I still don’t see any instruction about selecting GND net or any other net come to that. How do I make that choice?
I have the set, Rene. Just cutting them to DVD so I can sit down one Wet Sunday afternoon and work through them. There’s a lot of viewing there!
No, now you have an isolated zone not connected to any net. It is not “no zone” it is “<no net>”, but you were told to select “GND”. You were also told that when you do so correctly the track from R1 to J1 will be covered by the zone, which is okay. In the tutorial they say that track is not necessary as it will be replaced by the zone. So once the zone is in place you could delete that track and R1 would still connect to J1-3 via the zone.
The section you quote is a good example as to why the getting started tutorial is not really ideal.
They leave out some parts (telling you how to select a net)
But most importantly, they do not tell you why you would use a filled zone in the first place. They also forget to mention why you would use thermal relieve. (This is something i notice across the whole tutorial. They never tell the reader why something would be done in a real design.)
Lets fill in the gabs. (For this paragraph)
Why are filled zones used?
Filled zones have a lot of benefits to them. One of them is to help save edging fluid. This would be the benefit in your small pcb.
But the most important use case is to get good low impedance connections without needing to think too much about them. This of course works best if that filled zone has as little interruptions in it as possible.
If you have >2 layers you will most likely use one of them for a uninterupted ground plane. (Only disturbances to it will be vias and tht pads.)
In two layer design you are a bit limited as to what you can do. You will need both layers to connect your signal traces leaving very little freedom to get such an uninterupted ground plane.
A trick then is to have vertical connections on one layer and horizontal connections on the other. Filling free spaces of both layers with ground and connecting the two filled zones with lots of stitching vias where they overlap.
Why thermal relieve?
Connecting a pad to a large copper plane makes it hard to hand solder it. As the copper plane will act as heat sink. In such cases using only a few thin connections between the pad and the plane reduces how much heat you loose to it.
So using thermal relieve makes it easier to hand solder a pad.
That looks much better.
Are these
what you are talking about?
Thanks for that, Rene, this sort of thing is very helpful.
Three of them are the connections created by the thermal relieve algorithm. The one on 45 degree angle is a trace.
I agree with “a bit limited” but not with “you will need both layers to connect”.
It was true when you had lot of TTL ICs arranged in matrix, but it is not true if you have one microcontroller and some external circuits.
I work on two layers and it is not so difficult to get all nets (except GND) connected on top (of course to get this you have to have a way to hide GND connections diuring elements placement).
En exaple (not from KiCad yet) of center part of such PCB:
Vertical track on left is 5V and goes around PCB, slanted track under microcontroller is VCC. Possibility of going out from under microcontroller with VCC to really any direction has capital importance (here not used) to get everything at one layer.
This post wasn’t ignored, Dale. I have been musing over it. As a Mechanical Engineer I am not impressed with banks of unused connections either. Murphy seems to find things for such connections to do - sometimes with strange results. As a complete electronic dyslexic I have some basic queries about this concept of additional resistors.
I am assuming that a pull-up resistor would attach to Vcc and a pull-down to GND. That’s quite an assumption and will probably be greeted with gales of mirth - but if it is wrong, please enlighten me.
If the above is correct, how am I to know which type of resistor to attach to which unused IC pin?
…or you can do a google search for “pullup pulldown resistor” and follow the links.
This looks like a good explanaiton for pull up and pull down resistors: https://www.electronics-tutorials.ws/logic/pull-up-resistor.html
In the majority (all?) microcontrollers nowadays pullup and/or pull down resistors are already included on chip in the uC itself.
If you want to know for sure, then get the datasheet of the PIC12C508A-I/SN from microchip and search the datasheet for pull up or pull down and/or read the section about general purpose I/O ports.
I very much like the “Getting started with KiCad” manual, but you have to be aware of it’s limitations. It is never intended to be complete, nor is it meant to build a working PCB. It is only meant as an introduction of the workflow in KiCad. The omisson of buffer and decoupling caps is a clear indication of this.
Therefore I want to suggest to read it twice.
Leave this circuit / PCB now for what it is, and start over with the “Getting started” manual, but now with a real schematic of something you want to build.
The “reference manuals” for KiCad have a more in-depth explanation on lots of details.
1 Like
That looks excellent. I’ve printed it and plan to sit down with it after lunch.
Thanks
1 Like
I’ve read it several times and worked through it at least three times 
Next play IS a real project 
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.