Probably starting a religious war, but I've actually had better luck with engraving bits that actual mills for traces. I change over to a real mill for milling operations.

What I really need to do is master gcode well enough to have a routine that re-zero's the height after changing a bit.

Bob


On Thu, Mar 21, 2019 at 4:26 PM John Ferguson via Groups.Io <jferg977=[email protected]> wrote:

I suppose swarf isn't the best word, but I'm curious about how you guys get rid of the chips which lodge in cuts.?? I've been chasing them out with a needle, buy it's tedious.

I've got 90 psi air in the shop. Maybe I should try that.

I've got my production system running very smoothly. Design In Eagle, generate Gcode in pcpcode, sometimes include autocad millwork with gcode generated in SheetCAM, set up work piece, set height on cutter and go.

I make all of my boards from 2.5x4.0 workpieces. I made an aluminum fixture with a depression which fist these workpieces and has a lid which locates 4 holes one in each corner.?? I drill the holes and then screw the board down to an HDPE fixture bolted to the table on my 6040 router.?? The HDPE has 4 holes which take 6-32 nylon socket head caps screws.?? I face off the surface the board sits on every 4th or 5th board to maintain level.??

I adjust the bit height by rolling a piece of .250 drill rod back and forth until the bit just clears and then dropping it one click on the servo so it doesn't quite clear.?? I have vertical "home" set at Z=1.000 on the machine so I have to raise the spindle .750 before I start.?? I suppose I could lower the home to .250 and eliminate the "raise" but having it at one inch gives me enough time to catch some kinds of screw-ups.

I've been delighted with this bit. I buy them from Precise Bits and they seem to last pretty well.??

EM3E8-0100-15V 0.0100in. (0.254mm) tip dia., 3-flute 15?° tapered stub trace-isolation bit, 0.020 in. DOC, 1.42 in. (36mm) OAL 2 $16.95 $33.90

The size of the bur on the edge of cuts seems to increase as they wear out, but only downside of this seems to be more dressing of the finished board. I use 400 grit paper and final clean up with 800 crocus cloth.

I've been tinning the finished boards with MG Chemical 421.?? I'm a bit mystified why I get better tinning if i rough up the board with 400 grit before plating it.?? I wash board with acetone before tinning.

After I've soldered everything (all through hole - no smds yet), I go over the board with a loupe and remove any shorts with braid and clean out any suspicious cuts with a needle.

and they always work, at least the ones I haven't messed up the circuits on.

If you read this and see anything stupid, let me know.?? Also I'm still like a better way to get rid of chips.

regards,

John ferguson?? delray beach

All the spot drills are done with the same tool as is used for the track milling - I've used both 0.1 and 0.2mm? tools for this.?? The drill file is completely separate and calls for drills as required by the component definitions in Eagle.?

Spot drill code is part of the main "etching" GCODE file - this is run as received from pcgcode, no editing required whatsoever. The drill info is in a separate file and contains all the necessary spindle stops and starts to allow for tool changes and setting the Z axis zero using a touch probe.

Drill depth is set for about 1.7mm when using 1.5mm FR4 boards. I manually "adjust" this for larger drills (up to 3mm) to allow for the 120 degree drill point, but no other changes are made.

Peter

I've used most Eagle revisions from 7.7 to 9.2 with equal success w.r.t .pcbgcode if that's any help to you.

Peter

Hi,
Not trying to pour cold water on your findings, but I've never had a problem with spot drilling!!? I did a check of my code in the area you modified and it appears to be exactly the same as the one that misbehaves - see below (cut and paste) :-

if (SPOT_DRILL == YES) {

if (m_first_spot_drill) {

output_drill_first_hole(drill_x, drill_y, SPOT_DRILL_DEPTH);

m_first_spot_drill = NO;

}

else {

output_drill_hole(drill_x, drill_y, SPOT_DRILL_DEPTH);

}

}

Something else must surely be going on elsewhere in the programme to change the SPOT_DRILL_DEPTH ?
To my inexperienced eye the original code seems to be entirely logical.

Peter

Hey John,

I had experienced your issue as well, and I cruised through the code and I managed to find out how to fix the error.. I adjusted the code as follows

The issue is a flag reset in spot drill. the first one comes out fine, but then the rest are drilled to fill drill depth.

?

In pcb-gcode.ulp you will find this sequence:

//

? if (SPOT_DRILL == YES) {
??? if (m_first_spot_drill) {
????? output_drill_first_hole(drill_x, drill_y, SPOT_DRILL_DEPTH);
???? // m_first_spot_drill = NO;
??? }
??? else {
????? output_drill_hole(drill_x, drill_y, SPOT_DRILL_DEPTH);
??? }
? }

///

?

The underlined line is the culprit. it spot drills one time, then spot drill is set to NO, so it doesn't do it anymore. I got the correct behaviour by commenting out that line. All the drill files are to the correct level. It works as it should on my machine - running on the GRBL config.

Hope this helps!

Joel

John,

Just tested the "arc theory" and can confirm that milling layers contain the necessary GCODE.

Of course due allowance will be needed to allow for tool diameter, and the GCODE will need to be edited to set correct cutting depths for multiple passes.? Personally I set this to 0.3mm when using a 3mm diameter tool and normally set the lowest cut to 0.1mm below the bottom surface.

Peter

John,

I experienced a similar problem in EAGLE 7.7 as well. From what I can find is that the milling layer does not recognise holes as routing paths.
Something I haven't tried yet is the make to hole with two 180 degree arcs. I suspect this will work from my experience with arcs on other boards.

Ken