Experimental results from laser etching masks for copper etching

[Dustin Andrews]

(tldr: 1. Cermark makes great marks on copper boards meant for circuit boards. 2. Cermark is dissolved by the same chemicals that dissolve copper, so it doesn't make a great etching mask.)

Hi! I've been lurking around the board for a while. I have learned so much from this forum. Thanks to everyone who has posted such useful information.

I've been having a lot of fun using my laser to make geared toys for my family and enclosures for electronics. I prototype a lot of circuit boards and I get good results with the toner transfer method, but the laser seems to have a lot of potential. After looking around the internet and this forum I decided I would do some experimentation and see if I could come up with a workable process. If I can get the laser working it will cut my prototyping time in half or less.

In order to be useful I need to be able to etch traces that are 8 mil wide with 10 mill clearances between (6 would be ideal). I can (just) do 10/6 with a laser printer and toner transfer.

The ideal process would use the laser to etch the mask in one step. I've been experimenting with Kapton tape. The results are promising, but I haven't found a setting that will vaporize all the tape without destroying the fine details. I suspect I might be able to dial it in with some more work. It might take two or three passes at different power to get it to work.

This week I got a bottle of Cermark and experimented with it on 1/2oz copper clad fiberglass (aka PCB). Here is what I have learned so far.

Ferric Chloride Rub.jpg

On the left is the Cermark after coming out of the laser. I superimposed my speed and power notes in the image.
I started the test at 100% power and slow speeds. The Cermark got hot enough to burn the copper right off the board at really low speeds. My 40 watt laser can't mark the copper at this speed alone. The blobs at the bottom are where the burning was worst. Then I used a sponge with some Ferric Chlorid and rubbed the board gently until all the copper between the tests was gone.

The results were mixed. On the plus side the Cermark made a good clean mark on the board that I couldn't scratch off easily. It also worked somewhat decently as a mask. The Ferric Chloride actually removed the Cermark as well as the copper. I was hoping that it wouldn't remove it at all. Still, the copper pattern was preserved decently.

This first pass showed me that I was going to hot. So I did another pass (on the back side of the same board) with lower settings.

Peroxide soak.jpg

In this image you can see the Cermark made a really good mark on the copper at 1000 DPI, Speed 100% Power 50-70%. The top is before rinsing off the Cermark. The middle shows after rinsing and light scrubbing with a soft sponge. The bottom is after etching. This time I switch to etching with Muratic Acid + Hydrogen Peroxide. It's a bit gentler etching solution and I soaked instead of rubbed.

Results: I realized I should have been using a negative image so this board is the inverse of the one above. The detail in the etching is not quite good enough, but the marks are once again excellently bonded. The wires are just touching in places where they should be separate. Once again the Cermark dissolved in the etchant, but not before protecting the copper long enough to get a halfway decent etch.

70% worked the best, so I ran one more test bracketing that setting. I also changed to the lowest DPI setting to see if that would have any affect.
Peroxide Soak 2.jpg

Results: 250 DPI was no better or worse than 1000 and a three times faster. All of the marks at these speeds and powers were great. The Etchant once again ate into the Cermark a little too fast. The etches all look different, but the differences (IMHO) are mostly due to the way the etchant works as you gently agitate it. None of the traces came out usable.

Cermark may yet be usable. I have some "trf" film that can be used to seal toner in the toner transfer process. Most likely it will stick like a champ to Cermark. However, it's "blobby" and can't do the resolution I need for my smaller designs. I may give it a try and post some results just to see if it will stick to the Cermark for use in decorative etching.

[Rich Harman]

I have had good luck using regular old spray paint. The etchant does not bother it at all and afterwards it wipes right off with acetone.

[Dustin Andrews]

I saw that various people were using spray paint. The drying time seems to put it in about the same time-frame as warming up my laminator. Still, maybe with a jig I could get good double sided registration easier than toner transfer. How small are the traces and clearance you have done Rich?

[Rich Harman]

I haven't pushed the limits for clearances. When I need something really precise I just order some boards from Advanced Circuits. However, I will soon be attempting a board that includes a QFN and other fine pitch SMT components.

As for drying time, I use paint that dries in just a couple minutes.

[Dustin Andrews]

Could you be specific about the paint?

I'm looking to do some QFN prototypes soon I think. I just did some boards with SSOP and they all came out fine with Toner Transfer using parchment paper. I always order prototype boards from a similar place before I go to production, but I want to cut the expense and turnaround for more experimental stuff.

[George Carlson]

Something I've been intending to try is nitro-celulose laquer as a resist. It should burn away with very little power and leave no residue. Acrylic laquer may work as well.
Since I found I can mill circuit boards on my CNC router, I haven't considered messing with etchant. I have one of those Chinese 6040 machines. It has high quality linear ball slides and ball screws on all axis. I did some rebuilding of the electronics on it, but the mechanics were very good.

[Chuck Stone]

I saw that various people were using spray paint. The drying time seems to put it in about the same time-frame as warming up my laminator. Still, maybe with a jig I could get good double sided registration easier than toner transfer.

If you trace the outline of your board and etch that onto a piece of card stock, registration is pretty easy.
I'll lightly burn a pattern into card stock and then place my object on it for marking. Obviously that doesn't
work for cutting, but for engraving it works great. I have done two sides stuff, I just trace the outline in
Corel, copy and paste the outline further down the page and flip it (H or V, doesn't matter) and burn this
into the card stock. Save the file and save the burned card stock as a template for future use.

When you go to engrave, use the template for placement. Open the file and add your markings where
you want them and then delete the edge. (leaving only what you want to burn) Then Undo to get the
edge back and put your markings on the other side.

I do several parts that are the same except for text. This method makes it very easy for me, and most
important is that it is absolutely repeatable positioning. The only time I screw up is when I keep importing
the text from another file while I'm engraving .. you get going so fast that you can lose track of where you
are.. and delete the raster part that you just placed there .. before you've even engraved it. D'OH!!
you have placed the item

[Rich Harman]

I have not tested various paints. I am sure there is something better than the black enamel that I have been using.

I have also done isolation milling but I find that etching is quicker. If I had a spindle that wasn't limited to 3,000 rpm I would probably reverse that opinion.

[Rodne Gold]

Search my posts for "etching" , we do a lot of it an can get extremely fine detail with the laser and masks. We etch chemically and electrically , electrically would be better IMHO for even finer detail as it does not undercut.

[Mark Ross]

Try using ammonium persulfate for the etchant. We used to use it in a corngin ware dish on a hot plate. Etch times were only a few minutes.

[Richard Rumancik]

This is an interesting topic. Personally, I don't think that Cermark is the way to go. It is not a contiguous coating and does not work well as a resist.

I know that one other member recommended nitrocellulose lacquer as a resist for etching but I can't find the thread. There are many references on the web for people using "black spray paint" but many are not very specific. The brand and formulation are critical for good results and you can't just use any random spray paint and get good results. Some are better than others but I don't know if anyone has found the "perfect" spray resist. The perfect resist bonds well, but lasers cleanly away with zero residue. Let me know if you have found it . . .

I found an interesting article recently on the topic "Rapid Prototyping of Printed Circuit Boards Using an Engraving Laser" by ASEE authors Jeffrey Mountain and David Beams. It is hard to locate using a Google search but you can find it this way:

1. go to http://search.asee.org
2. click on "conference papers"
3. in the search box enter Rapid Prototyping Printed Circuit Boards

you should find the paper at the top of the list. It is paper # 2006-2097. Worth a read if you are interested in making boards with a laser.

They use "acrylic clearcoat model paint" as a resist. Unfortunately it does not indicate the manufacturer name.

[Richard Rumancik]

I forgot to mention that Kapton tape won't work really well as a masking film because the normal 10.6 micron wavelength isn't suitable for Kapton. You will get a lot of charring and edge raise. Manufacturers of Kapton film products use a modified co2 laser that outputs at 9.3-9.4 microns which I understand cuts very cleanly.

I've been wanting to try to laser dry film photoresist and see what it does but haven't got around to it. (This process would just use it as though it were a film and ablating it, not using it's photoresist properties.) I'd be interested if anyone has tried it.

[Glen Monaghan]

They use "acrylic clearcoat model paint" as a resist. Unfortunately it does not indicate the manufacturer name.

The professor's name is David Beam. From a google search:
Phone: (903) 565-5587
Email:David_Beams@UTTyler.edu

-Glen

[Rodne Gold]

There is a way to clear any residue from a paint coated board - it's low tech but works , and that is to use a household abraisive cream type cleaner "handy Andy"? on your fingertips and massage it in and wash off. Works for even the finest line the laser can resolve.....if the paint comes off when you doing that , it would have come off in the etchant anyway ,, so you saved a board...
The key to achieving success with thin or fragile paint resists is the time its in the resist. The longer its in , the likelier its gonna lift , so you have to speed up the etch time by heating/spray etching or using alternative etchants.

[matthew knott]

Interesting how many people make their own PCB's. I would have though that using the laser to remove a mask was more trouble than its worth being as a UV light box is cheap and the old method of laser printing onto acetate sheet and using pre-resisted copper would be much easier and get better results. On the other hand we use our fibre laser to directly produce pcbs as we can etch the copper away with the laser and leave the fibreglass untouched, is saves on the etching, i just etch around the pads and tracks (about .5mm) and you get a great top ground plane so good for EMC, from gerber file to finished board 150x100 takes about 10 minutes and no nasty chemicals, totally digital method.