Frage Experimentation with pcb engraving.

I've been using my 420/2 for pcb engraving for three years now, so I figured it was time to check backlash settings. Pleasantly surprised to see that measured X axis had not changed and Y axis had increased by 0.002" (50 microns). Y axis was corrected in an updated profile.

I recently saw advertised some (Chinese?) carbide engraving bits specified to have a 0.1mm cutting tip and a 10deg angle, so I decided to try these for engraving boards for surface mount components. Initially I used the same pcb-gcade feed and spindle settings settings as for 30 degree 0,2mm tools I normally use (1000mm/min and 18,000rpm) - these resulted in the rapid demise of the tool after only a few seconds!!

Next trial used a much lower feed rate of 300mm//min and upped the spindle to 24000rpm. the results look quite reasonable (see attached photo) after the engraved board was polished using 2000 grade wet and dry to clean the cut edges. Board layout parameters were set to 0.012" (0.3mm) track and 0.012" isolation using Eagle 9.2.2 to autoroute the traces. Engraving parameters in pcb-gcde were set to multiple pass and a step size of 0.075mm, feed was set to 300mm/min and tool diam to 0.1mm. Spindle speed was 23,000rpm and cutting depth 0.15mm.

Results look reasonable, although the slow feed and small step size resulted in a 25 minute marathon even on the small trial board.

Peter

The Steppi performs better ;o)
Just have a look at the pictures in this german thread: PCBMillingTable

First image shows the prototype of a PCB based heatbed for a 3D printer. What you can see is, that, given the PCB itself is evenly thick, a PCB-milling table helps to have even-thick traces across the whole PCB.

Next you see my test for SMD. You can even have a trace between 2 legs of the SMD.

The next images show the bits I use. Not cheap, but very good. And with hard-paper-PCBs they also last for a while.

#12277 shows the measurements of the traces.

At the moment I try milling PCBs with EstlCAM software + Arduino, which allows to probe the surface before milling.

I'd really like to build my own PCB milling spindle running at something around 60000 rpm using a brushless motor and maybe extend the PCB milling table by an oil-bath for cooling, better cuts and extended lifetime of the bit.

End goal looks similar to this: PCB
(With end goal I only mean the look they showed. I think the Steppi is better than the wegstr, as they also don't have traces between legs of the SMDs and it looks like the Steppi is milling faster than the wegstr)
I already tried to build my own "spring clamp" for scraping off the soldermask - so far no luck because I don't have a lathe. I drilled a long hole into some 8mm aluminum round rod, glued in the stump of a killed bit on one side. But this way I ended being off for 0.1mm from the center. So, mounted to the motor I have too much vibrations. Experiments will go on ;o)