I am struggling to convert files from Eagle to KiCAD format for editing.
This is what I need to convert...
http://reprap.org/mediawiki/images/b/b0/ArduinoMegaPololuShield.zip
I need to use something that is not Eagle as I explicitly want to make
the PCB much bigger.
Here is a Hackaday article I tried to follow. It lied. It wasn't easy
at all!
http://hackaday.com/2015/12/27/eagle-to-kicad-made-easy/
I got as far as Step 6 of the instructions here:
https://github.com/lachlanA/eagle-to-kicad
This left an Eagle window open at a description of Pin header
ULTIMACHINE_3PIN-HEADER
I then carried out step 7, opened pcbnew from the command line, and it
refused to open the Eagle .brd file. The option to open an Eagle format
file was there, but I got this error when I tried:
Error loading board.
IO_ERROR: /home/startup/Documents/RAMPS_1-41.brd(1): expected <
from /build/kicad-fW3pRG/kicad-4.0.2+e4/kicad/pcbnew/eagle_plugin.cpp :
Load() : line 1192
Any ideas?
I was using Eagle 7.6.0 for Linux 64-bit, not 6.x.x as is mentioned by
the instructions - if that helps...?
Cheers
Alex
Oh man, designspark....
Also, why is there so much through-hole?!
Good luck Alex.
I would agree you are best to start from a blank sheet. To make a successful design transfer, you actually need a greater level of competancy in the target package than you need to do the design from scratch.
It also does not save time, so is only really justified if you need the result to be identical. When you do the transfer a lot of information is lost. The base netlist and pads are usually OK, but you lose a lot of error checcking and contextual information, which makes checking and subsequent modification much more difficult and risky
We did a transfer from Eagle to DesignSpark on the Superman project. Even working with a DesignSpark consultant, we wished had not gone that route.
Richard
IMHO (and IANAL) the PCB layout is the only real thing that could be considered as IP, the actual schematic is really basic stuff that is just common stuff for interfacing the two (types of) modules. IE anyone tasked with creating a board to do that job would come up with a schematic that would be more or less the same. It would be like trying to licence a Hello World program.
But that being said (and to quote Adam Hill) don't be a dick and open source anything you do based on the schematic.
Cheers,
Jeremy
I think this is where the Adam Hill ‘don’t be a dick’ test applies.
I have an objective, which is to make a working 3D printer controller in a certain form factor, with additional functions compared to RAMPS and with better RF shielding.
If I could have achieved this more quickly and reliably by easily modifying RAMPS, which had to involve reliably transferring it to free software with a larger PCB capability, and just changing the layout, I would have done so and complied with all the requirements of the GPL, at least for the first iteration – my plans for future iterations always involved a total rework.
Now, since it appears I would need to re-draw the circuit fully anyway in order to be sure it will work, I am going to ignore RAMPS and design a new board completely from scratch, which I will licence myself. There are bound to be some superficial similarities at a schematic level with RAMPS (and its many variants), as it will be compatible with Arduino Mega and Pololu drivers, and form follows function, but I will start from first principles (the datasheets) for my own sanity, and there will be clear differences obvious to anyone familiar with 3D printer control boards, as my approach to modularity is quite different.
From this conversation it seems that by taking that decision now, and starting from scratch, I’ll be doing myself a favour by more clearly making a new work that is not a derivation of RAMPS, whereas my initial approach could be seen as a derivative work, via my own first version!
From: reading-...@googlegroups.com [mailto:reading-...@googlegroups.com] On Behalf Of James Belchamber
Sent: 06 July 2016 08:31
To: reading-...@googlegroups.com
Subject: Re: [RDG-Hack] Eagle-KiCad conversion
Yeh, it's not an opinion I'm willing to defend to the hilt. I'd argue that the 'source code' is really the file, and if you convert it then the license follows the conversion. However, if you just write it from scratch you wouldn't be subject to the same licence, even though the resultant circuit still looks the same.
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2015.0.6201 / Virus Database: 4613/12565 - Release Date: 07/05/16
Unless anyone has any further tips I will look at Gavin's suggestion, and also try the same in DesignSpark.
I think we all know there are scripts to do this, but they are not very clear on the capability and risks.
I still question why a conversion is needed, Kicad is probably better at PCB routing, Eagle better at error control and libraries. Mixing packages, gives the lowest common functions.
While I try to follow the secondary thread here on licensing, I am concerned that the idea of convesion from Eagle to KiCad in some ways fixes the licensing issues.
Developers under Eagle who use the non commercial license, do so in good faith that the files they publish also follow that license. It would be unethical for users to convert to KiCad to subvert that intent.
I guess I still struggle to see a technical or ethical reason to justify conversion, but am open to fuel discussion.
Classic forum post scope creep J As OP I only asked about the practicalities of conversion.
A friendly reminder spawned an interesting discussion about licensing ethics, which was rather helpful!
I agree it’s clear that the licence of the original PCB design work should be respected throughout, regardless of conversion to another package/format. It may even prevent this, without getting the author’s permission. If you translate Shakespeare into French, it’s still Shakespeare. A conversion of RAMPS to KiCAD would be a useful thing to do, but the moral ownership would remain with RAMPS, subject to its GPL.
It was interesting to get perspectives on whether a work is derivative if it’s similar and carries out a similar function. I ended up realising that the practical barrier to me taking a short cut to a working board by porting RAMPS in KiCAD and then editing it has probably done me a big favour, because the tweaks I made to it would be a derivative, via my own first versions.
Instead I’m going to start from a blank sheet in KiCAD, follow the chip pinouts and datasheets only, and make a new work, and with a different feature set which reflects only my own printer’s requirements. My mindset will be to ignore any legacy conventions and build what function dictates. So while there might be a passing similarity to any of the mass of existing 3D printer driver boards simply due to having similar sockets, it will be a clearly new work.
Version: 2015.0.6201 / Virus Database: 4613/12593 - Release Date: 07/10/16