PCB Design : Thinnest tracks.

[lord loh.]

I am trying to design a PCB using the Laser printer/Iron/FeCl3 process...

I managed to print a board with 0.25mm track width on a PCB. However parts of the tracks were faded and lost...due to inefficient ironing. (my fault) However most of it was clear and well imprinted.

I retouched the faded parts with a marker and corroded it and ended up with a few broken tracks. (circled red in picture)

I would like to know what is the thinnest track that can be achieved by the laser printer/Iron/FeCl3 process ?

Also when I checked the continuity of the good tracks using the diode tester range of the multimeter, i got figures like 325, 215 ...(mV of voltage drop between two points on the track)

Couldn't this high resistance of tracks be a serious problem?

Pleas give me a few tips and suggestions for my next PCB... Should I go for a 0.30mm / 0.40 mm ?

I am using a software called diptrace (https://www.diptrace.com/)

Thanks in advance.


=============
Added later :

What is the mm to mil conversion? and what does mil stand for?

Thanks again.

 

[lord loh.]

BLUNDER !!!!

Forgot to post the picture...

 

[evandude]

I regularly use 10 mil traces on my boards, which is 0.254mm... a mil is 1/1000 of one inch, ie- one milli-inch

the most important part is the transfer paper, and you don't mention what kind you are using. with photo paper, I could generally only use about 16 mil traces reliably, and even then I'd end up with at least a few broken traces on a board.

now I use press-n-peel blue and have no trouble with 10 mil traces... almost never do I get a single broken trace or other defect. as you can see, it transfers very cleanly and etches very well.

**broken link removed**
**broken link removed**

 

[lord loh.]

I use the regular A4 size photocopy paper.

Does the resistance of the track pose any problem when you make 10mil tracks? My continuity tester showed me a 300mV drop between distant points on the track. Is this accptable?

What is the press and peel blue paper? Don't know if it is available in my city (Ahmedabad)

 

[evandude]

I don't get any voltage drop like that on my tracks, it looks to me like your tracks are so eroded from incomplete etch resist coverage (fixing with a sharpie doesn't do a very good job) that they are making for very high resistance traces...

when I etch with press-n-peel, the copper underneath the traces ends up completely un-etched, you can still see the polishing marks on it, there are no rough, patchy sections.

press-n-peel isn't something you would be likely to find locally, it's only sold at a few places and for the most part it's mail-order.

their website is https://www.techniks.com

if you can't get that, then at least try glossy photo paper... it's still a pain in the butt but at least it's better than plain paper... if you are stuck with photo paper or plain paper, then you definitely need to use larger trace widths...

 

[lord loh.]

Saw the site...

Nice... The trick lies in the paper I understand.

And why do people all over drill after itching? I drill before itching. I am worried that the thin itched tracks may break during drilling...

 

[evandude]

And why do people all over drill after itching? I drill before itching. I am worried that the thin itched tracks may break during drilling...

I drill after etching because the etched copper ring tends to help guide the drill bit to the center of the pad. True, some pads do suffer damage, but generally it's not bad enough to make it unusable by any means.

 

[chemelec]

I Don't really understand Why you would use such Fine Tracks.

Unless you have a design that Specifically needs them, use Wider Tracks.
Most often I make my tracks between .75 mm and 1.25 mm.

Heavier current tracks, typically 2 mm or larger.

If I'm placing tracks between IC Pins, I can etch as small as 0.125 mm, but I only use a length long enough to clear the IC pads, than increase the width again. I can easily get two track between IC pins, if needed.

Remember the More you etch away, the More Etchant you use.

Also a good idea on a board like you show, Fill in some areas with copper.
They can just be Fills, Not connected to anything, But less to etch away.

Drilling holes Before etching is also not really good. You get etching on the Edges around the hole. and as Evandude said: The lack of copper, or indent where the Hole is, makes an Excellant Drilling Guide.

Not that I use that laser Printer Process, I do too many, Multiples of boards for that.

 

[evandude]

yes, for a board as spread out as he was doing, using such fine traces doesn't make much sense except for squeezing between the IC pins. Personally I stick with 10 mil traces for the most part because I really cram things together on most of my boards, and I use a lot of surface-mount parts and stuff, so only small sections of any given trace could be widened so it's hardly worth the hassle. (a good percentage of my boards are only a couple square inches, and usually quite well populated with components) but I do generally go with something more like 16 mils or so when I have a more spread out board design.

but yeah, lord loh, I would definitely increase your trace width everywhere possible, especially the long ones... you could easily get away with doubling the width almost everywhere, if not more.

 

[lord loh.]

Okay, narrowing the tracks only between the IC pins seems a good idea.

I intend to drive six 7 segment displays and a few(6) 74LS series logic ICs and a 555 timer in my most populated board. So will a .25mm to .40mm trace provide enough current?

What I have been doing so far is to leave no hole place in the pads in the printout. evandude says that it helps guide the drill... I 'll try that out.

And what do you suggest the pad sizes? I use 2.4mm x 1.5mm rectangular pads with hole diameter 0.9mm

Thanks for all the help so far.

 

[pike]

Yes as said before keep the traces as big as possible, lowered resistance, increased current carrying capacity.

You must have wasted alot of etchant making that board. dont do that, while it makes to board look more proffesional, it certainly is not practical.

I've never had succes with the iron-laser etching, i just use the photo method now, i can line down to less than 5 mil. (eagle wont go lower tha that!)

 

[evandude]

how did you even get eagle to go to 5 mils? the lowest I've ever been able to get from it is 10 mils... (.01 inch)

 

[lord loh.]

I did not use the copper pour, as I was worried about track and the pour getting shorted... I could adjust the track clearance but did not bother... Yes it took me a lot of enchant...: (

I am still pondering about the pad size... is 2.4mmx1.5mm fine?

Also has any one managed to drive two tracks between two pins of an IC? I have got a circuit that just wont route without this extremity...

I am in a state of hesitation whether I ought to try to implement it or not...

 

[evandude]

Also has any one managed to drive two tracks between two pins of an IC? I have got a circuit that just wont route without this extremity...

you know, using a wire jumper or two isn't the end of the world...

 

[mstechca]

Here are my suggestions:

Go with WIDE tracks. Not only will the chances of broken tracks be smaller, the circuit will function more naturally because the narrower the track, the higher the resistance. Ohms law can prove this.

Drill AFTER etching, especially when your drill isn't great (like my hobby motor based drill). The copper rings around the spot you drill in help you drill in the right spot.

Evandude is right to an extent when he says that drilling doesn't have a large impact on the track.

If the tracks leading to the ring are large enough, then there will be no problems after drilling, unless you can't drill at all.

BUT, if you are using skinny tracks <0.3mm, then the whole ring could rip off if you don't be careful with your drilling.

 

[lord loh.]

you know, using a wire jumper or two isn't the end of the world...

After breaking my head ove the circuit for hours, I decided that it indeed is not the end of thew world to use a jump wire or two...

I guess people sometimes have to let go of the perfectionist nature that is inherent in humans...

Thanks a lot for all the help...

 

[Nigel Goodwin]

you know, using a wire jumper or two isn't the end of the world...

After breaking my head ove the circuit for hours, I decided that it indeed is not the end of thew world to use a jump wire or two...

I guess people sometimes have to let go of the perfectionist nature that is inherent in humans...

If it bothers you, then don't use a 'jump wire', use a zero ohm resistor instead, then no one will know :lol:

Are you manually routing the board?, or using an auto-router to do it?, if you are using an auto-router, try doing it yourself instead.

 

[mstechca]

If it bothers you, then don't use a 'jump wire', use a zero ohm resistor instead, then no one will know :lol:

Nigel :lol:
an RF circuit will know :shock: :twisted:
I just have to remind you that extra leads of the resistor equal inductance/more resistance, unless the solder touches the resistor itself. I am not going to take this idea for RF circuits. I recommend a jump wire myself. In fact, use 2, 3, or even 4 jump wires bunch together.

I think one objective of making a halfway decent circuit is to avoid unwanted resistances.

You should think of widths as resistance divisors. Let's say each track being 1mm in width is 10 ohms per meter (I'm picking numbers here).
if the width is 2mm, then ohms law says the resistance is 5 ohms. 3mm = 3.333... ohms, 4mm = 0.25 ohms, etc.

Are you manually routing the board?, or using an auto-router to do it?, if you are using an auto-router, try doing it yourself instead.

I do 99% of my circuitry manually when I do circuits.
Manual is better. You will avoid bugs made by the programmer of the software. Also, manual allows you to connect the parts any way you want, AND you won't have to fix whatever mistakes the computer has made, with respect to your design.

 

[Nigel Goodwin]

If it bothers you, then don't use a 'jump wire', use a zero <a href="#">ohm</a> resistor instead, then no one will know :lol:

Nigel :lol:
an RF circuit will know :shock: :twisted:
I just have to remind you that extra leads of the resistor equal inductance/more resistance, unless the solder touches the resistor itself. I am not going to take this idea for RF circuits. I recommend a jump wire myself. In fact, use 2, 3, or even 4 jump wires bunch together.

A zero ohm resistor IS a jumper wire, just with a shaped body fitted around it, so it can be inserted by component insertion machines. There's no need to use multiple jumpers unless you have a high current requirement.

 

[lord loh.]

I did think of the zero ohm resistor... to fox my teacher who insists on optimizing any given circuit a "bit more" to have no jumper wires...

I do a manual routing of close and obvious routs...and leave the long distance routes to the grid router. This way, the routing is faster as the router does not touch the existing tracks. The router several times made a mess of routing the nearest points. One of the nearby traces went all round the board before connecting to the final point...

However the autorouter does not recognize resistance values. Only resistors. So the two terminals of the zero ohm resistor is considered in two different nets. This creates more problems that what it solves. As some nodes on both sides of the resistors are in the same net. So the router attempts more complex routings and makes a more complex mess.

My software is Dip Trace (https://www.diptrace.com/) freeware edition. It allows upto 250 pins. And was the best software I could get after a lot of long head breaking sessions with a dozen softwares that I evaluated...