PCB developing times

[Simon]

Hi guys, I made up a board last night but over etched, the previous board was fine but this one I did not leave as long in the caustic soda, maybe this was the mistake as it did not remove all the unwanted photoresist ? or coukld the etching solution be exaustng. i'm always worried about over developing the photoresist and ruining the board, but am I going the other way ? A comercial chemical recomends 30-60 seconds, i'm sure I was in at least this long

[ElektroQuark]

Hi:

I use the Bungard boards. The user guide stand that the developing time is about 30-60 seconds but the photresist could be in the developer for about 5 minutes without been damaged.

Can you post some fotos of the etched PCBs?
Also, describe your procedure, including times and products used.

Iñigo.

[mikeselectricstuff]

Don't  use caustic soda (NaOH). Use a silicate based developer.
Silicate has a HUGE margin between 'develop' and 'strip', making concentration and temperature much less critical than NaOH. It also has infinite shelflife, whearas NaOH tends go off after a while. 
You can make silicate up so strong that it will develop in 10 secs but not strip for a couple of minutes. Different makes of laminate do vary significiantly in their develop time.
A stronger developer also makes it easier to see when developing is complete. Wiping gently with a dry paper towel is a good way to remove the last traces of developer & gives cleaner trace edges. 
To tell if it is fully developed, a quick dip in ferric chloride will turn developed PCB pink instantly.
See my PCB page for lots of tips on making good PCBs

[Simon]

Well I used a LED UV box i made myself and exsposure seems to be from 30-60 seconds, I think 45 is a good average value, i like to make sure it's well cooked but I've noticed that there are some tiny holes in the film printout and these have been etched so maybe I'm over exposing.

them I put a normal (eating) spoon of caustic soda in 1 litre(ish) of water and develop until the traces are nice and sharp and maybe a bit more (I was dubious about rubbing in case i damaged the tracks but that sounds like a good idea)

then I go into the etchant for maybe up to 5 minutes.

I think my problem might be over exposure and under development

[Simon]

I think i also need to get a cheapy inkjet printer as my gel printer is not too good on printable film

[mikeselectricstuff]

get a laser - dirt cheap nowadays, and print on thick (90gsm or more) tracing paper.

[Simon]

yea but you get higher contrast with the film, basically printing 2 (one landscape one portrait) and putting one on top of the other makes a nice dense mask. I've just been playing with the file in photoshop and found that for black non of the 4 colours was fully filled in, so I've done it manually, if I can get 1005 ink on all 4 colour layers I have a chance of a decent print

[mikeselectricstuff]

yea but you get higher contrast with the film, basically printing 2 (one landscape one portrait) and putting one on top of the other makes a nice dense mask. I've just been playing with the file in photoshop and found that for black non of the 4 colours was fully filled in, so I've done it manually, if I can get 1005 ink on all 4 colour layers I have a chance of a decent print

That sounds like a lot of  hassle - if you're thinking of replacing the printer then a laser is the way to go, as you get a useable print in one pass with no messing about. Tracing paper is also pretty cheap.

[ElektroQuark]

I can't help you without photos.


Here:
htp://www.todopic.com.ar/foros/index.php?topic=28880.0
I describe my method to make PCBs. I show a tiny one (for an SMD CI), and how I use solder paste and a toaster to solder.
I'm sorry it is in Spanish, you can translate it using Google.

[Simon]

yea but you get higher contrast with the film, basically printing 2 (one landscape one portrait) and putting one on top of the other makes a nice dense mask. I've just been playing with the file in photoshop and found that for black non of the 4 colours was fully filled in, so I've done it manually, if I can get 1005 ink on all 4 colour layers I have a chance of a decent print

That sounds like a lot of  hassle - if you're thinking of replacing the printer then a laser is the way to go, as you get a useable print in one pass with no messing about. Tracing paper is also pretty cheap.

Well not like I do loads of boards and i'm setup for it anyhow, it just takes a minute to convert. I actually did not get a great result with my work laser so gave up the idea, I'm not buying a printer on the off chance and i do rather like the film. As an ex offset press operator lining up tow films and taping them in place is a doodle

[Zad]

I have had excellent results with a cheap inkjet printer, using a single layer of transparency - the ink seemed far from opaque. The key is calibration. Draw up a smallish board (it only need be 2x5cm) in Photoshop or whatever, with graduated line sizes across it. Using a mask, expose areas of the board for 15/30/45/60/75/90 seconds or whatever you feel is appropriate. Drop it into silicate developer and slowly rock the liquid back and forth. You should be able to see the resist change colour almost as soon as the board goes in. Then gradually the agitation will remove the exposed resist and you should see the copper show clearly through. Then etch in a similar way, observing when the copper is fully eaten through.

If you put it in a bath and leave it for 'x' seconds then you are going to get a lot of reject boards before you hit on the right timings.

[Simon]

Well I play it by eye mostly, I had been ramping up exsposure times but I think i need to drop back on that as I'm over doing it, as for developing I need have a little more nerve and keep it in longer. how hard can i rub on the developing board without damage to the photoresist I want to keep ?

[mikeselectricstuff]

It is generally better to overexpose than under, as an underexposed board will never develop fully, whearas if grossly overexposed you can sometimes rescue it at development.
With silicate devloper the margin is pretty huge as long as your artwork is reasonably opaque, so you an add a decent exposure margin, like about 50% to cover uneven exposure, temperature variation etc.

You should see the exposed areas start changing colour within 15-20 secs of going into the developer. When doing test pieces to figure out optimum expose/develop conditions, start with the developer really strong, and get the exposure right first, then if necessary dilute the developer once you've worked out the ideal expose time for your UV source and laminate.

A gentle rub with dry kitchen towel or tissue paper is about the optimum amount of abrasiveness to help remove traces of exposed resist without damage.
Unexposed resist is pretty tough and you have to try quite hard to damage it, so don't be afraid to rub quite hard.
 
Different brands of board can have significantly different develop times - I suspect due to resist thickness variation, but I've never had to adjust expose times for different ones. 
I highly reccommend Microtrak board from Mega

[Simon]

well I've had good results from my LED UV box with 30 seconds so 60 is well exposed, I'm not sure if it was due to under development or over exposure that the tracks are pitted (if under development it's because the etchant ate the stuff i wanted as well as what I did not want)

I think more time in the developer is a good start as I've good good boards from a 60 second exposure but can drop that down a bit if I think I'm in trouble

[ElektroQuark]

I will try again. I don't know why.

If the PCB is underexposed, it will need "over" etching, and you will obtain pitting and tracks will be thinner. That's why I requested a photo of the etched PCB. A PCB artwork to compare track width will be usefull aswell.

[Simon]

the artwork is quite dense as it is 2 layers overlapped, the traces did come out pitted I think because I over etched it because i had under developed it, but obviously any pits in the artwork (seen only with a magnifying glass under strong light) will not have helped with the long exposure

[Simon]

I've already used the board it was a bit undercut along the tracks but i was going to tin then anyhow as I have no other way of protectng the copper from tarnishing

[mikeselectricstuff]

I've already used the board it was a bit undercut along the tracks but i was going to tin then anyhow as I have no other way of protectng the copper from tarnishing

A fairly effective way is to leave a coat of flux on - I find the spray stuff tends to stay sticky, but the liquid type you get in pens dries pretty well. Coating a fresh PCB in flux also makes it way easier to solder.
Although some manufacturers claim that photoresist acts as flux, it's pretty useless in practice and should be stripped off.

[Simon]

yea I stripped the photoresist with appropriate stripper chemical and then penned on flux and tinned the lot