Tried a different way of making PCBs today ....

[picbits]

I normally etch my own PCBs using Photoresist PCBs, an Inkjet printer and some inkjet transparency. This works brilliantly apart from having to wait for hours for the ink to dry (otherwise it sticks to the PCB). I can speed up the process by heating the film but it still takes its time.

I won a pack of Colour Laser transparencies off Ebay some time ago and have been meaning to try them with the photoresist PCBs but having printed off a sheet a few weeks ago didn't feel they would be up to it as the density was touch and go.

I had a batch of circuit boards to make today so cut off a spare bit, knocked up a test print on the laser printer and exposed the board for 1, 2, 3, 4 and 5 minutes with a bit of card used as shielding over the unexposed areas.

I have to admit the results weren't as bad as I expected them to be. The boards are certainly good enough for prototypes.

I can now get a board from the CAD/PCB design package to fully etched in 20 minutes

Heres a picture of the finished board - the transparencies used were Viking Colour Laser films on a Samsumg CLP-300 colour laser. The films were £5 delivered off Ebay and the printer was only £99.

 

[picbits]

Forgot to say, the track widths starting from the top were 0.08mm, 0.35mm, 0.7mm, 1.4mm, 2.8mm and 5.6mm

The 0.08mm is pretty solid but I wouldn't go that thin as its started to underetch. The 0.35mm is very solid and properly etched.

 

[3v0]

I have to admit the results weren't as bad as I expected them to be. The boards are certainly good enough for prototypes.

The line tests came out quite well. What was it that makes the process only good enough for prototypes? How much better is your regular process using the inkjet?

In general I do not go below .01 (.25mm) using toner transfer.

 

[picbits]

I find the inkjet films give a better sharper image on the PCB. The ink is denser than the toner an a little more forgiving than the toner method.

Saying that though, the majority of stuff I make uses .35mm or above traces so I might try using the toner/transparency method for some of the future product. Its also more robust - less chance of smearing or the positives getting damaged in storage.

Ideally at some point I want to try using an old DVD/RW laser to burn the image onto the photoresist. I have a 90% built CNC router which would be ideal for the task. The idea is to use isolating routing via the laser but I don't know how well the photoresist will react to the laser - its designed to respond to the other end of the light spectrum.

The laser/transparency method is certainly a cheap easy quick way of doing this though.

Not quite as cheap as press'n'peel but lets not get into that argument

 

[3v0]

Neat stuff if you can get the laser to work. Taking the transfer/exposure step out of the process would be great.

I am in favor of any process that produces the needed results. That is providing that it is a consistant process that most anyone can make work.

 

[picbits]

At some point I will be needing to make up large quantities of boards (around 100 per week). I've been toying with the idea of screenprinting them with etch resist in batches. I've got a load of screenprinting gear but no suitable inks as yet or chemicals to make the screens up.

That would be ultra cheap as standard copper clad board is around a third of the price of sensitised board.

 

[Pommie]

Have you checked out the price you can get boards made for? Just go to **broken link removed** and get a quote. There is no way you can make 100 boards per week and it not be cheaper to get them professionally made.

Mike.

 

[jpanhalt]

Ideally at some point I want to try using an old DVD/RW laser to burn the image onto the photoresist. I have a 90% built CNC router which would be ideal for the task. The idea is to use isolating routing via the laser but I don't know how well the photoresist will react to the laser - its designed to respond to the other end of the light spectrum.

With light at the other end of the spectrum (do you mean IR?), you probably won't get the photochemical reaction that is typical for the resist, but you may simply vaporize the resist so it will look like it works. If that's the case, you could consider using a clear acrylic or a solder-through conformal coating as your resist instead of the expensive photoresist.

John

 

[picbits]

I'll have to weigh up my options. The problem is the boards will be relatively small, may vary in design so I don't want to stock too many of one kind and around 50% of them will need to be custom cut to the customers requirements.

From a standard 100mm x 160mm board I can get 4 custom boards cut and at a cost price of £0.80 per board that makes a cost price of around £0.35-£0.40 per finished etched and drilled board.

I'm building the CNC router so I can do the custom cutting and drilling will also be done by this machine.

At the moment I'm making everything by hand. Exposing, developing, etching, cutting, drilling and soldering the boards by hand and can just about hit 100 boards a week. Its seriously hard work though and I have to work in a confined space. Being able to "outsource" a lot of the work to my CNC machine will make a big difference.

If I was making 100 of exactly the same boards a week then I'd wholeheartedly agree with you for getting the boards outsourced but part of the novelty of my products are that the customer can specify to some extent the design and layout of the board (display boards).

I've banged the numbers into PCBCart and they are pretty good - works out at around $1.55 per board for 100 boards. Add shipping, import duty, fees for the courier to collect the duty and you're looking at probably around $2.06 per board which is around 2.5 times the cost of making them myself.

On the plus side you dont have all the hassle and boredom of making them yourself. Its good fun as a hobby but when you start doing it 5 days a week it gets very tedious.

 

[3v0]

I tried screen printing boards back in the Z80 days. On a 4" by 10" board there was nearly always one spot where the ink was not correct. We were using ink made for PCB work.

I am not saying the process does not work. If things have not changed I am not sure if it is worth perfecting unless you intend to open a PCB house.

 

[picbits]

With light at the other end of the spectrum (do you mean IR?), you probably won't get the photochemical reaction that is typical for the resist, but you may simply vaporize the resist so it will look like it works. If that's the case, you could consider using a clear acrylic or a solder-through conformal coating as your resist instead of the expensive photoresist.

John

I'm pretty sure DVD lasers are IR. You can pick up new drives here in the UK for around $30 (£15) with a reasonably high powered laser inside.

Clear acrylic probably wouldn't work as you'd get the laser reflecting off the copper underneath so some kind of conformal coating that could be burned by the laser would be good. The photoresist is thin enough to possibly be burned or chemically altered enough to work but yet again - its something that would have to be played with.

I'm sure there would be something out there which would react in the right manner