot be disassembled. I'm building a Quorn tool and cutter grinder which uses a couple of magneto bearings and gets finish machined after final assembly. Tricky with a commercial ER holder, but if your bearings are decent you should get good results.

 

[Braeden P]

i have a little 30 18 mini ¨mill¨ to much flex going to upgrade the whole entire spindle with new rails and bearings

 

[Steve-F]

Flynth, found it......here is picture of what I have. Adjustable depth control with a 3/16" collet and Excellent bearings!

 

[macardoso]

Get that from Steve!

You need a proper spindle - runout kills tiny tools.

 

[Braeden P]

Get that from Steve!

You need a proper spindle - runout kills tiny tools.

yup alum eats the vee bits alive i might know from practice

 

[Flynth]

Flynth, found it......here is picture of what I have. Adjustable depth control with a 3/16" collet and Excellent bearings!

This is nice. If we weren't located half a world away I would be interested in buying it off you (if you didn't need it)

As an example what difference a thou of runout makes I took two microscope pictures of test pcbs milled. Those were done with 10 degree 0.1mm v bit (2.5thou ),10k rpm, 130mm/min (5.1in/min) feed, 0.04 doc (1 thou) 2 passes so total depth 0.08mm (2 thou). I estimate the width of the cut is 0.11mm (around 2.5 thou) on the first picture and 0.15mm (3.5 thou) on the second. It is not a huge difference, but I think it will affect the tool life quite a bit.





BTW, this on on 3018cnc "Pro" version using the same g code. The only difference is the height map. As the pcb is not flat and it is glued to a piece of mdf it wouldn't be possible without a height map.

i have a little 30 18 mini ¨mill¨ to much flex going to upgrade the whole entire spindle with new rails and bearings

I always considered 3018 an engraver rather than a "mill" (even in quote marks ;-) I assume we have the same device. I have the so called Pro version that uses thick bakelite as y supports rather than aluminium extrusions. I think they are OK. They feel very stiff in hand. I'm not expecting miracles from it, but with some upgrades it should be possible to cut aluminium with tiny cutters at high speed and shallow doc. My main issue with it so far was the runout and the flexibility of the spindle mount. Imbalance caused by motor's shaft cutout is enough to cause vibration in the mount... I think it is because they saved money by designing large empty cavities in the plastic. If it was a solid block of plastic it would be much better. I'm considering 3d printing a new z mount from abs with close to 100% infill. That should be much more solid. Of course aluminium would be better.

Here is a photo of the above pcb being cut.


Coming back to the original subject

It's a cool project!

The trick with home build spindles seems to be to machine the critical parts to final dimension in their own bearings, then treat is as a non serviceable cartridge that cannot be disassembled. I'm building a Quorn tool and cutter grinder which uses a couple of magneto bearings and gets finish machined after final assembly. Tricky with a commercial ER holder, but if your bearings are decent you should get good results.

Yes, definitely. With the ER holder I hope it will fit the bearings nicely and I'll be able to "dust grind" the taper to make it true once it is assembled.

Having to wait for that adapter a long time I may end up making my own equivalent before it arrives...

Also, why magneto bearings? Did you just happen to have them, or are they in some way better for a spindle?