SparkFun Forums 

Where electronics enthusiasts find answers.

Have questions about a SparkFun product or board? This is the place to be.
By skimask
#144574
I started using a Sable 2015 CNC machine to make PCBs.
Good enough resolution and accuracy for me and beats waiting X number of days for a PCB to get made.
http://www.youtube.com/watch?v=JcnumN8LPYg

Problem/Issue:
I built a "vacuum table" to hold down the PCBs rather than using clamps, screws, etc.
Works great on double sided PCBs. The copper holds down solid. I can't pry it off with my fingers and it doesn't move at all during machining.
Problem happens when I'm using single sided PCBs. The fiberglass/phenolic side won't stay down nearly as well as the copper side. I figure it's because of all those little ridges created by the fiberglass itself, all coming together to leak air.

I don't have any good ideas as to how to hold it down. Lots of silly ideas, no good ones. I've tried double sided stick tape which tends to pull up some of the MDF. Thought about spraying the back of the PCBs with a polyurethane, letting it dry, then sticking it on, but figure the poly would kinda soak into the MDF and cause it to swell up, thereby ruin the flatness. Tried using electricians tape, scotch tape, packing tape, etc, to seal the bottom of the blank PCB as well as going around the edges. No dice.

Any other swell ideas?
By Mee_n_Mac
#144576
Would using a vacu-forming idea work ? That is lay a sheet of flexible plastic, perhaps even a few layers of Saran Wrap, over the top of the PCB to be milled. Then mill right through it where needed. The air pressure should still seal it tight along the top outer edge despite any interior milling. Then again this ends up much like taping it down to the table. I'd have thought that would work, given a good seal along the top edge.

Would a bead of some caulk (removable) along the edge work ? I'm not sure what would be best, as in easily removable w/o messing up the MDF.

What do you use to seal the holes in the vacuum table not covered by the PCB ?
By skimask
#144582
Vacu-forming - I like it. Simple and straight forward. I'll give it a shot later today.
When I tried the tape yesterday, I put the tape along the bottom edge with the non-sticky facing outwards thinking that I only had to cover the little "micro holes" in the edges of the fiberglass layer. Ya...Not so much. Didn't try taping over the top of the PCBs themselves though.

Caulk - But it's soooo messy. And, ya, would probably mess up the MDF. Eventually, I'll probably make the table out of a chunk of aluminum or at least a good hardwood. Until then, the MDF is practically disposable and I've got a nice drill press and LONG bits to make those holes.

Uncovered holes - I picked up a box full of single sided PCBs (roughly 200), all 2.75" x 8", for dirt cheap a few weeks ago. This particular version of the vacuum table is a custom fit for those, i.e. no open holes.

Next time, I'll just buy double sided PCB blanks instead and figure out a really good repeatable method of registering them so I get the etching aligned properly. Manually routing a single sided PCB is getting to be a PITA.
By skimask
#144615
@lyndon - I'm on cnczone, but not a huge following of folks over there doing PCBs. Finding threads here and there, nothing substantial though. I'll check out your link.

@fll-freak - Gonna check out using an o-ring as well as the thing gasket material. Problem is what to do with that last couple sub-thousandths of flex in the gasket material itself, especially after a tool starts pressing/drilling/etching on the material on the top of the board. Will it flex or won't it? I'll find out eventually...

@Dave - If I knew I could get exactly the same "squish" everywhere on the PCB so that both boards were perfectly parallel to each other, then hot glue would do the trick. Unfortunately, you know how it is...no "steady as a rock" hands here. I'd end up with one side a couple sub-thousandths higher than the other, and you know what that does to the depth of the cut...
By AndyC_772
#144643
Many years ago I was working as a student in the workshop of a (sadly now defunct) electronics company, and I made a vacuum chuck out of a block of aluminium. It was hollowed out and covered with a plate with an array of small holes in it, each around 2mm or so if I recall correctly.

It worked very well, and I wonder if there might be a couple of issues with the design of your setup:

- could the MDF be slightly rough and/or porous itself? You might get away with it when it's sealing against a very flat copper surface, but the combination of the surface roughness of the PCB as well as that of the MDF is just too much. Maybe something as simple as sealing the surface of the MDF with watered-down PVA glue would help reduce the porosity and produce a smoother finish?

- the holes you're using are quite large; one or two of them not sealing properly will result in a lot of airflow, which will tend to reduce the negative pressure achieved at all the other holes too. If I were redesigning the chuck, I'd start with a hollowed out block of something and then drill a larger number of smaller holes, so any one of them failing to seal wouldn't have such a negative impact on the suction.

Best of luck :)
Andy
By skimask
#144650
MDF rough and/or porous? That's what I've been reading. Not so much that you can push air THRU the MDF itself, but like you said, the fiberglass roughness + the MDF roughness = no seal.
Sealing the surface isn't an option as it'll cause the MDF to swell up and "delaminate" (I know that's the wrong word but it seems to fit) from itself and basically fall apart.
The holes are very large! But, I already had the 1/8" drill bit for going all the way thru the MDF No sense in spending a load of $$$ on something that might not work in the first place :) And I resurfaced the top after drilling the holes paying attention to the bit to make sure it actually made cuts as it went across each hole. Had to cut an extra couple thou out of it to get cuts on all the holes.

Since this actually half-way works, the next version will be made of a much better material for starters.
I'm thinking 2 blocks, both with a bunch of channels cut across the length/width (I'm leaning away from the hollowed out blocks if only for the fact that it might collapse a bit under the vacuum), both block sandwiched together after the channels are cut, and like you said, a load of much smaller holes drilled in it.

I've only got $60 in pieces/parts invested into the block so far. Went ahead and added 3 more vacuum fittings yesterday, one on each corner, trying to get more volume. It helped quite a bit, but still not perfect. The mess of fittings/hoses/etc all hooked up to the shop vac looks a bit goofy, but hey, it works (sort of).
Getting there...........
#144652
A variant of the hot glue idea. You can get spray adhesive. That should put a fairly uniform layer down. If you don't use too much, it won't be permanent. Instead of a copper clad FR4 board, you can get aluminum clad fiberboard (not sure it's exact name) that is designed as backing for drilling PCB. Cheaper and not as hard on the drills as FR4.
By Dave Mueller
#144690
Spray adhesive would work, as would the thin (like Scotch tape) double sided tape used for window insulation. I was thinking of setting the single sided board on top and just a blob of glue in the corners to hold it steady X-Y, or a bead around the edges. Or foil tape, the aluminium stuff is pretty sticky and strong. Or solder tabs. You just need to plan for a border on the single sided PCB.
By skimask
#144778
Weeeellll....
Turns out... :)
That the surface wasn't completely flat.
After trying saran wrap, tape, packing tape, etc, I went back and re-milled the surface with the same 3/4" router bit, only at a much slower speed...and that cured the problem with the fiberglass side. Great suction, great holding power, no need for tape, saran wrap or anything. Go figure... I don't remember what travel speed I had selected when I originally milled the surface, but the 2nd time around, I ran at about 10inches/minute. Maybe I had it hauling the first time around. I dunno.

Now, a new problem...possibly related, quite possibly...
Got around to doing more etching and found that when the Z axis is set to etch at a good depth around the edges, it totally misses in the middle. Doesn't do anything once I get about an inch from the edge, any edge.
The only thing that could possibly cause this is the vacuum actually sucking the middle of the board down so hard that it bows up the outside edges, either that or the MDF is flexing inward in the middle. I find the MDF flexing kinda hard to believe since I've only drilled holes thru it, not hogged out the whole interior of the board. It should be well supported in-between the holes themselves and keep the whole thing from trying to collapse inward...or does it?
Next step is to try a harder material like a solid chunk of oak. Maybe all those MDF fibers are actually moving just enough to wreck it.

The other next step we're gonna try next weekend is to go back to chemical etching (laser printer, toner transfer, ironing/laminator, etching, etc.). At least if I get good registration on the boards, I can still use the machine and the DRL file to drill all those holes...
By chartle
#144780
skimask wrote: Now, a new problem...possibly related, quite possibly...
Got around to doing more etching and found that when the Z axis is set to etch at a good depth around the edges, it totally misses in the middle. Doesn't do anything once I get about an inch from the edge, any edge.
The only thing that could possibly cause this is the vacuum actually sucking the middle of the board down so hard that it bows up the outside edges, either that or the MDF is flexing inward in the middle. I find the MDF flexing kinda hard to believe since I've only drilled holes thru it, not hogged out the whole interior of the board. It should be well supported in-between the holes themselves and keep the whole thing from trying to collapse inward...or does it?
Next step is to try a harder material like a solid chunk of oak. Maybe all those MDF fibers are actually moving just enough to wreck it.

The other next step we're gonna try next weekend is to go back to chemical etching (laser printer, toner transfer, ironing/laminator, etching, etc.). At least if I get good registration on the boards, I can still use the machine and the DRL file to drill all those holes...
I left you a few comments on your you tube video but as for your flex issue I can't see the MDF flexing because of the vacuum but maybe.

Oak would be a bad choice first its natural wood and will expand and contract and not stay flat no matter what you do it. Also I think white oak is porous and would leak air. Red would word but again it real wood.

Here is thought, can you use two layers of 3/4 mdf? On the bottom layer rout a grid of connected groves with one "port" to the side. Glue and screw the top layer being careful not to clog the grooves (maybe just around the edges?) and then punch your holes through that. Now you have a full 3/4 inch surface.

One other thought to your OP, you could use craft foam for your surface as an all over gasket. But I'm not sure it would compress uniformly enough.

OK one other thing why don't you have some sort of registration on your MDF? It would help hold the board.
By skimask
#144784
chartle wrote:Oak would be a bad choice first its natural wood and will expand and contract and not stay flat no matter what you do it. Also I think white oak is porous and would leak air. Red would word but again it real wood.
Yep, kinda figured that no matter what kind of wood, but it'll work enough for testing, at least before I go out and get a thick ol' chunk of aluminum and spend money that doesn't need to be spent...yet...
Here is thought, can you use two layers of 3/4 mdf? On the bottom layer rout a grid of connected groves with one "port" to the side. Glue and screw the top layer being careful not to clog the grooves (maybe just around the edges?) and then punch your holes through that. Now you have a full 3/4 inch surface.
Had that idea too. Would be a load easier to route grooves vs. trying to drill straight, long holes, and will probably be stronger in case it actually the MDF is actually flexing. I was thinking of routing grooves in both halves, only a fair amount deeper in the bottom half, maybe barely routing the top half just enough to show lines where the holes should go.
One other thought to your OP, you could use craft foam for your surface as an all over gasket. But I'm not sure it would compress uniformly enough.
Don't really need the gasket any more since I figured out that the surface wasn't flat in the first place! That's what I get for double-checking...or rather...NOT double-checking...my work.
OK one other thing why don't you have some sort of registration on your MDF? It would help hold the board.
I've got 4 tiny holes at the corners to help line it up, but nothing to hold it in place other than the vacuum, which now works like a champ. When I get to doing double-sided boards, I'll definitely have a few pins placed in strategic spots. Nothing like routing a double-side board...offset top to bottom by .1" :)

I'll get 'er eventually. I know I'm not the first one to do this...(sigh)...Probably gonna end up like everyone else. Try to do it on the cheap, spend a lot of time and effort only to spend the big $$$ later on...
By Mee_n_Mac
#144787
I assume there's an airspace below the bottom surface of the drilled MDF and that the MDF is supported only along it's edges ... why not put some support posts in the interior of the air chamber.

Have you confirmed that the bed is bending ? Laid a straight edge from side to side and seen the deviation, vacuum vs non-vacuum ?