small accurate holes

A fair bit back, Jim Rozen described a drill press for Printed Circuit Boards that sounded like a good design. A microscope with a cross hair to align the board, and a drill that comes up from the other side to drill the hole.

So the first step was to figure out what to use for a spindle for the drill press and how to drive it. My choice was a small trim router that HF sells as low as $20. The second thing was to make a new collet to hold carbide PCB drills. Measured the angle of the existing collet as 30 degrees ( 15 degrees on the lathe compound ). But how to get an accurate hole with very little run out.

Well the first problem was to get an accurate diameter hole. A 1/8 inch bit makes a hole that is too large. And the next size down was too small. So googled on this group for ideas and tried to make a small boring bar out of a broken drill. The drill was about a 3/16 drill and I never did manage to make a decent boring bar.

Okay a D drill will do it and I had several broken PCB drills. And fortunately a small ( about 4 inch )diamond wheel. Much better luck than I had trying to make a boring bar. Got one collet drilled out, but had guessed at the outside diameter. Bad idea. Made another collet and nearly got it right. I left a little straight bit at the small end of the taper that was just a little too big in diameter. So pushed a broken PCB drill in and then chucked it in the drill press and filed that to fit. Also smoothed up the tapered part with the file. It seemed to be true before I did that, but it couldn't hurt.

So now I more or less have a collet. I will have to buy a slitting saw to put a slit in it, but the hole is tight enough to hold a PCB drill now. Can't see any run out. I think that part will work.

Dan

Reply to
dcaster
Loading thread data ...

Dust? Drilling PCBs makes lots of abrasive dust that will get into your bottom motor spindle bearings. If the board is already etched, you don't need a microscope or cross hairs to align the drill bit in the center of the pads. The lack of copper in the middle of the pad will center the bit automatically if the operator can get it reasonably close. And if you need better accuracy than that, then CNC is the way to go (it doesn't need the microscope or cross hairs either).

[...]

Aren't there 1/8" collets in most Dremel-type motors? Making a collet from scratch sounds like reinventing the wheel. Carbide PCB drills #60 and below like to run at 25K RPM -- your home brew collet had better be pretty accurate in run out if you expect the bits to last longer than about three holes.

Reply to
Tim Killian

The old Excellon PCB drilling machines used a similar arrangement. They had a projection microscope that displayed the pads on a ground glass screen, with cross hairs. I'd be inclined to use a web cam for viewing the pad.

Leon

Reply to
Leon Heller

Good memory. I had run that thing at the university of arizona at tucson, in 1978. I honestly cannot remember if it was a commercial unit, or was home-made by the shop's boss. He was that kind of guy so it may have been home-made.

The spindle and motor were out of a Dumore drill press, I remember that. They were rigged to rise up from below under the action of an air piston, tripped by an electric foot switch. I seem to recall that the drill was held in a chuck, maybe the tiny albrecht one, on the motor spindle. A collet of course would be less expensive.

It was pretty filthy work, I think there was a vacuum rigged up to pull the swarf out of the working area. If I were doing that today I would vent it outside, G-10 is not the best stuff to be working with.

This was nearly before the days of NC pcb machines, and for protoypting work it was pretty nice. They also did their own multi-level boards via tape on acetate, which were then optically reduced. Then they had their own expose, develop, and etch setup as well.

These days board houses are numerous, good, and inexpensive - if you are doing even a few, you might consider just shipping them a gerber file.

Jim

Reply to
jim rozen

Make a thin rubber disk, pierce the center, and slip it over the drill so it covers the collet and bearings below. Make it small and very round so it won't come apart at high RPM. Might have to use a plastic disk

Reply to
Rex B

Not so much a good memory, as using Google to look for threads on drilling PCB's. I do try to read all of your on topic posts and some of the OT ones as well. Always interesting.

I will have to figure out some way to capture the swarf. I don't want something as noisy as a shop vacuum running but will have to rig something.

This is just for prototypes. If I get someth>

Reply to
dcaster

Good idea. I have a web cam that I can use. Maybe just draw a cross hair on the screen with a sharp wax pencil when I use it.

Dan

Reply to
dcaster

Reply to
Tim Killian

Reply to
Tim Killian

I am still thinking about the dust problem. My thoughts now are to use a smallish squirrel cage fan and suck the dust into a pleated paper filter. And build a box around the motor and blow air into the box. Maybe a blower with two squirrel cage blowers or maybe the filtered air would be clean enough to use to cool and keep the motor clean.

I think my collet is accurate enough, but the proof will be if I can drill a # 80 hole. Making a collet from scratch is not cost effective. If I counted my time as worth anything, it would have been cheaper to buy a Dremel tool. But I did learn from the experience.

Dan

Reply to
dcaster

I am thinking of doing something like that.

Reply to
dcaster

I already have a regular drill press, but not one suitable for PCB's. The router motor has a lot more RPM's and less run out. I hope to make something suitable for a lot less than a small high speed drill press would cost. Hell I hope to spend less than a 0-1/4 precision drill chuck costs. Changing bits will be more of a pain, but I hope not too bad. The router motor will be held in a clamp so it can be easily removed to change bits.

Dan

Reply to
dcaster

I don't know what your budget is, but for $200 you can buy a very capable hobby machine with multiple uses:

formatting link
I already have a regular drill press, but not one suitable for PCB's.

Reply to
Tim Killian

Parallax error. The planar view of the trace on the board was a big help. The drill was swapped easiy from under the table.

Jim

Reply to
jim rozen

Here's what I do.

formatting link
Of course, I was incredibly lucky to get these Westwind air bearing spindles. The guy who sold them to me had no idea that bad units carried a $6000 core charge with the rebuilders! (Maybe a rebuild is $6000 with return of the core, I'm not sure anymore.)

I made a bracket to clamp it to the spindle of my Bridgeport. I have successfuly drilled boards with .018" drills. I can't drill a stack of two boards at that size, however. This blind drilling allows me to use backup material and entry material, so I get minimum ridges around the holes. I use blank (no copper) PCB material as the back-up, and .010" aluminum roof flashing as the entry material. I align pre-etched boards without the entry material first, then insert the entry material under the clamps on one side first, then the other, so that the board can't move.

I also pre-drill the blank boards, then laminate the resist and expose and etch. That way I don't have to fool around with the aligning on the mill. But, I can't have pre-laminated board blanks at the ready, either.

I make master artwork for exposing the resist with my laser photoplotter

formatting link
(Ugh, GOT to make new photos of this project!)

Before all this automation, I had a hand-steered drill with a pedal to stroke the bit. It worked pretty well, too. Really no need for a microscope.

Jon

Reply to
Jon Elson

I seriously doubt it! An Excellon drill has a 7 ton granite base, air bearing sliders on the X and Y axes, the spindles are mounted to another granite beam. The spindles use air bearings, and hold runout to under 0.0001". They can drill .012" holes pretty well at 80,000 RPM, but that is about the limit. These machines go for $200,000 and up. I was pretty amazed I could drill .018" (about #77) holes with my lash-up of a Westwind spindle on a Bridgeport. Using a standard VFD, I can only do 24,000 RPM (400 Hz), which is way too low for a drill that small. But, that is small enough for anything I'll ever do. Just the right size to poke a wire wrap wire through for a via.

Jon

Reply to
Jon Elson

No, actually the spindles and axis drives I have were RETIRED before 1978. It was a complete CNC PCB drill by Excellon. It used a GE Mark Century control, all discrete germanium transistors on single-sided paper-phenolic circuit boards. GHASTLY nightmare, I didn't even want to LOOK at the horrible control! (Not CRT CNC, not even CNC, but tape NC, just enough smarts to position to the X Y coords and stroke the drill.)

Jon

Reply to
Jon Elson

And the problem is...?

Reply to
Stephen Young

So it's the thing that takes the "drill tape" file from my PCB CAD program.

Reply to
Jim Stewart

Was that called a Uni Drill and then they had a Quad....thats going back a bit.

Reply to
John

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.