0.55mm Kyocera Endmill breaking instantly
Hey everyone, first time post. I am using a similar design to a RS-CNC32 (https://www.makerfr.com/en/cnc/rs-cnc32/) to mill PCBs on FR4 copper clad board. I flattened out a piece of waste board, mounted my blank to it with some custom clamps and tried a variety of bits. At first I tried a 60deg bit but the height variation was horrible. Then I bought some 20deg ones and the height variation was still pretty bad, and causing traces to be too thin/thick. I think the 20deg bit was slowly chipping away as it milled because when it finished the end was almost flat.
I decided that I would go for the smallest flat endmill I could for QFP pins (0.55mm) and bough a pack of 5 from drillman1 on ebay. As soon as the bit hit the copper spinning at 25000rpm it snapped. And then another one snapped as I was zeroing the height. They are so delicate! How is it going to take any cutting forces if it can't even touch off on the material?
Anyone have any tips on what to do? I can provide any details necessary to help.
Thanks in advance,
Clayton
Re: 0.55mm Kyocera Endmill breaking instantly
Hi,
I use 0.5mm Kyocera Tycom endmills also from drillman and get hours use out of each one.
It was not like that in the beginning though, I broke them, lots of them until I learnt how to use them.
Firstly investigate Autoleveller, a software utility that probes the board and modifies the Gcode Z value to follow the wave/bow/warp of the PCB.
I've used it to very good effect for five years.
Note that it doesn't matter whether you use endmills or 60 degree or 20 degree engraving bits, Autoleveller will help hugely. I don't bother
facing the spoil board, Autoleveller will accommodate a few tenths of a mm in PCB warp/bow AND another few tenths of a mm because the spoil
board is not quite flat.
Next....do not clamp the PCB down, the clamps WILL induce bow/warp in the blank. Use double-sided tape to hold the board down.
I use double sided tape and drill two 1.5mm holes and fit two 1.5mm pins to hold the board in place. The pins not only prevent the board
from moving laterally they will, if you chose the hole spacing wisely allow you to flip the board over and have the top and bottom layer line
up. Now all of my boards are doublesided SMT.
Most recently I've started using TSSOP ICs which require 0.2mm between pin pads and so I've been using engraving bits. I've standardized on 0.1mm
tip, 45 degree engraving bits by Kyocera Tycom from drillman. I buy 30 a a time, I usually get ten plus hours from those before the cut quality degrades and I
replace them.
When I started I was making boards with really thick copper layers (12 oz or 420um!!!). Angled engraving bits were no good at all in this thick copper so I used 0.4mm endmills
from drillman. That meant I could use SIOC ICs with 0.6mm between pin pads, smallest trace width was 0.4mm and smallest spacing between traces of 0.4mm. Each side took ten hours,
and to start with I couldn't get more than a 1/4 hr from each. Once I started using Autoleveller and also flood cooling I increased that to five hours! Flood cooling
flushes the chips out of the cutzone and is very VERY important to getting long life out of these tiny and fragile endmills. Note that you have to chose your doublesided
tape carefully and /or use duct tape around the edges of the board to prevent the coolant from getting underneath and allowing the doublesided tape to let go.
I get up to ten hours cutting with one endmill before it gets so blunt that I replace it, so it can be done.
Craig
Re: 0.55mm Kyocera Endmill breaking instantly
It is not enough to simply "flatten" the PCB. No matter how you mount it, the height variations will still be enough to spoil the traces and/or break the bit.
You need some way to ensure the constant depth of cut - either a depth ring, or surface probing. AutoLeveller is the program many people successfully use.
Spindle runout is also critical. It does not take much runout to spoil the cut and/or break the tool tip. Many entry-level "spindles" - DC motors with ER11 chucks slapped on - are nowhere near accurate enough for repeatable PCB milling.
Also, a homebrewed machine with limited rigidity may require VERY conservative feeds and speeds. My current settings are 4000RPM and 100mm/min, for instance.
https://i.ibb.co/NW6LbBN/IMG-20200729-124337.jpg
3 Attachment(s)
Re: 0.55mm Kyocera Endmill breaking instantly
Quote:
Originally Posted by
CitizenOfDreams
Spindle runout is also critical. It does not take much runout to spoil the cut and/or break the tool tip. Many entry-level "spindles" - DC motors with ER11 chucks slapped on - are nowhere near accurate enough for repeatable PCB milling.
I must agree spindle runouts is a nightmare for milling precise PCB's, but we all have slightly different out comes with DC motors with ER11 chucks,
below are pictures to prove otherwise.
Attachment 464588
Testing new 20º 'V' bits, Spindle is DC motor with ER11 Chuck.
Attachment 464590
Milling PCB with DC motor with 20º 'V' bit
Attachment 464592
Repeatability guaranteed, some claim no success with DC motors with ER11.
Will add that no need hold down PCB with bolts & nuts, double sided tape should do, I use a 100um tape at ONLY three points :)
Cheers,
Hanspeter.