Aside from runout, feeds and speeds can be the culprit. FSWizard Pro gives the following for a 0.03" 2-flute, uncoated carbide drill in FR4.
RPM 20k
SFM 150 ft/min
Feed rate: 29 in/min
Torque and breaking torque are listed as equal.
In pure copper, RPM is 40k at roughly the same feed rate, for half the chip load. I don't have any thoughts on what this might mean for copper clad FR4.
Also check that the spindle mount is rigid and on axis with the feed. Strict perpendicularity to the work might not be so critically important, but can help reduce wander when starting the hole. A shorter bit is more gooderer, but they the bits seem to come in only one length.
I use a rotozip clamped to a linear guide. It develops 20k+ rpm and very seldom breaks bits. I attribute this mostly to the high RPM and the relatively rigid mount and guide. Any slop or flex in the mount is sure to cause problems. Hand held is definitely out of the question (I didn't read the entire thread, so maybe this isn't an issue).
Mike.
On 2/22/2016 11:33 AM,
craigl2@... [Homebrew_PCBs] wrote:
The real problem run out caused for me was the drill wandering before biting in and therefore being "off target". This was still a problem although less when I temporarily used a high-end spindle with a run out of less than .0005".
The runout is still present with the six dollar hobby RC motor I'm currently using but I found a way around the problem. Simply drilling through a .007' thick piece of cardboard helps somewhat but the drill would still wander off sometimes as much as .005".
The solution I found was to use a guide bushing positioned about .01" from the PCB. This results in the hole being drilled to .001" accuracy or better.
I am using a 1/8 inch thick piece of flat stock steel as the drill table. This provides three benefits:
1. Magnets hold the table to the frame so it can be quickly and easily removed for more open access to change the drill bits.
2. Two 3/8" X 3/8" neodymium magnets placed on top of the PCB allow it to be slid around for positioning but hold it (hands free) in place firmly for drilling. These may be a little stronger than necessary.
3. I made small bushings with a short 3/16" piece of steel drilled through with the appropriate size drill. These were pressed into a 1/8" thick plastic disk and four 1/8" X 1/8" neodymium magnets were pressed through 1/8" holes every 90 degrees. When I change to a different drill size I slip the appropriate size bushing down over the drill, replace the tabletop and slowly raise the drill until the magnets attach the bushing disk in perfect alignment to the underside of the table. Because there is still some runout of the drill there is a small countersunk area in the bushing facing the drill to guide it into the hole.
Craig