Etching first PCB problems - conducts between tracks

[wingnut]

Hi all

I tried to etch my first PCB today, but it conducts at about 30k to 100kOhms between some tracks. Did I not leave it in the ferric chloride long enough, or is there some conductive residue left on the board. The process I went through was as follows.
1. I prototyped the circuit on a breadboard.
2. I drilled all the holes on the copper plated fibreglass.
3. Then I cleaned the copper with an eraser and methylated spirits.
4. Then I drew the tracks on the copper with a permanent marker.
5. Then I agitated the board in ferric chloride solution in a dish for 12 minutes, rubbing it gently with a toothbrush.
6. It looked fully etched, so I washed it with water and then with meths, and dried it.
7. I mounted the components, soldering them on.
8. Everything worked fine except a TSOP which was supposed to be giving a reading of 5V was reading all kinds of varying lower voltages on its data pin.
9. We discovered that the etched board was conducting a bit, between pins, with a resistance of 30k to hundreds of kiloOhms.

Can anyone explain what this slight conductivity might be? Did it not etch fully? I don't think I would have noticed the problem if I was not trying to read a remote's IR signal through the TSOP, which is very susceptible to some kind of external pulldown resistance.

I have never bothered to make a PCB because my soldering is sub-standard, and it often gloops pins together in one ball of solder. But I will say that my soldering improved manyfold after I took an angle grinder to the point of the soldering iron, turning it into a lean, mean soldering machine

If you could also recommend a program to draw circuits, which is a circuit simulator and then turns the circuit into a PCB, preferably for free, that would also be greatly appreciated.

 

[(*steve*)]

Sounds like you should have etched it longer. An old rule of thumb is to etch the board twice as long as it took to first appear as if it was etched.

I find that for single sided boards you can get a good idea by holding the board up to the light. You can't do this as easily with double sided boards if they have any significant flood fill.

 

[BobK]

Did you remove all the components before testing the resistance between tracks?

Bob

 

[davenn]

5. Then I agitated the board in ferric chloride solution in a dish for 12 minutes, rubbing it gently with a toothbrush.

don't do that --- it will scrub off your marker pen and damage the tracks whilst etching

8. Everything worked fine except a TSOP which was supposed to be giving a reading of 5V was reading all kinds of varying lower voltages on its data pin.

you don't state if this above problem is specifically related to the below problem

9. We discovered that the etched board was conducting a bit, between pins, with a resistance of 30k to hundreds of kiloOhms.

weird sort of values ... for connection between tracks due to poor etching, I would expect zero Ohms or close to that

NOTE what Bobk said .... at what stage were you measuring those values ?

Dave

 

[(*steve*)]

Hey I missed that those readings were attributed to the poor etching.

Yeah, I would expect a lot more continuity.

Whenever I've done etching tests I've only ever seen open or continuity. I guess it's certainly possible to get any resistance, but the conductivity of copper is so high that any significant resistance would require either a very long, very narrow, or very thin layer (and probably a combination of these).

 

[wingnut]

Thank you all for the replies.

I had soldered a socket onto the board, into which I placed an Atmega 328 chip.
Because of its inability to read the TSOP soldered to pin 4, I removed the chip and measured for continuity between the pins of the now empty socket soldered to the breadboard. A lot of the pins were reading a big resistance between the pin and the board itself. For instance there was about 50k resistance between pin three and pin 4, and nothing was connected to pin 3. The TSOP data leg was attached to pin 4. Other pins were reading infinite resistance to the board. To the naked eye, there was no visible copper. I also washed the board with a toothbrush to remove leftover flux but that too did not help.

Here is a photo of the board. She is pretty ugly, I know.

There is random, high resistance between the supposedly bare board and some tracks. If I place a probe on any track and drag the other probe across the "naked" board, with the ohmmeter set to 2M, I get readings between infinite and 30k, changing all the time.






Next time I will leave it in longer and not use a toothbrush.

 

[(*steve*)]

Holding up to the light with the copper side toward you. Do you see any cloudiness where it should be etched?

 

[Bluejets]

When using your ferric chloride, place it in a tray slightly larger than the board.
Place the tray in hot water in the sink or whatever for a minute or so before popping the board in.
(just be careful with Mum's stainless steel sink......)
I usually place it copper side up and jiggle the board constantly while the etch is taking place.
Drilling I do after the etching.

12 minutes seems a long time to etch what you have there.
I remember the Dick Smith board used to take a long time so maybe it's from that era.
These days I use the chinese stuff, maybe 2 minutes all up.

Never heard of a problem with high resistance between tracks.........either zero or infinity as Steve says.
I'd say flip the board over and run your "meter" around on the bare side and see what you get.
Maybe you have 98% humidity today or sweaty hands??

 

[wingnut]

Holding up to the light with the copper side toward you. Do you see any cloudiness where it should be etched?

Steve, there definitely are cloudy areas in the tracks. I think it was because the permanent marker was a little dry and did not cover completely. One could see that some drawn parts were lighter than others, and these lighter areas etched away.

 

[wingnut]

When using your ferric chloride, place it in a tray slightly larger than the board.
Place the tray in hot water in the sink or whatever for a minute or so before popping the board in.
(just be careful with Mum's stainless steel sink......)
I usually place it copper side up and jiggle the board constantly while the etch is taking place.
Drilling I do after the etching.

12 minutes seems a long time to etch what you have there.
I remember the Dick Smith board used to take a long time so maybe it's from that era.
These days I use the chinese stuff, maybe 2 minutes all up.

Never heard of a problem with high resistance between tracks.........either zero or infinity as Steve says.
I'd say flip the board over and run your "meter" around on the bare side and see what you get.
Maybe you have 98% humidity today or sweaty hands??

Hi Bluejets

I am using the finest quality, ultra-cheap Chinese boards

How do you tell when it's all etched away? The way it looked to me was the permanent marker repelled the FeCl3 so in these places the colour of the marker could be clearly seen. In the area to be etched, I could not see which bits were etched down to board, and which were still copper. Only when I washed it in water could one tell.

So the bare copper and bare veto-board looked uniformly covered with that dark-yellow liquid. Being colour blind does not help either. I took it out at 10 min and washed it in water, but it still looked slightly coppery, hence the 2 extra minutes. I did have it in a small plastic dish and jiggled it non-stop.

There is no way that stuff is going anywhere near my sink, but it did go on some weeds.

I tore off the socket, sanded the socket tracks down to bare board, and point-to-point soldered another socket in place. Now all is working fine, and the TSOP is giving a stable reading.

I think I will try to etch another circuit today, taking it out every 2 minutes to wash and inspect it till I get some feeling for how long it takes to etch. I think I will use new FeCl3, keep the liquid at room temp, so that the result is reproducible in future. I might even hold the board up to the light and take a photo so that it gives me a record of how its supposed to look at all stages in the process.

 

[(*steve*)]

How do you tell when it's all etched away?

Take out the board, rinse it in water, and look at it. Holding it up to the light can be very revealing...

 

[Bluejets]

Not quite sure what you mean as when I do it, I usually have just enough echant to cover the board.
Then while moving the liquid back and forth, the board can clearly be seen "between waves" so-to-speak.

Again, I don't see where you need 12 minutes.
Are you heating the etchant as I described above?

I did one earlier today (115mm x 45mm) ...... 2 minutes tops.

 

[wingnut]

Take out the board, rinse it in water, and look at it. Holding it up to the light can be very revealing...

Did exactly as you said Steve taking it out every 2 minutes. Here is how it looked at 10 minutes...



I see what you mean, there is a kidney shaped patch of copper still to etch.

At 14 minutes with liquid at room temp (27 Celsius) it was fully etched. Here is what it looked like against the light - perfectly uniform.





I could see nothing "between waves" Bluejets. Maybe your solution is more concentrated than mine, or maybe my copper is a micron thicker. There is also zero conductivity between etched and tracks. Maybe it needed the 2 extra minutes. I only gave it 12 yesterday, and today is hotter. Yesterdays etched area looks a little cloudy or less opaque than today's, thanks for that tip Steve.

Very nice circuit board Bluejets.

Thanks for the feedback Steve and Bluejets.

 

[Bluejets]

Would not matter if your enchant was derived from a black hole.
As the liquid washes over the board on each successive tilt of the tray, the board becomes exposed and one can clearly see whether the copper is completely etched or not.
I cannot understand why you cannot see this effect.