PCB milling & Pick & Place

[eurofox]

Someone got experience with this one?

http://cirqoid.com/?gclid=CIi53bvc5cECFanMtAodgn8A3w

[ElektroQuark]

I have saw the video. I only wanted to put the machine apart and add manually the components. It's soooo slooooow.
And it seems you must be there to feed components.

[Kjelt]

Milling PCBs did you see the copper between the 0603 packages? Useless IMO unless you mill away all the excesss copper.
Then P&P : watch the different needlesizes (color) they changed them manually. No feeders so you have to place the components and program exactly where they are and even then you need to manually adjust them prior to placing. IMO better get a manual P&P and a seperate mill (if you really want this , nowadays I would just order with a pcb house and wait a week).

[cncjerry]

PCB trace milling thru isolation is something I have tried for years and I have found it unreliable even using expensive equipment.  I can enumerate the problems beginning with the challenge of holding the PCB flat.  I used a vacuum table with a manifold and reserve storage but even that technique wasn't good enough to overcome the variance in depth.  Too deep and the cutter became so hot it glowed; too shallow and you would miss material to remove.  The list goes on.  Even if I was successful in milling, the copper swarf would roll up and embed itself in the cutter paths requiring a tool change and retrace with a needle to remove the swarf.  I had some success with .005" cutters spinning over 50k, even as high as 90k rpm, but then they would invariably break.  Forced air helped some as well as a light oil.

I'm sure people have had good luck on occasion but ultimately is was cheaper to use the PCB services to etch the boards.  I still do it for large power supply traces but not for anything requiring precision.

Jerry

[mazurov]

I'm sure people have had good luck on occasion but ultimately is was cheaper to use the PCB services to etch the boards.

You're presenting it backwards. You want to make PCBs consistently - find the smallest clearance as well as other design parameters that _you_ can produce consistently, write down design rules for the process and stay within them rules.

I mill boards for DC-DC and certain RF prototyping since in both cases the feature sizes are typically large enough and I want to stay on one side anyway so I don't really suffer from DRC restrictions that much. Given that milling is done on a general purpose CNC machine the process costs very little money. Turnaround is so fast it is hard to compare to anything else available which is handy for RF work. I can mill and measure grounded CPWG on a piece of FR-4 faster than most people spend figuring out which online calculator is more accurate.

On the other hand, milling QFN footprints is difficult and most require thermal vias underneath a package so designs of this kind go to the boardhouse.

[mrpackethead]

I had a $30,000 LPKF S63 machine, suppsoed to be the bees-nees of machines, and it was a pile of .....   After six months of trying it went back,  Never going to try this again, its just not a good idea..  PCB's are so cheap and fast to make now.