PCB Milling for SMD components - noob questions

[phil123456]

Hello everyone,

[ I am Phil, I do electronics, 3D printing (and Angular programming for a living) ]

I recently became interested in making PCB with cnc, so I found a couple of cheap(200$-200$) CNC on aliexpress but I am not sure they are precise enough (recent SMD components have quiet fine pads)

after digging a while, I noticed that there should be an empirical ratio between the speed, the feed(?) and the size of the "shizel"

would I need to buy an expensive machine or is it feasible with these arduino based diy kits ? what modifications should I do on the machine ?

thanks

Phil

Similar Threads:

• Noob CNC Questions
• Questions from Noob
• Noob with 10M Questions
• NOOB Questions

[CitizenOfDreams]

It is not easy or straightforward, but yes, you can mill PCBs with a cheap Chinese machine. Here is an example of what I can do with my plastic CNC mill (a Zen Toolworks clone).

The 8-pin chip is MSOP (0.65mm pitch).

[phil123456]

do you need to modify the cnc ?

[CitizenOfDreams]

No modifications to the CNC, except for a better spindle. Here is the spindle I use:
https://www.ebay.com/itm/400W-CNC-Ma...item58f5ae0aad

[CitizenOfDreams]

One essential piece of the puzzle is a software called Autoleveller (Autoleveller | CNC workpiece levelling). The name is pretty self-explanatory, it levels the G-code to ensure uniform cutting depth.

[phil123456]

169$ it's almost the price of the cnc I want to buy...why that spindle ? do I need such an expensive one ?

[CitizenOfDreams]

I did not buy that spindle specifically for PCB work. You can get by with any spindle as long as it has low runout and low vibrations. A similar looking DC spindle would cost much less money.

[phil123456]

hard to tell if it has low vibration from the picture, I guess it's linked to high RPM, I mean the fastest it turns, the more equilibrium

low runout, you mean low current ? cos 400w does not seem low here :-)

I think I saw 2418 and 3030 on aliexpress, that were quiet affordable and used by many, do you think they would do ?

[CitizenOfDreams]

Runout is radial inaccuracy of the spindle/chuck/collet, meaning that the tool is rotating eccentrically. Too much runout makes a spindle unusable for any kind of precision work such as PCB milling.

Vibrations are caused by poor balancing of the spindle. Spinning an unbalanced spindle faster will only make vibrations worse.

A Chinese router is probably a good place to start. If other people can make PCBs with them, so can you.

[phil123456]

ok thanks,

I guess I'll order a 2418 and see what I can do , reprint maybe a better header or even order a separate spindle

[phil123456]

One last question,

how did you manage to attach teh metal clamp of the new spindle motor to the cnc ? I was thinking about printing some plastic adapter...

[CitizenOfDreams]

I made polycarbonate plates similar to the original Zen Toolworks motor mount. The first part goes to the middle of the motor. The second and the third parts are glued together and go to the bottom of the motor. Sorry, these are the best pics I have at the moment.

[WhiteHare]

I have a 2418 that I've used to mill PCB's, and I just today did the spindel upgrade:
https://www.thingiverse.com/make:490336

I actually tried two different spindels: the 54mm brushless one and a cheaper 52mm brush spindel. Strangely, as you can tell from the photos there, I seemed to get better results with the 52mm brush spindel (the two benchmarks on the right in the photo below):.

Anything I can do to improve the trace resolution further?

[CitizenOfDreams]

There are several things to check.
1. Spindle/collet runout - do you have a dial indicator you could measure it with?
2. Spindle balance - can you feel any vibration when it's spinning?
3. Broken tool - with a sharp V-tool it is easy to miss a broken tip.
3. Cutting depth - going too deep will make the cut wider than expected.

[WhiteHare]

There are several things to check.
1. Spindle/collet runout - do you have a dial indicator you could measure it with?
2. Spindle balance - can you feel any vibration when it's spinning?
3. Broken tool - with a sharp V-tool it is easy to miss a broken tip.
3. Cutting depth - going too deep will make the cut wider than expected.

I'll dig my dial indicator out of storage. What should be the target spec?

Of the 4 you mentioned though, my hunch is that #2, spindle balance, is the dominant factor because I do feel vibration. I tried reducing it by picking a speed where it is less, and that did have visible effect, though I'm unsure whether it's better or worse:

Should I try to balance the spindle by putting on bits of tape, like some people do with the high speed motors on their quadcopters? Or perhaps I need to buy a more balanced bit in the first place?

[CitizenOfDreams]

I personally would call 0.01mm runout "excellent" and 0.02mm "good enough" for this setup. But I'm not a machinist, so anyone is welcome to correct me.

Can't give you much advice about the balance. Does the spindle vibrate without the tool, the nut and the collet? What kind of RPMs are we talking about?

[hfjbuis]

You have to measure the runout of the collet holder Inside this holder.
Then measure the runout on the shaft of the bit As close to the collet as possible.
Now you know how good the holder is and how good the collets are.
Be sure to absolutely clean the collet, bit and holder.
On my mill and lathe a runout of 0.02 measured on the tool would be great!
Huub

[WhiteHare]

UPDATE: Upon closer inspection, I noticed there's a lot of looseness between the z-axis smooth rods and the bearings that slide up and down them. I just now ordered new rods to see if I can get a tighter fit, and, see whether it helps reduce the vibration. The new rods should arrive in a couple of days. I'll see if installing new chromed, hardened steel rods (instead of the aluminum rods that came with the 2418 kit) helps mitigate the vibration before resuming on the very good suggestions put forth on this thread.

[WhiteHare]

Got the new steel rods. They measure 7.96mm diameter, whereas the old aluminum rods were 7.92mm. That 0.04mm difference may not sound like much, but I can already tell that the new setup is much less loosey-goosey.

[WhiteHare]

Yup, the new rods eliminated most of the vibration.