PCB Design : Thinnest tracks.

I'd suggest you eas yourself over to eagle... I know it can certainly be "head-breaking" when you get started, those crazy germans sure didn't make the interface nearly as intuitive as it should have been, but once you get used to it it's really not hard at all...

generally when I need jumper wires I just design the board as a 2-layer board... anywhere I wish to have a jumper, I go up to the top layer, over the traces I need to cross, and then back down... then I etch the board as a single-layer board (bottom layer only) and simply put jumpers in place of each of the top-layer tracks.

For using the autorouter, I generally do a bunch by hand, then maybe let the autorouter run and see what it comes up with, and then I edit the autorouted tracks manually until they look like what I want... shortening them, rerouting, and adding jumpers as needed. But on a lot of my boards I don't use the autorouter at all... generally unless it's something with a lot of buses (like in a digital logic or microcontroller circuit where you might have an 8-bit bus running between a few chips), there isn't too much need for an autorouter.

But I promised myself to always allow myself a few jumpers when I had spent several hours trying to get a certain board optimized to a single side with no jumpers, when I realized that it would take me 2 minutes to solder in a jumper or two, thereby saving myself a lot of time tearing my hair out on the design instead.

evandude said:

I'd suggest you eas yourself over to eagle...

Click to expand...

I'm against eagle. I tried it once, and I didn't like its autorouting, or appearance of the components. It made me think that the 1st and subsequent printouts will be bad.

but once you get used to it it's really not hard at all...

Click to expand...

I am sticking with IVEX winboard.

generally when I need jumper wires I just design the board as a 2-layer board... anywhere I wish to have a jumper,

Click to expand...

why?
What I do is create a jumper piece and use it. All it is is two drill holes about 0.2mm apart.

I go up to the top layer, over the traces I need to cross, and then back down... then I etch the board as a single-layer board (bottom layer only) and simply put jumpers in place of each of the top-layer tracks.

Click to expand...

I drill holes right through on a single sided board, and fit the jumper in place.

For using the autorouter, I generally do a bunch by hand, then maybe let the autorouter run and see what it comes up with, and then I edit the autorouted tracks manually until they look like what I want... shortening them, rerouting, and adding jumpers as needed.

But on a lot of my boards I don't use the autorouter at all...

Click to expand...

I go manual.

generally unless it's something with a lot of buses (like in a digital logic or microcontroller circuit where you might have an 8-bit bus running between a few chips), there isn't too much need for an autorouter.

Click to expand...

I won't use an autorouter unless I expect extra mistakes. In other words, I don't use it.

But I promised myself to always allow myself a few jumpers when I had spent several hours trying to get a certain board optimized to a single side with no jumpers, when I realized that it would take me 2 minutes to solder in a jumper or two, thereby saving myself a lot of time tearing my hair out on the design instead.

Click to expand...

I guess it will work. Just watch with the jumpers at RF. even a millimeter of wire can have an impact on the frequency. In fact, I am totally against putting a closed jumper and an inductor in series unless the jumper can be opened or closed by the end user.

mstechca said:

I'm against eagle. I tried it once, and I didn't like its autorouting, or appearance of the components. It made me think that the 1st and subsequent printouts will be bad.

Click to expand...

but you also say you don't actually use autorouting. personally I don't mind the appearance of components in the schematic, and the PCB printout is the important part anyway... and that comes out fine as long as you set the DPI high enough on the bitmap.

mstechca said:

I am sticking with IVEX winboard.

Click to expand...

go for it, not to be a jerk, but I wasn't talking to you, you wouldn't listen to me anyway :lol: I was suggesting eagle to lord loh, since he said that his software limited him to 250 pins.

mstechca said:

why?
What I do is create a jumper piece and use it. All it is is two drill holes about 0.2mm apart.

Click to expand...

I don't do that because it divides one net into two, so you have to show it in the schematic (which looks messy) and you have to put it in just the right place in the schematic... if you shuffle stuff around, you'll have to move the jumper both on the board, and in the schematic... that's annoying. using the top layer for "virtual jumpers", you still get two holes to use for a jumper, and you can make it any size and shape you want, without having to make a new footprint every time.

mstechca said:

I drill holes right through on a single sided board, and fit the jumper in place.

Click to expand...

yes, I do that too... when you go from bottom to top layer, it inserts a Via, which is exactly the same as a normal solder pad for a component...

mstechca said:

I guess it will work. Just watch with the jumpers at RF. even a millimeter of wire can have an impact on the frequency. In fact, I am totally against putting a closed jumper and an inductor in series unless the jumper can be opened or closed by the end user.

Click to expand...

yes, but then again practically anything you do in your PCB design affects things at RF... after all, a PCB trace is still a form of wire as well. in reality, if you are dealing with RF stuff that's sensitive enough to be totally messed up by a small bit of wire, it would probably be wiser to be using a 2-layer board with significant ground planes in the first place.

evandude, I recently read a post on this forum, where someone said that his mother board buses were twisted and looped and deliberately made long. The reply he got was that it was done to make all the bits reach the destination at almost the same time by giving each trace the same amount od inductive effect instead of making some traces of a bus short while other long. DipTrace bus routing did no such thing. I do not need such a precision as I am not designing a high speed bus... just trying....

About eagle...(freeware)

It enforces a

The useable board area is limited to 100 x 80 mm (4 x 3.2 inches).

Click to expand...

So 250 pins limitation is fine... I used two boards with connectors between them.

DipTrace is more intutive...So I feel. (https://www.diptrace.com/) the software is about 7 Mb.

I have finished designing 5 of the 7 boards required for my project.

evandude said:

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.

Click to expand...

I tried photo paper and also used .4 mm tracks...

Photo paper (Kodak) give me no appreciable difference at least in the .4 mm scale except that the paper was harder to pulp up and remove after the transfer.

I transferred the image on to the smooth side (or should I have done it onto the rough side?)

Any suggestions?

lord loh. said:

evandude said:

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.

Click to expand...

I tried photo paper and also used .4 mm tracks...

Photo paper (Kodak) give me no appreciable difference at least in the .4 mm scale except that the paper was harder to pulp up and remove after the transfer.

I transferred the image on to the smooth side (or should I have done it onto the rough side?)

Any suggestions?

Click to expand...

Photo paper isn't perfect. If you cant' get press-n-peel and you don't want to do the UV exposure method, then you will probably be stuck with larger traces. You could try more brands of photo paper, I'm sure with google you could find some recommendations for paper type, but I think it will involve a lot of trial and error.

there are two other methods which I have done.

WAX: Cover the entire copper board with wax and leave the wax on the sections you dont want etched.

MARKER: Draw your design with a thin black permanent marker, and etch the board.

I found that after using the UV method, the above two methods wasted my time.

you can calculate the real resistance of a track: R = rho * length / ( width * thickness). Why don't you use your DMM measuring resistance instead of diode checking?