Re: [CAD_CAM_EDM_DRO] Milling / drilling PC boards on home shop CNC equipment.

Hi Larry,

Thanks for the comments, more below:

Larry Edington wrote:

>
> I've been using my 2' x 3' CNC router table to make PC boards for years. It
> uses McMaster 5/8 ball screws and nuts. It's plenty accurate enough for
> drilling and milling PC boards. You don't need micron accuracy to do
> standard 10 mil trace / 10 mill spacing PC boards. Make the traces and
> spaces larger and your accuracy requirements go down further.

Did you build from plans, or BUY? Can you post pix?

> Dremels are also good enough. I've used the cheap ones up to the new digital
> speed control model as well as the Dremel flex shaft. I have seen NO issues
> concerning Dremel bearings causing routing or drilling errors.

I have a new one with flex shaft I'm itchin' to try.

>
> I do prefer the air grinders. I paid $50 for mine. It looks like a Dremel
> flex shaft hand tool. I've got a large compressor so it's no problem to run
> it for long cutting sessions. My next favorite is the digital control
> version of the Dremel. You want the highest speed you can get for the
> cleanest cut.

How about a cheap die grinder (air)? Should I be thinking of using
that?

>
> The most important thing for isolation milling is a floating head. You want
> the head to move down and put pressure on the PC board blank to help keep it
> flat. Or to at least contact the board to ensure the bit penetrates into the
> blank the same amount each time.

I was asking about that about a week or two ago.

>
> For this I use a small slide that the lead screw has been removed and a
> return spring is attached. I mount this slide to my normal Z axis slide. The
> router head ( a dremel or air grinder ) is attached to the small slide. The
> small slide has a 'foot' that is attached with a threaded screw. The bottom
> of the foot has a Teflon pad glued to it. When the Z axis moves down, the
> foot contacts the PC board and does not allow the bit to penetrate any
> further. More downward movement of the Z axis just results in the upward
> movement of the small spring loaded slide. There are many ways to accomplish
> this. You can do this fancy or simple.

I think I follow that, but sure would appreciate a pix!

>
> You adjust trace isolation width by adjusting the depth of the bit via the
> screw on the foot. Even if your PC board is not flat the floating head will
> ensure an even depth of cut.
>
> Get drill bits that have the plastic depth collars on them or add them with
> 1/8 ID wheel collars from a local hobby shop. Build a fixture you can drop a
> new bit into and then place the wheel collar on the bit. This way when you
> change tools, they are all the same distance from the dremel or grinder's
> chuck and your depths remain constant.

Good ideas.

>
> Think & Tinker sells 60 degree carbide isolation bits. That's what I use. I
> set my depth for a .010 inch wide isolation path and set the board clearance
> rules accordingly. For wider isolation I use .032 carbide end mills. I don't
> use these often.

Found them at:

http://www.thinktink.com/
http://www.thinktink.com/stack/volumes/voli/store/mechmill.htm

>
> Another option is an array of holes drilled for a vacuum hold down plate.
> This would hold the PC board as flat as your MDF surface. But would be
> dependent on perfect axis alignment and perfect flatness of your table. The
> floating head eliminates a lot of potential errors.

And MDF is a dense foam, wasn't that a thread a while back (I knew I
should have read it closer)?

> The perfect solution would be a floating head with vacuum hold down. That
> way you can eliminate the double sided tape.
>
> Side to side alignment is no problem.
>
> After you have finished laying out the board, place two .125 holes exactly
> in the center of the board, but one at each end of the board.
> Cad tools make this easy. The holes should be placed left to right exactly
> the same distance from the center of the board horizontally and in the exact
> Y axis centerline vertically.
>
> Order your excellon drill file so that these two 1/8 inch holes are drilled
> first.
>
> You'll need something you can drill into under the board. I use 1/2" MDF
> stuck to the aluminum surface of my router table with double sided tape. I
> also stick the PC board blank down to the MDF with double sided tape.
>
> Your first two holes should drill 1/2 way down into the MDF after going
> through the copper clad blank. Now insert two 1/8" pins ( old broken bits
> are perfect ) into these holes. These are your locator pins. Make sure they
> don't project up above the PC blank far enough to allow the drill or router
> bits to hit it or your depth foot to contact a locating pin.
>
> Now drill the board and mill the top side's isolation paths.
>
> If your drill speed is high enough, you won't need centering holes drilled
> in the board blank. The carbide bits won't wander all that much before they
> break. If they are sharp they will go right into the copper dead center
> perfect every time.
>
> Next flip the board left to right and place back on the pins. By the way,
> the gerber file for your bottom side must be mirrored for this to work.
>
> Now mill your bottom side which is now facing up.
>
> It's as easy as that. It'll take some work getting your cad software
> procedures figured out and getting the the hardware setup on your router or
> mill to do this. The floating head is the hardest part.
>
> Print some gerber files out on transparency plastic to get the board
> flipping figured out.
>
> Once you are done you can go from board design to a finished board in under
> an hour. That beats a 3 day wait from the board shop.

Thanks for the instructions. This is good stuff for a website!

>
> Kellycam now has Gerber isolation paths working in his cnc software. It
> works fine for generating the isolation paths but Master5Cnc is much better
> for running the stepper motors to cut the paths. You won't likely find an
> isolation program for $125 anywhere else. Isolator Pro from T-Tech is $1995.
> Ztrace which in my opinion is a worthless piece of junk is $320. Eagle PC
> layout software generates isolation paths as does a couple of programs from
> Germany. Kellycam is the cheapest simplest solution I've seen so far.
>

I JUST posted a comment on Ztrace, sounds like you're not to hot on it.
But Kellycam is $125, and you feel it's up to the job for hobbyists
(more?)? Eagle was one program I was exploring.

> I use Protel and Pads PCB layout software. This is what works for me. Your
> mileage may vary.

I don't think I ready to shell out that much for Protel, the other
programs I'm hoping will do it for me! Then again, it's something I'd
like to write!

Alan KM6VV

>
> later,
> Larry E.
>
> ----- Original Message -----
> From: "Alan Marconett KM6VV" <KM6VV@...>
> To: <CAD_CAM_EDM_DRO@yahoogroups.com>
> Sent: Monday, July 23, 2001 12:24 PM
> Subject: Re: [CAD_CAM_EDM_DRO] Leadscrews, PCB router experiment
>
> > Hi Jon,
> >
> > Yes, I take your point. I'm hoping for simple PCB's, designed with
> > "relaxed rules", and less stringent requirements. Side alignment WILL
> > be a problem. Then again, it can be a learning experience!
> >
> > Alan KM6VV
> >
> >
> > Jon Elson wrote:
> > >
> > > Alan Marconett KM6VV wrote:
> > >
> > > > Hi Jon,
> > > >
> > > > I have a 3/8" slab of aluminum, that might have to do for my base. I
> > > > want to do more then drilling, I want to be able to mill traces as
> > > > well. I'm planing on using a dremel tool.
> > >
> > > We've gone around with this a few times. I don't think the standard
> > > Dremel moto-tool spindle is good enough for PC board drilling.
> > > The problem is that you have to hit the 'target' quite accurately, or
> the
> > > hole will cut through the side of the pad, rather than through the
> > > middle, leaving a complete ring of copper around the hole. I even
> > > have problems on my system, with the air bearing spindle motor,
> > > high precision ground ballscrews, encoders delivering resolution
> > > of .00005" (yes that's 50 uInches) and the physical stability of a
> > > Bridgeport. If you mill the board, then the holes should at least
> > > line up with the copper, but you still have to deal with alignment on
> > > the other side, if you do 2-sided boards.
> > >
> > > Jon