Homebrew PCBs

Well, after babbling away about how good my el-cheapo ML-1210 laser 
printer makes transparancies, I made a little board (4"x4") with 
0.008" traces on 0.020" centers over half of it and the other half 
as a ground plane. The transparancy came out great, no pinholes etc. 
Exposured and developed, only found one small break (and a 
corresponding piece of dust on the transparancy). Quick jab with my 
sharpy and its ready to go.

The shocker was when I went to turn on the heater for my ferric 
chloride. I hadn't realized it but its been months since I etched a 
board - instead of being a translucent dark brown, it was an opaque 
greenish brown goop! I didn't have any more on hand and it was a 
sunday so..., topped it up with water, heated it up, and etched 
anyway - worked perfect! In fact, this was about the best etching 
I've managed (probably because I exposed longer than needed and got 
more aggressive with brushing in the developer.

Why would it turn green? Something growing in there?

ghidera2000 wrote:
> Well, after babbling away about how good my el-cheapo ML-1210 laser 
> printer makes transparancies, I made a little board (4"x4") with 
> 0.008" traces on 0.020" centers over half of it and the other half 
> as a ground plane. The transparancy came out great, no pinholes etc. 
> Exposured and developed, only found one small break (and a 
> corresponding piece of dust on the transparancy). Quick jab with my 
> sharpy and its ready to go.
> 
> The shocker was when I went to turn on the heater for my ferric 
> chloride. I hadn't realized it but its been months since I etched a 
> board - instead of being a translucent dark brown, it was an opaque 
> greenish brown goop! I didn't have any more on hand and it was a 
> sunday so..., topped it up with water, heated it up, and etched 
> anyway - worked perfect! In fact, this was about the best etching 
> I've managed (probably because I exposed longer than needed and got 
> more aggressive with brushing in the developer.
> 
> Why would it turn green? Something growing in there?

The active ingredients can settle to the bottom. Add some HCl and
stir it.

ghidera2000 wrote:
> 
> Why would it turn green? Something growing in there?
> 
> 

Considering copper compounds make very good fungicides (down to 
the ppm level) I think it would be a safe bet to say its not 
growing in there :)

The green is obviously due to copper compounds, some form(s) of 
copper chlorides. Add some conc. HCl until the green sludge is no 
longer green, this will dissolve any copper hydroxide and bring 
it back to life.

Ah ok, the light comes on. Copper<->Green... I should have made the 
connection.

I only have about 900 litres of HCL at work, think that'd be enough? 
(Dr. Evil Laugh!).

--- In Homebrew_PCBs@yahoogroups.com, Adam Seychell  Add some conc. 
HCl until the green sludge is no 
> longer green, this will dissolve any copper hydroxide and bring 
> it back to life.

i'm sure it isn't the laquer.
also the metaloxide paints should isolate.
maybe there are some "metallic" look paints which conduct but....

it is not very wise anyways to cover up the pcb before you tested it.
i never spray mine at all and if not touched with your fingers they won't 
oxidize too much at room conditions.

you should never assume any circuit working on the first try.
you can only assemble a board fully and spray it without testing when you 
have already made a second working one before.

but take the laquer of for your convenience when soldering to repair it.
take any solvent and brush. if you have plastic parts (connectors etc) on 
the component side don't wet them with solvent.
it may dissolve them.

then get the scematic and test every function.
assuming you have no scope this makes it necessary to get all "nice and 
slow".

fast rs232 voltages you can't test with your meter.

if you provide the scematic somewhere maybe one may have a look on it and 
tell you what even if this is mostly a pcb hardware
manufacturing not debug board....

hope you get it right....

stefan

Actually, I sprayed my PCB AFTER I tested it by using a 9V battery 
set the signal input to positive and negative.  The booster responded 
correctly and the polarity of the output was correct in both cases.

The lacquer I sprayed on the PCB is a clear coat, which doesn't 
contain any pigments.  Therefore according to your comments, this 
shouldn't affect the circuit.  Am I coorect?

Now, I suspect that the positive voltage on the TxD pin on the RS232 
port of the ATX motherboard may not be high enough to drive the 
signal transistor (see the booster.bmp in the booster directorey in 
the files section).  Therefore I swap the positions between the 390R 
and the signal transistor on both sides, so that the signal 
transistors can be activated at a low voltage (booster2.bmp).  Do you 
think this can solve the problem?

Mark


--- In Homebrew_PCBs@yahoogroups.com, Stefan Trethan 
<stefan_trethan@g...> wrote:
> i'm sure it isn't the laquer.
> also the metaloxide paints should isolate.
> maybe there are some "metallic" look paints which conduct but....
> 
> it is not very wise anyways to cover up the pcb before you tested 
it.
> i never spray mine at all and if not touched with your fingers they 
won't 
> oxidize too much at room conditions.
> 
> you should never assume any circuit working on the first try.
> you can only assemble a board fully and spray it without testing 
when you 
> have already made a second working one before.
> 
> but take the laquer of for your convenience when soldering to 
repair it.
> take any solvent and brush. if you have plastic parts (connectors 
etc) on 
> the component side don't wet them with solvent.
> it may dissolve them.
> 
> then get the scematic and test every function.
> assuming you have no scope this makes it necessary to get all "nice 
and 
> slow".
> 
> fast rs232 voltages you can't test with your meter.
> 
> if you provide the scematic somewhere maybe one may have a look on 
it and 

> tell you what even if this is mostly a pcb hardware
> manufacturing not debug board....
> 
> hope you get it right....
> 
> stefan