[Recommendations]How much power do I need to laser cut copper

[blueskull]

I am looking for a laser solution to cut copper up to 4 mils. Any suggestions?
I do not care about wavelength, but for the sake of efficiency, no CO2 laser.
I do not have too much funding, so Q switched is not practical.

It has to be mounted on a robotic arm, so bulky solutions (cylindrical lens+diode bar) are not preferred, unless packaged in a sturdy, neat package.
Fiber solutions, if affordable, will be very appreciated. I see there are cheap up to 10W fiber collimators on eBay. Used devices are fine, eBay devices are fine.

Any suggestions?
Thanks in advance.

[rdl]

You might look for videos on youtube. I know there's one where a pcb is etched with a blue ray laser diode, but that's less than 1.5 mils.

[KL27x]

I do not care about wavelength, but for the sake of efficiency, no CO2 laser.

I think you need to care about wavelength for the sake of efficiency. And I believe a CO2 laser has the proper wavelength for cutting copper. And I think you'll need around 30W, at least. There's a reason we don't cut pcb's with lasers. Copper is highly thermally conductive and reflective. Good luck.

You might have a better success by putting the copper an oven to oxidize and preheat the copper and target the continually forming surface layer of copper oxide.

[matseng]

I think you need to care about wavelength for the sake of efficiency. And I believe a CO2 laser has the proper wavelength for cutting copper. And I think you'll need around 30W, at least. There's a reason we don't cut pcb's with lasers.

Ehhhh...  I was under the impression that copper is highly reflective in the 10um band, thus making a CO2 laser virtually useless for cutting copper unless "industrial" power ratings of the laser is used.

[KL27x]

Yes, you are right.    I'm wrong.

This topic comes up every few years on another forum I frequent, and misremembered like exactly wrong.

So maybe what I misremembered is that industrial power CO2 is still the most practical laser to cut copper, despite it being inefficient? 

A guy built a CO2 laser and posted some nice pics, and one of the things he wanted to do is cut copper. Thread kinda died years ago. Maybe he gave up.

[matseng]

I'd hazard a guess that you would need *slightly* more than 7 watts unless you want the fr4 to be vaporised as well due to the long exposure of the laser even if you got thru the copper eventually. Copper will just shuffle away the heat from the laser faster than you can apply it with a measly 7 watts.

The LPKF units for copper ablation from pcbs is pulling 1.2KW from mains and using a pulsed UV laser for zapping the copper.  I guess that the energy in the pulses are 2-3 orders of magnitude higher than the CW 7watt unit.

[NiHaoMike]

Could you consider EDM?

[IconicPCB]

LPKF have what ails You.

Consider a 10W IR source passing through a frequency tripler to come out a nice UV.

Now You can ablate copper using an acoustic effect. Namely score the copper in small chads and then hit the the chad with extra emergy causing localised heating and popping the chad off the substrate.

U NEED UV.

[nctnico]

I'd hazard a guess that you would need *slightly* more than 7 watts unless you want the fr4 to be vaporised as well due to the long exposure of the laser even if you got thru the copper eventually. Copper will just shuffle away the heat from the laser faster than you can apply it with a measly 7 watts.

The LPKF units for copper ablation from pcbs is pulling 1.2KW from mains and using a pulsed UV laser for zapping the copper.  I guess that the energy in the pulses are 2-3 orders of magnitude higher than the CW 7watt unit.

What about cutting ceramics?
I can etch copper, if cutting it is too hard.
How about ~1.6 mil thin ceramic?

My process is to plate copper on ceramic, then either laser cut copper, or just use H2O2/HCL to etch copper.
Since there is no chemical that etches ceramic, I definitely need laser to cut thin ceramic sheet.

I'd use a diamond saw blade for cutting ceramic (perhaps water cooled). The easiest way will be to have it made though (ceramic substrate with copper traces)!

[ajb]

The other challenge with laser cutting metal is thermal conductivity.  If you're trying to ablate away a thin line of material without melting the adjacent areas you need to put in a lot of heat very quickly to get the target area to the ablation point before you lose all of that energy to the surrounding area.  It's not unlike trying to solder to a big ground plane, if you have a low heat capacity soldering iron most of the energy gets conducted away by the plane before providing an appreciable temperature rise at the joint.  For efficient soldering and to avoid damaging the board you need to add heat much more rapidly in order to establish a steep thermal gradient with the joint at soldering temperature while the surrounding area stays relatively cool.

I've experimented a bit with cutting stainless steel shim stock (2-5mil IIRC) on our 35W CO2 Epilog, and while I can punch through it, there's a lot of distortion of the edges of the cut.  I'd guess you'd need more like 100+W to get a clean cut with the stainless, and assuming the same absorption you'd need even more power to cut copper cleanly due to the higher thermal conductivity. 

As far as ceramic, I'd be worried about the material cracking.  Even if it's thin enough to be highly flexible, blasting away bits of it with a laser is a pretty violent process at the microscopic level, and materials that have a bit of elasticity will tolerate that better.  It will probably take some trial and error to find a power level that works.  You could probably send a sample of material to Epilog and have them test it for you on either a CO2 or fiber system.  I've seen people bring them oddball materials at trade shows and they're happy to try running them on their demo machines.

[LaserSteve]

One, you could just send it out to the local laser  job shop.

Two, Your looking at Q-Switched YAG or a Q-switched  Fiber Laser.   Co2 can be used to cut copper, but the power required is immense.  Copper is cut by ablation or sublimation cutting.  The goal is to focus just above the copper and start a micro plasma.  Assist gas is needed to clear the kerf left by the laser.   Pulsed Green light from a frequency doubled laser excels at cutting thin copper, silver, and gold films.  In fact LCD plants, ceramic module makers, wafer repair, and resistor trimmers used to use the pulsed green light from a Xenon Ion laser before doubled YAG became popular.


I used to work on industrial and scientific pulsed YAG lasers at rep rates from 0.5 HZ to tens of Kilohertz, with pulse widths from macro to picosecond. Those drove a variety of wavelength conversion means, so I've worked from 230 in the UV to 15 microns in the far IR.  I've also used a fair amount of Co2 systems to a Kilowatt.  My personal record is 52 Joules 1064 and 32 Joules green in a single pulse, unless you want to count servicing 129 flashlamp power supplies for a Terrawatt system.  I've been at this laser game for 20+ years.

  Pulsed UV is difficult to generate.  The 355 nm CW UV is even more difficult to generate, as its done using sum frequency mixing from 1064 and 532.  If you buy a cheap UV YAG, You'll spend more time adjusting doubling and tripling crystals then you will cutting metal, unless you have a decent budget.  Stay away from 266 nm lasers, trust me... They are awesome for what you want to do, but the operational cost is brutal.

I doubt you could ever achieve a decent cut on thin metal using a diode laser array that is less one hundred to two hundred watts. The problem then becomes the spot size from the diode array's fiber, which will be huge. .  At Co2 your looking at ~ 100-200 watts or more for a thin kerf in a few mills of copper.

A 400 watt Firestar CO2  I played with could barely do a clean cut on a Coke can or thin stainless sheet.  In fact we painted aluminum sheet with a thin layer of spray enamel just to get the cut started cleanly. (carbon based black engine paint)   Yet a few tens of watts of average power from a 532 nM doubled yag went thru the coke can at any gantry speed desired.

  I have a 60 watt,7-10 nanosecond, 7 KHz Green system installed near here for perforating brass and stainless steel sheets for limited production runs of lithium batteries. It works so well that I get a free lunch every time I drive by that shop.  They run it at 25-35 watts, typically. It also has really good up-collimation and spatial filtering optics followed by a lens that insures a tiny spot size.

Ablation or sublimation cutting is dependent on peak power.  You need a pulsed laser. Think of a pulsed laser more as a very fast discharging capacitor then a tiny cutting torch. 



Steve

[Fungus]

It has to be mounted on a robotic arm, so bulky solutions (cylindrical lens+diode bar) are not preferred, unless packaged in a sturdy, neat package.

...or fix the laser to the floor and have the robot hold the copper up to the beam.

( Wouldn't that be way cooler to watch than the other way around?  )

One, you could just send it out to the local laser  job shop.

If it's not an ongoing thing then this option makes a lot of sense. Cutting highly conductive metals needs serious laser power (=expensive).

[LaserSteve]

Of Course if your doing this industrially and have budget, SP, Coherent, and Ekspla have some nice green and UV solutions...

Example:

http://www.spectra-physics.com/applications/pcb-manufacturing#resources

http://www.spectra-physics.com/documents/ShortPulseWidth.pdf

Steve

[LaserSteve]

I see your user flag indicates China..

If you build or rebuild it, yes...

Given what Chinese DPSS laser parts sell for in America, probably yes, with some judicious Ebay/AliBaba searching.

China has been producing lamp pumped ND:YAG as well as importing it since the 1970s... There are a couple of Chinese Ebay sellers that make new parts for common lamp pumped systems.

I'l give you an example.  1000 meters from me, in a warehouse, are two 100 watt Lee Laser lamp pumped, Q-Switched lasers removed from service in working order.  They even have the fiber couplers and original crates.  My employer's surplus disposal office  offered them to me for 1000$ for the pair..  I turned them down, for lack of storage space.

Why? No one on campus wants reuse them, because of the three phase power and lack of hardware skills. That and the fact that each system, complete with chiller, weights 340 LBS/154 KG. On a campus of 24,000 people, right now, three of us have the skills to get them running without help and there are perhaps another 20 who could be trained.   So the hardware is out there, but the trained people move on. When they do, the lasers are nearly always disposed of.


 Which is sad, because for an EE or EE student, such  a laser is just an exotic oscillator that is very easy to repair and get going on a modest budget.  I mean what EE hobbyist would pass up a 7 kilowatt arc lamp that runs submerged with bare terminals in deionized water?  (Edit, Most people  should, unless they have some safety training, these are not toys!   )


So my bet is that if you do some digging, a crusty old but repairable  Lamp Pumped system is available near by, either a medical unit or industrial.

If you buy a lamp pumped laser, open the pump cavity and make sure the gold plating is intact and not corroded.

DPSS lasers tend to be disposed of when the pump diodes degrade, and often those are a custom item.

I'll give you another example,  I have a 800 mW, Q-Switched Green laser head on my desk that I paid 300$ for. It needs a 8 amp constant current power supply with NO voltage or current overshoot to run the pump diode.  It needs two TE coolers held to within ~ 1.25' C, which is easy.  It needs 4 watts of 80 Mhz RF to drive the Q-switch and probably produces 100-200 kW peak pulses at few kilohertz. or so. 

   Its someone's prototype.  Its milled from a monolithic aluminum block, and the design is very simple.  I'm half way thru reverse engineering it.  Other then some very good mechanical engineering, the design is trivial and published.
Other then the fact that none of the supporting electronics was included, its working just fine on my bench.

I'll see if I can post a picture of the insides by tomorrow,

This is do-able if you have time to find the surplus parts.

There is a hobbyist laser conference every year in the US, focused on laser shows.  Every year I train 5-10 people at the conference to align laser oscillator mirrors.  Most people get the technique within one to two hours.  EE students/good technicians  who understand simple rate equations and basic geometry usually can lean to adjust the laser to peak performance in another hour or two.  Its not difficult.

Figure 100-200$ for a good pair of safety goggles.  You'll also need 100% beam containment.  These are not toys, and the safety procedures is where much of the expense in lab use comes from.

Steve

[LaserSteve]

Oh, if your in the States, getting you a used ND:YAG laser is EASY!

Steve

[T3sl4co1l]

My process is to plate copper on ceramic, then either laser cut copper, or just use H2O2/HCL to etch copper.
Since there is no chemical that etches ceramic, I definitely need laser to cut thin ceramic sheet.

Gnaw, tons of things.  Nothing is impervious*!

(*Teflon comes damn close, but of course burns at relatively modest temperatures, and can be attacked by strong bases.  For everything else, there's always a stronger superacid.)

There may even be chemicals that etch the ceramic without touching the copper.  As I recall, copper is resistant to fluorine, though not fluoride so that might be rather difficult.  The metal can be coated with something else resistant, temporarily; which is done with normal PCBs (tin flash protects during an etch step).

Most likely candidates being hydrofluoric acid and relatives, for most anything on the aluminosilicate spectrum.

Tim

[timb]

I physically locate in the US, specifically North Carolina, which is bad because unlike CA or NY, here it is not easy to get cheap or free dumpster dived goodies.

I think I need to learn more about laser, when I have time. Anyway, laser is just an easy way to get my research jobs done, not anything essential.

I have traditional ways to do it, it just take more time and paper works since we have to borrow the other labs' gears.

You can get some pretty rad gear dumpster diving around the Research Triangle. Also check Craigslist (and put an ad up saying what you're after). I've had really good luck doing that.

If you don't mind a short drive up to Virginia, NASA Langley auctions stuff off every so often. There's also DRMO auctions in Norfolk and Richmond.

[rickey1990]

my plasma cutter comment is officially deleted. Americans and there imperial measuring system, I fort you meant 4 millimetres

[Fungus]

Most likely candidates being hydrofluoric acid and relatives, for most anything on the aluminosilicate spectrum.

I'd pay to get it cut professionally looong before I'd start messing about with hydrofluoric acid.

"Bottled Evil" doesn't come close to describing that stuff.

[timb]

You can get some pretty rad gear dumpster diving around the Research Triangle. Also check Craigslist (and put an ad up saying what you're after). I've had really good luck doing that.

If you don't mind a short drive up to Virginia, NASA Langley auctions stuff off every so often. There's also DRMO auctions in Norfolk and Richmond.

I'll try my luck there and take an eye on RTP news. It is hard to check too often since I am still ~40 miles away from there.

If you send me a PM with your email and maybe a list of stuff you're after, I'll keep my eyes peeled for you.

I actually came across a laser cutter at a DRMO auction last year. It went for like $100. Looked like a nice unit, too. Unfortunately I didn't have anywhere to store it at the time. (I occasionally see some high end laser stuff, but I don't know much about it, so I don't feel like I can fairly asses it for resale.)

[LaserSteve]

Just send it out to someone like Potomac Photonics...  Its a precision laser job shop near DC.

There are a couple of used laser dealers in the US.  I'll send you a PM in the morning.  I don't feel like broadcasting my sources world wide.

Edit, they changed their name:

http://www.potomac-laser.com/

Last time I needed something for academic work, I told them it was a student project, the cost suddenly  became very reasonable.


Steve

[LaserSteve]

Potomac is a huge company that signs NDAs.  In the US, violating your NDA, especially one tied to academic research, is corporate financial suicide. Given a large portion of their work is secret defense work or a medical device, they will have insurance and procedures against such disclosures. Look who their client list is. They did not get that big by violating an NDA. 

You don't have to tell them what they are working on. You would have to inform them if the substrate or film  is hazardous like Beryllium Oxide or similar.  The employee who runs the laser cutter need not even know what he is working on, just any standard procedures for protecting your device.

Your process will not end up on Facebook. 

Steve

[brainstorm]

@matseng, @iconicpcb or others that mentioned LPKF:

https://youtu.be/PhH05jNyjCk?t=33

I wonder which type of UV lasers/power/pulsing do they use and if they could be sourced somehow in 2019?

[IconicPCB]

That laser is great for processing RF structures...otherwise it is of questionable utility.

The video shows the laser structuring component and hole pads.

In real life holes are drilled first, then they are activated and the laminate is then plated to deposit copper onto the hole walls.

Etching the copper off prior to hole wall activation is utterly impractical.

If however the hole walls are activated and holes plated then the copper on top and bottom surface becomes thicker possibly making it difficult to remove thicker copper.

Say a 2 ounce board is to be fabricated for a power electronics application.... I dont think so.