Milling a Solder Paste Stencil Mask for soldering in toaster oven

[guru_florida]

I tried milling a stencil mask for solder paste today with mixed results. I ended up breaking my only micro endmill (10mil) so that pretty much stopped me right away.

I was milling 10mil brass sheet. I used a helical cutting path 3.5 mil DOC, ~1 IPM feed, 2500RPM. I pretty much took a stab at the speeds and feeds.

It cut about 10 holes before breaking. The holes it cut were *perfect* tho! They matched with the pcb exactly and these are fine pitch SMD pads and some 0406 resistors.

I used the bare side of a pcb as the sacrificial material underneath the brass sheet. I used a test dial indicator to square the machine and flatten the bed to within 1.5mil over the whole board (near 0 deflection in the cut area though, so this wasnt a factor in the break.) The helical motion started above the copper so it wasnt trying to plunge.

I could see that clearage of the swarf is an issue but because its so small its hard to clear eve by forced air. I also had some tapping fluid trying to clear the swarf but I think that made it worse as it collected near the bit instead of getting kicked clear when it was cutting dry.

Anyone have any suggestions? These tools are about $20 each...so expensive for a trial and error strategy!

Colin

[kn6za]

Colin,

Check out www.millpcbs.com Phil has a video of his machine making a solder mask. follow the links to solder stencil and you will find the you tube video.

He is using one of my cutters to do it, and they cost much less. If your interested check out my web site.

[guru_florida]

Actually, I have some of your bits, the 60deg, they work great but I still find it a lot of work getting the z axis flat enough to get the resolution I need -- not a machine problem or your vbit, just my setup process.

I figured the V bit would make the stencil edges too slanted so I tried the endmill -- if only it didnt break it makes great cuts. Do you have a different degree vbit you would recommend? The 40?

I use the PCB Fab in a box (8 minute pcb) with the laminator and it works great and is quick. I think I am going to nail this Z setup prob with a depth probe and adjust the gcode. With EMC2, I could conceivable generate a gcode file that would automatically probe, store the results in variable map, and then start cutting in one swoop.

[kn6za]

Colin,

Phil uses the 45deg tool in his video. I have not attempted to mill stencils myself.

[Zig]

Try flipping the data so that the stencil is cut from the rear towards the front surface.
This way the slant edges will face downward and will hopefully provide a etter deposit of solder paste.

[Drools]

here we go again, gotta teach the new guy (me) the how/when/why of a SMD stencil. I watched the video making the stencil, now I need a video on the rest of the process. Why brass? I'm making it painfully obvious that I know nothing about solding SMD devices.

[Zig]

It might behove us to initially describe what the stencil is to do and then look for improved solutions.

Typically solder paste or as it is sometimes reffered to as cream is a suspension of solder balls in a carrier compound along with active flux chemicals.

Since it is a suspension ( a mixture of separable components) not a solution the paste may have less than uniform consistency.

This may be due to inconsistency in solder ball sizing, settling out of solder balls due to the way the paaste has been stored and or extruded onto the PCB and sheared sufficiently well prior to application.

The stencil is supposed to mask portions of PCB so that only surfaces where a soldered joint is to be formed will have solder paste them.

Bearing in mind the nature of the paste and the need to squeedge the paste onto the pcb through tiny apperture the rheology ( fluidity ) paste will play a major role.

Typically the surface area of the joint is much larger than the area of the walls of the apperture in the stencil. From this we can infer the paste ought not stick to the walls of the opening.

But....


There are a number of ways of cutting openings intot he stencil.

Chemical milling

Laser cutting

Mechanical milling

Each of these leaves a signature hole wall profile.

Suffice it to say that it is essential the walls be as smooth and as vertical as possible with perhaps a slight tilt on theminto the apperture to facilitate cleaner release of paste onto the PCB.

For us who are into mechanical milling it may be beneficial to use a conical cutter and flip data upside down to cut the stencil so that we do infact end up with a stencil when used the right way up with openinigs walls which tilt slightly inwards and promote cleaner "snap off" ( a silk screen printing term) when lifting the stencil post paste application.

[guru_florida]

Thanks Zig for the great post.

I made a probe today to help set the tool z offset and map the small irregularities in z height of the stock in order to use a v-bit.

I was also trying to cut my mask into 10mil brass and the guy in the video was using 5mil, so I will get my hands on some of that.

Drools: Brass is often used in stencil makiing, but just about any metal that can be found in 5-10mil sheets and cheap will work. Some have used plastic transparencies but I think the results arent as good. I used brass because I had it on hand.

If you are making any number of pcbs rather than a single prototype then a solder mask and solder paste just saves toooo much time to pass up.

Colin

[guru_florida]

I found some SFM/chipload data on the small carbide endmill I used. It didnt have chipload for 10mil endmills, but the chiploads for 1/8", 1/4, 1/2 and 1" were linear so I computed chipload for 10mils (maybe it isnt linear that low I dont know).

Working from that I determined the recommended speed and feed as:
1.2-2.08 IPM @ 7600-13k RPM

I was running at 1.2IPM but had the RPM way too low at 2600. using RPM of 2600 the IPM would have to be .4 max! I was running the tool at twice its rated IPM.

I am a little relieved to find my error. Still, I am leaning toward the vbit now after all your comments.

Colin

[Drools]

Would Aluminum be easier to mill? If I remember correctly when I lived in the big city the hobby shop there had sheet brass and sheet aluminum in the same thickness.

[guru_florida]

Originally Posted by Drools Would Aluminum be easier to mill? If I remember correctly when I lived in the big city the hobby shop there had sheet brass and sheet aluminum in the same thickness.

I am not sure about Brass, but Aluminum is very gummy. It may gum up an endmill and break it. I have it happen on larger endmills too when they start getting dull.

[Jay C]

There is an instructable where the guy uses the aluminum from a soda can. he is using the photoresist etch method, but that might be a cheaper starting point for raw material. I have tried milling a mask yet as I only do one-offs but Phil sure makes it look easy.