Single-Board Universal Module

It's official.

This thread has been hijacked. I'm unsubscribing until PB, or others using his Universal Module board, has some progress on it.

RedSleds wrote: It's official.

This thread has been hijacked. I'm unsubscribing until PB, or others using his Universal Module board, has some progress on it.

Hi RedSleds,

I'm very sorry about that!
I move thread to here .

I've been following this thread since day one and I'm quite excited about what's going on here. I also want to build a universal module, but I can't even compile the code.
I've checked out the latest master from github, and building the devo8 target works fine, but when I run make in the universaltx dir, it can't build the libopencm3 linker script:

BUILD lib/stm32/f0
make[2]: *** No rule to make target `libopencm3_stm32f0.ld', needed by `/home/ali/src/rc/devo_8s/deviation/src/libopencm3/lib/libopencm3_stm32f0.ld'. Stop.

I'm sure I just forgot something stupid (as usual), but as I haven't used libopencm3 before, I'm a stuck at this point.
Anybody care to help?

Are you building on Windows? If so, start bash and build first time from there, it should build (if build is not broken at all).

No, I'm on Ubuntu 14.04. But I've tried explicitly starting a bash (as Ubuntu has some kind of other semi compatible shell) and build in there, but it throws the same error.
I've also tried bringing my own libopencm3, but that was no good either. With that not even deviation would compile..

Edit:
I got it to compile now. I had to copy the f0 linker script over from the same version libopencm3 and move a few commits back. I still have to figure out which commit breaks the universaltx.
Edit 2:
It was e4d482c2c50551b367c17ea12f29f234a30fc9fd coverity fixes.

submitted a pull request

octagon wrote: Why not download a copy of EagleCAD?/

'cause I don't wanna fill my rare discspace with GBs of mud..

octagon wrote: Gerbers are already available for the project, or you can just order the boards from OSHpark./

I know.
But they do not help me.
I want to build some of my own and need only two ways/modules, Hubsan and FRsky.

Last year I learned that some members of CERN jumed on the KiCAD train, so that's the way to ride.

I ordered a big part of the BOM a year ago, and have three 0.9.4 boards on my desk, so I'm committed to building one. BTW, would this code (sans USB and all the protocols) run on a smaller ST, say an stm32f030c8t6? I'm asking because I have a few of them around and I'm not planning on using the module standalone.

I have KiCAD and I never got used to it. I'm not saying it is total crap, unless something major changed, KiCAD can allow logic errors.
bugs.launchpad.net/kicad
Maybe the scientist at CERN can fix the bugs.

@PB:
i tend to forget there even is USB on many ST chips. I'm still too stoked by what you can do with SWD, I guess. Anyway, I ordered a handful of 72s to be sure. The parts that go on that board are way too precious to waste, and I have to wait for the PE switches anyway. I'm really looking forward to building this thing and even more to finally being able to retire the porcupine and put the original antenna back on.

re Eagle vs. KiCad:
I just switched from Eagle to KiCad, and after using it a bit i must say they both suck, but in totally different ways. KiCad's OpenGL Workspace and the push and shove router work really nice, but doing anything involving a curved line and a copper surface is a major pain. It can be done, but it usually involves editing the kicad_pcb file in a text editor or generating objects with scripts. But then, doing almost anything in Eagle, esp. switching layer visibility, needs so many clicks, it drove me nuts. The input prompt remedies that a bit, though. And board artwork like curved copper surfaces or board outlines didn't involve as much swearing as with KiCad, I recall.

And what product does not suck?

My job is a pure programming one, so I'm not so lucky - that's why I ask.

switching layer visibility.
It is not too bad, but I wish I could set "hot-keys" for groups I can pre-set for visibility. Color choices are also too limited, as is transparency. I use open source "gerbv" for checking designs, to get another view of the process.
Altium is the "best", but I could spend $11,000+ on something I need more.

Many swear by DipTrace, which has a comparable price range to Eagle, but all I ever did with it was export some Gerbers or PS files with the trial version. But Iǘe seen people on the net call it intuitive. Maybe worth a look.

PhracturedBlue wrote: If you want Gerbers, the easiest way to get them is to install Eagle, then generate them, then uninstall Eagle.

I do not want Gerbers, I do want the schematics.
I was told from eagle-loverboys even from a couple og KiCATs more than a hundred times: install eagle and you can export your beloved schematics easily..
My answer is always the same: NEVER EVER!

I will never understand how people uses not OpenSource stuff like Eagle/Altium in OpenSource/OpenHW surroundings.
But times are changing, more and more developer are switching to KiCAD or runnin' it aside their evaluated system.
I know, KiCAD has a lot of faults and glitches and some parts are really bullshit.. but it's OenSource and worth to be developed further on!
That's all.

BTW: I used gerbv from gEDA until I got knowledge of ZofzPCB.
Ok, sometimes it crashes, it seems to be a one-man-show, but I don't wanna miss the peeling function anymore.

I've looked at DipTrace, and I could not get along with it. Maybe I should have spent more time with it.

Gerbv gEDA works fine, I have offered many suggestions to improve it. Project seems dormant. Specifically I want to assign colors to file extensions, and layer ordering, so opening layers always looks the same. Now random cute pastels colors in random layer ordering demands that I recolor and reorder all the layers.
Once saved its OK for one file. Error handling is weak or non-existent in gerbv gEDA.