I learned early on in my experience with Eagle to place and connect a schematic symbol or wire. Then attempt to move it. Properly connected pins or wires will rubber band. Unconnected ones will move. They will usually connect if dropped in the proper place. If not, delete the (non)connection and do it over.
I still haven't taken the time to learn KiCad.
So after 9 pages of comments, I seem to be the only one that uses Easy EDA... Did any of you try it?
There are dozens of different pcbcad packages out there now.
From free to very expensive.
I remember buying a copy of EasyPC in about 1990 for DOS.
Basically ok but had a few bugs.
From free to very expensive.
I remember buying a copy of EasyPC in about 1990 for DOS.
Basically ok but had a few bugs.
How about simply use the pcb software only to design the wiring and use schematic diagram only as a source.
Simply start drawing the pcb layout with "trial and error".
This is the way I always do it using Fritzing with satisfaction.
Simply start drawing the pcb layout with "trial and error".
This is the way I always do it using Fritzing with satisfaction.
I haven't used EasyEDA either, but I have an account and use LCSC.
Part of that is that I don't like the online part of it.
I happily use EasyPC6 and PCB express, but mostly print my own boards.
Diptrace is ok, but I fine the UI annoying.
Cadstar much the same.
You can download and run EasyEDA locally if you like... That's how you have to do it to use autorouting - you run the autorouting server locally, too.
I think the trick there is while placing/routing the PCB try and keep a mental map of the schematic in your head. If you see something different to what you expect then time to check. My worry is getting a footprint wrong!
You have to triple check the footprints, both with the data sheet, and by direct measurements.
I have seen parts that did not comply with their data sheet, and would not have fit
on the board had I not measured them.
Last edited:
I take this a step further. Print one side of the board on card stock or thick photo paper. Then test fit the parts in the paper board. This finds the bad component models as well as the "all the parts fit but you still can't build it or use it" moments.
These are things like parts too close together, trim pots that have a fat cap right in front of an adjustment screw, and test points that are inaccessible or close enough to high voltage or hot tube to cause loss of vocabulary control when testing a live amp.
Always put some lettering on each layer of your board. This is especially important if you are making your own boards. It's too easy to get the layers swapped or a single layer board mirrored.
I still have this reminder visible when making boards. I got it stuffed with enough parts for a first test, but smoke happened when I powered it up. It took a while to realize that the layers were swapped causing the tube to be miss-connected and every semiconductor was in backwards. I lost lost my cool, threw it against the concrete wall, then tossed the remains in the trash. A day later after I thought about it, I got it back out of the trash, replaced all the fried mosfets and soldered a second tube socket on the other side of the board.
I still make P-cad (paper cad) models of my more complex boards before actually making a proto board to find stuff like this.
These are things like parts too close together, trim pots that have a fat cap right in front of an adjustment screw, and test points that are inaccessible or close enough to high voltage or hot tube to cause loss of vocabulary control when testing a live amp.
Always put some lettering on each layer of your board. This is especially important if you are making your own boards. It's too easy to get the layers swapped or a single layer board mirrored.
I still have this reminder visible when making boards. I got it stuffed with enough parts for a first test, but smoke happened when I powered it up. It took a while to realize that the layers were swapped causing the tube to be miss-connected and every semiconductor was in backwards. I lost lost my cool, threw it against the concrete wall, then tossed the remains in the trash. A day later after I thought about it, I got it back out of the trash, replaced all the fried mosfets and soldered a second tube socket on the other side of the board.
I still make P-cad (paper cad) models of my more complex boards before actually making a proto board to find stuff like this.
Attachments
I once designed a tqfp foot print.
I used correct dimensions but when pcb came it didnt fit.
As pins went along they slowly got out of sync with pads.
My CAD software was in imperial so metric had to be rounded up/down to fit.
This meant errors could accumulate along the sides of IC's.
So instead stepping pin to pin I tool the measurement from pin 1 and that got rid of accumulative errors.
I too usually print out components or pcb and test fit them if not sure and this has spotted a few problems over time.
I also take great care with 2d component outlines so they are exact, this saves messing around with 3d models etc.
I used correct dimensions but when pcb came it didnt fit.
As pins went along they slowly got out of sync with pads.
My CAD software was in imperial so metric had to be rounded up/down to fit.
This meant errors could accumulate along the sides of IC's.
So instead stepping pin to pin I tool the measurement from pin 1 and that got rid of accumulative errors.
I too usually print out components or pcb and test fit them if not sure and this has spotted a few problems over time.
I also take great care with 2d component outlines so they are exact, this saves messing around with 3d models etc.
Yes, if you can print out the pcb to the exact correct scale, use that to check fit both for parts
and in the chassis.
But never buy a lot of pcb copies until you try at least one real board with all parts,
and get it fully operational.
and in the chassis.
But never buy a lot of pcb copies until you try at least one real board with all parts,
and get it fully operational.
Last edited:
"I've just had a look at DipTrace and KiCad and the UI is rubbish on both of them. IMHO"
kicad has icons all the way around the screen.
I hate icons as you have to hover mouse to see what they do.
I also prefer screen real estate for holding schematics/pcb's not icons.
kicad told me they can be reduced in size but not turned off.
kicad has icons all the way around the screen.
I hate icons as you have to hover mouse to see what they do.
I also prefer screen real estate for holding schematics/pcb's not icons.
kicad told me they can be reduced in size but not turned off.
Yes, software in general is terrible in UI. It's much harder to do it well, so few do.
Last edited:
Kicad has keyboard shortcuts.
66 Shortcuts for KiCAD
Diptrace also has customizable keyboard shortcuts.
66 Shortcuts for KiCAD
Diptrace also has customizable keyboard shortcuts.