Problems making PCBs

[Chryseus]

Well I've made several attempts now at making a PCB and all have ended in failure, although my last attempt was a lot better than the first.
The problem I'm having is that my developer (NaOH) removes or damages isolated traces and takes quite a bit longer with more tightly packed traces / pads resulting in even worse damage.
I've tried various exposure times from 1 to 7 minutes using a UV LED array, 1 minute 30 seconds seems to give the best results so far, as for media I'm using a 90gsm tracing paper doubled up which was recommended in a tutorial I found.
As for the developer I've tried various concentrations from super strong to very weak, aside from the greater control with a weaker solution this has not helped.
Apparently adding sodium metasilicate to NaOH can prevent over etching but I'm unsure of the correct concentration.

Any advice would be greatly appreciated. 

Oh and the trace size I'm using is 13mil

[RCMR]

The only etchants I've had any success with are ferric chloride and amonium persulphate.

I think the problem with the acid/base etchants is that they produce H2 as part of the chemical process and the bubbles of H2 cause the resist to lift along the exposed edges --with resultant undercutting and damage.

From memory, the good etchants perform an exchange of ions that produce no gassing.

[Chryseus]

You're confusing developer with etchant.
I'm using a photoresist so you have to develop it with NaOH or similar developer before etching with ferric chloride.

Etching is not my issue, developing is.

[sonicj]

i've never heard of tracing paper for exposing circuit boards....

lots of info available on the toner transfer method.
-sj

[flolic]

For photoresist developer, NaOH is very tricky to use because results very much depends of concentration and temperature. Few more seconds or few degrees higher temperature and all your effort is gone...

I strongly suggest DP-50 developer, it is practically foolproof and work extremely well. You can get it of ebay, or google for supplier.

[mikeselectricstuff]

Tracing paper in a laser works very well, better than transparency sheets and cheaper.

NaOH sucks as a developer - get some silicate based stuff like this .

The difference is that with silicate, there is a huge margin between 'develop' and 'strip', in terms of time, temperaturi and concentration, and is way more forgiving than NaOH. Another advantage is that made-up solution has indefinite shelflive - NaOH goes funny after a while.

Can't comment on using LEDs for exposure, but you can get insect killer tubes pretty cheaply, so  I don't see the point in using LEDs.

See My PCB page for more info on making PCBs.

[amspire]

Did you print the artwork reversed on the tracing paper so that the actual printed ink is against the PCB?

Also what sort of printer are you using? If it is a inkjet printer using a dye based black, it may not be black for UV. Laser printer toner and carbon-based black inject inks will definitely be fine.

Richard.

[Chryseus]

Yes I put the artwork directly against the board and I'm using a laser printer.

I probably should get a silicate based developer although I do have a big pot of sodium metasilicate sitting here that should do the job, I'll have to experiment with it and see how it goes, that developer you linked to Mike is very cheap so if it does not work out I'll order that.

As for UV lamps I though they needed a ballast, I'm not exactly sure how to pick the correct type of ballast, does it need the same power rating as the tube ?
LEDs seem to work but they're not perfect they seem to leave small areas that are less well exposed.

Thanks for the replies guys, I'll post an update on how it goes later today. 

[amspire]

Your problem may be that the LEDs are just too close to the artwork, and light is able to get in around the edges.

It may be worth testing your resist and developer  using the sun as the UV source. Put a test strip and artwork out in the sun and gradually uncover strips at timed intervals. On a sunny day, you do get a even exposure of parallel rays.

Even with the developer you are using, I think there should be enough difference between the exposed and unexposed resist to develop easily. There is usually a pretty big margin of error for a properly exposed board.

How are you clamping the artwork to the board? Do you have a glass sheet?

Are you using a positive or negative photo resist on the PCB?

Richard.

[Chryseus]

Yeah the LEDs are pretty close to the board, about 5-7cm, I've noticed electrical tape gives a nice sharp edge that resists the developer very well.
At the moment I'm using an old scanner without the top, I tape one side of the artwork to the board and lay it directly over the LEDs (The LEDs are inside the scanner) then put a heavy object on top of the board.

I'm using Positiv 20 (postive) spray on photoresist which I dry with a hair dryer for about 5 minutes, it's pretty even when done so there is no real problem with uneven resist.
I'll try making the LED light more diffuse and use some black paper to prevent the UV being reflected, if that doesn't help I'll give sunlight a go.

[mikeselectricstuff]

I'm using Positiv 20 (postive) spray on photoresist which I dry with a hair dryer for about 5 minutes, it's pretty even when done so there is no real problem with uneven resist.

There's your problem.
Spray photoresist is a complete waste of time. use pre-coated boards.

[Chryseus]

I was hoping to avoid having to buy sensitized boards since they're more expensive but I guess you're right, the spray would probably be fine for thicker traces, I know some people have had good results with it.
Oh well I guess I should go make an order at Rapid, thanks for your help. 

[kripton2035]

if you have already simple pcb without photoresist, why not try the toner transfert method ?
really easy if you have a laser printer
I had your naoh method, failed too difficult, I used also the photoresist spray total waste of time and money...
now I use toner transfert, and I will one day use a pcb company to make the pcbI cannot make myself
too thin or too many same boards

[amspire]

It is decades since I tried putting a DIY resist on a board, but I seem to remember that negative resist worked far better then positive resist. My memory is that for negative resist, the exposed resist hardened pretty successfully.

Back then, it was all through hole components - thick tracks, big pads.

Richard.

[Madsaaby]

Yeah the LEDs are pretty close to the board, about 5-7cm, I've noticed electrical tape gives a nice sharp edge that resists the developer very well.
At the moment I'm using an old scanner without the top, I tape one side of the artwork to the board and lay it directly over the LEDs (The LEDs are inside the scanner) then put a heavy object on top of the board.

I'm using Positiv 20 (postive) spray on photoresist which I dry with a hair dryer for about 5 minutes, it's pretty even when done so there is no real problem with uneven resist.
I'll try making the LED light more diffuse and use some black paper to prevent the UV being reflected, if that doesn't help I'll give sunlight a go.

Hi

Maybe I can help a bit
I use the exact same technique, with the Posetiv 20 and the LED's.
AND I had the same problem in the beginning..

I think your problem is the hair-dryer, when blowing hot air on the PCB, the resist flows to the edges, making the layer very very thin.
Try instead placing the PCB on a thin plate of wood (or enything else ;]), and heat it from the bottom with an iron to around 70c for about 10 min, after you sprayed it. (This is how I do)


I use 12g NaOH to 1 l water..
My LED array has a spacing of 7 mm centre to centre of the led's.
And I use an exposure time of 3 min, with a distance of 20 cm from the lamp to the PCB.


Find the right layer-thickness, and your good 95% of the time..

Hope this helps..
If you have any questions, please ask

-Mads

[Chryseus]

Yeah the LEDs are pretty close to the board, about 5-7cm, I've noticed electrical tape gives a nice sharp edge that resists the developer very well.
At the moment I'm using an old scanner without the top, I tape one side of the artwork to the board and lay it directly over the LEDs (The LEDs are inside the scanner) then put a heavy object on top of the board.

I'm using Positiv 20 (postive) spray on photoresist which I dry with a hair dryer for about 5 minutes, it's pretty even when done so there is no real problem with uneven resist.
I'll try making the LED light more diffuse and use some black paper to prevent the UV being reflected, if that doesn't help I'll give sunlight a go.

Hi

Maybe I can help a bit
I use the exact same technique, with the Posetiv 20 and the LED's.
AND I had the same problem in the beginning..

I think your problem is the hair-dryer, when blowing hot air on the PCB, the resist flows to the edges, making the layer very very thin.
Try instead placing the PCB on a thin plate of wood (or enything else ;]), and heat it from the bottom with an iron to around 70c for about 10 min, after you sprayed it. (This is how I do)


I use 12g NaOH to 1 l water..
My LED array has a spacing of 7 mm centre to centre of the led's.
And I use an exposure time of 3 min, with a distance of 20 cm from the lamp to the PCB.


Find the right layer-thickness, and your good 95% of the time..

Hope this helps..
If you have any questions, please ask

-Mads

Thanks for the advice I'll try put it in the oven with a temperature probe.

[Madsaaby]

Yeah the LEDs are pretty close to the board, about 5-7cm, I've noticed electrical tape gives a nice sharp edge that resists the developer very well.
At the moment I'm using an old scanner without the top, I tape one side of the artwork to the board and lay it directly over the LEDs (The LEDs are inside the scanner) then put a heavy object on top of the board.

I'm using Positiv 20 (postive) spray on photoresist which I dry with a hair dryer for about 5 minutes, it's pretty even when done so there is no real problem with uneven resist.
I'll try making the LED light more diffuse and use some black paper to prevent the UV being reflected, if that doesn't help I'll give sunlight a go.

Hi

Maybe I can help a bit
I use the exact same technique, with the Posetiv 20 and the LED's.
AND I had the same problem in the beginning..

I think your problem is the hair-dryer, when blowing hot air on the PCB, the resist flows to the edges, making the layer very very thin.
Try instead placing the PCB on a thin plate of wood (or enything else ;]), and heat it from the bottom with an iron to around 70c for about 10 min, after you sprayed it. (This is how I do)


I use 12g NaOH to 1 l water..
My LED array has a spacing of 7 mm centre to centre of the led's.
And I use an exposure time of 3 min, with a distance of 20 cm from the lamp to the PCB.


Find the right layer-thickness, and your good 95% of the time..

Hope this helps..
If you have any questions, please ask

-Mads

Thanks for the advice I'll try put it in the oven with a temperature probe.

Yup, works just fine.. Hehe, But "someone" wont let me dry photoresist in the house, because of the smell.. :/

[Chryseus]

Just a quick update, I've found spinning the PCB on a PC fan or other motor works really well and gives consistent results every time, down to 10mil and below.
So my process so far is:

• Print layout onto Gateway 90gsm tracing paper (double up if needed)
• Cut PCB to desired size
• Clean thoroughly with household cream cleaner and a sponge, dry with paper towel
• Attach to a low speed PC fan with double sided tape (I've noticed faster just makes more mess for little gain)
• Spray on enough resist resist to just coat the board (liquid resist would probably be easier)
• Turn on fan and gently heat from above with a hair dryer, it should dry within about 2 minutes.
• Expose for 2 minutes with UV LEDs or lamps, lamps give a much more even exposure.
• Make a solution of sodium hydroxide, I'm using about 1/4 tsp to 300ml of water.
• Dip in solution and wait until the exposed resist starts coming off, it's fairly hard to judge so I like to wash and develop multiple times.
• Put into a solution of warm ferric chloride and frequently check until it is done.
• Clean off old resist and coat with a rosin flux to prevent oxidization.

So if anyone else is having trouble, give this a try. 

[IanB]

That seems like a good process.

A thought occurs that if you used liquid resist and added a drop or two of food colouring as a dye, it would help to check for even coverage of the board and also make it easier to see how developing is progressing after exposure?