What PCB creation method would be best for

I am making a heated bed for a 3D printer. It’s basically a bunch of small copper runs in parallel across the width of the board. For your viewing pleasure, here is a close up.

To put it in perspective, the copper traces are about 1.8mm wide with about .25mm of seperation between them. The board itself is single-sided 1oz copper FR4 and 305mm x 305mm (12" x 12").

Given the above dimensions, what would be the best method to make this board?

  • laser printer toner mask transfer and etch?
  • laser mask and etch
  • pcb mill

Since the idea is to cover the desired copper, it doesn’t seem like a lot of work for the laser and I don’t know if the PCB mill can handle a board this size. The laser printer toner transfer will require some precision since I’ll have to print at least two sheets and line them up.

Another option is to just buy one already made but 1) they are close to $100, 2) I would have to depend on their design for output power.

I plan on powering the heater with a variac initially, and a SSR and PID initially for testing and eventually use a step-down transformer to somewhere around 60VAC. According to The Circuit Calculator, with the dimensions specified, I should easily get about 350W at 6A and 60V. Actually, if I account for the change in temperature from 25C to 200C, the power comes out to just over 200W at 60V and 3.6A.

But back to the original question, in this particular case, what would produce the best results for making this board?

Thanks for looking! All opinions welcomed and desired.

Joe

1 Like

Smallest endmill I’ve ever been able to use in the PCB mill (which is probably under 12"x12" working area, but the Shapeoko2 probably has a big enough bed) is 0.5mm. Since your spacing is 0.25mm, it would be tough to get it exactly as shown. An engraving bit might be able to do it, but you’d have to have the board super level to have a shot at it doing it properly.

This is probably better suited for etching with a sponge (due to size mainly). I’ve never had any luck with the spray paint and laser method personally, but others have had great success with that. I know @apipe and @Darcy_Neal have both had success with the laser approach (among others I’m sure). I’ve always used the transparency paper and iron approach. The alignment issue might make this easier on the laser.

FWIW, I think @HankCowdog once used strip board and soldered the rows together on alternating ends when attempting something similar. I have no idea how it stacks up to this approach as far as wattage or efficiency etc.

I found this article a couple weeks ago… Could be something to look into…

1 Like

i’m actually etching a board as i type. we used the vinyl cutter to cut out the traces and transferred that to the pcb board. the double sided boards i used the pcb mill to drill the holes to use as registration points, the single sided board i just applied to the board and will hand drill the holes with my dremel tool and flexshaft handle. its so much easier than any other method and produces beautiful traces. http://www.instructables.com/id/Fast-and-Easy-PCB-Prototyping-with-Vinyl/

2 Likes

I also etch inside of a ziplock freezer gallon bag. it is much cleaner and uses way less chemical than pouring it into an open dish. easier to agitate the formula without getting your hands dirty…

That would be the biggest challenge, leveling the bed. I know that the .25mm is almost 1/100th of an inch, which in theory, is well within the realm of the HAAS, but could you imagine trying to get that small of a bit in the HAAS? That type of tolerance is probably going to remove the Shapeoko and EMCO out of the equation. I guess I could try the black spray paint and laser method.

Update:

I am going to calculate another version of the board with 3mm traces with 1 mm of space between them and see what the total resistance comes out to. 1mm is going to be much more forgiving that .25mm.

Thanks for all the feedback! I like the idea about the vinyl. I actually might still have some I bought for a different project. Then I just need someone to show me how to use the vinyl cutter.

If you only need to do one ever, you could cut it with an exacto knife. Make sure to wear eye protection because the tips like to snap off when you’re cutting copper that way.

I make SMT test boards like that when I need one. It’s time-consuming, but nowhere near the time*cost that fabbing one would end up being.

Thanks Daniel,

Yes that would work too. I used build RC planes. I all too familiar with the perils of the Exacto. They excel at cutting through finger nails and skin as well as vinyl, foam, balsa, too many more items to list…

1 Like

consider the laser paint method with a cheap piece of of single sided PCB?

Draws a calculated 170 watts, IIRC. Works well and heats faster than the Prusa red commercial heated beds.

http://www.thingiverse.com/thing:12727 was my inspiration.

I used two 4"X8" pieces of Stripboard 4"x8" 3200hole Phenolic Pitch 0.1" made by Veroboard and purchased from Amazon. You want to make sure you use “clad” stripboard (as opposed to electroplated) so that the depth of the copper on the board is consistent across the traces.

I have had a lot of success with the Pulsar Toner Transfer system for PCB. I also have an etching tank that I use. Yes I know a lot of people use the sponge method but I haven’t had much success with that. I may just be too impatient. I design my circuit on the pc, take an inkjet copy to the local FedEx and copy it onto the Toner transfer paper. I have an old lamination machine I use to get a constant pressure heat transfer and it works great. If I really want a good PCB I also add the green foil to the traces which prevents the pitting but it is not usually necessary. Mouser and Digi Key both carry the Pulsar transfer paper.
There are a lot of ways to do this … Voltera has a kick started for a PCB printing machine with multi layer ability… I may yet break down and order one but they will not be available until next year. FYI

Don’t rule out the Emco. It is a German build machine and I don’t know how accurate it is but it beats the pants off ANY of the Chinese mill! Anyway, leveling the bed with a waste board would be a piece of cake… I would be more worried with the tolerance of the copper clad board if you are looking for 1/100th tolerance.

I have few questions for you. Do you take it up to FedEx (AKA Kinkos) because you don’t have a laser printer, or does their copier do a better job than a laser you might have? Did you have to modify your lamination machine? I ASSUME most wouldn’t accommodate a PCB thickness item? I really like the idea of using a lamination machine and I might start looking for one on CL or something.

We’ve discussed the Voltera printer before briefly here: Direct printed 2-layer PCBs
IMO, I think it’s weird that they call it “dual layer”. It seems like you couldn’t print the two layers with collinear traces for instance, and it’s just IMO a single layer with automatic jumper creation. It seems like there could be a lot of dual layer designs out there their software couldn’t really handle. Don’t get me wrong, it’s still very cool. Maybe they should call it 1.5 layer or something equally confusing. :smile:

As far as the Emco or PCB milling in general goes, I think the issue is more likely not a flat bed situation, but a slightly warped copper clad board.

The Toner Transfer system looks pretty cool but I’m doing a 12" by 12" board. It looks like their paper is 8.5" x 11". :frowning:

I would have to print it out across 4 sheets and then line them up on the board. I’m not saying it can’t be done, I just don’t know that I could do it.

As for the EMCO, I agree that the copper board flatness would probably be off more than the EMCO but there is just not a way to compensate for this. Normally you could surface the piece you are working on but that is probably not a good option with a copper clad board.

I’m going to swing by the space after work and try my best with the black paint and laser cutter. Anyone have a good strating point for settings on the raster? I’m guessing at 100% raster power and current but what about speed?

You nailed it! I don’t have a toner printer. I just stick with the inkjets so I just need a copier to make the transfer. At $.09 a copy a very economical solution. I bought the laminating machine on e-bay for about $20. It did not accommodate the PCBs … at first… There were some plastic tabs meant to guide the laminated item… probably ID badges… but after running the PCB through, they broke off so, problem solved. If you want to try the toner transfer I have paper and you can use my laminator. Just let me know. Glad to accommodate.
I ALMOST pulled the trigger and ordered a Voltera…Still thinking about it.

I’m definitely not the person to ask about the laser approach as I’ve never successfully been able to pull off the laser etch. I would love to try it again and get it to work some day, but I haven’t had a need yet.

If it was me, I guess I’d cut up some maybe 1" square pieces of PCB and try some various power settings and processes and see what works best. Kinda hard I suppose to batch out a lot of tests in raster compared to vector operation, but probably worth a few tries in any event. Getting an even coating of thick enough but not too thick spray paint is probably key as well.

Yeah, 9 cents a copy ain’t bad at all. I bought a really cheap Samsung printer one day at Fry’s for $49 if I recall correctly. I figured if transparency paper melted in it or something crazy like that, I wouldn’t cry too much :).

I have read that 3 - 4 thin coats should be adequate. I couldn’t get there last night but I may walk on over tonight after work now that I figured out how to make a negative of the PCB negative in Inkscape.

I redesigned the board so that it will create about 250 watts at about 50V and 5A.

I plan on using a PID and SSR to control the temps. In theory, at least according to the PCB Trace Resistance Calculato" and PCB Trace Width Calculator, the board could handle up to 17 amps with 2mm traces at 300 degrees celcius. But at 1/16" thick FR4, I don’t think I’ll be trying that…

Do you think that the results on those links are accurate enough for what I’m trying to do?

I guess I’ll find out. :wink: