Would You Be Interested in Roadtesting KiCad EDA?

I have used KiCad since V5 and it's been relatively good. I mean - I didn't love EAGLE and I generally don't like vendor-lock-in. Being free and open source is a big positive.

If it is to be RoadTested, I will echo the requests of others, but slightly differently - I'd like to see multiple board manufacturing runs being paid for as part of the deal. This is because (as we usually know) we don't always get the board right the first time. It might also be good to test different board designs (even if we do get it right) say the use of 2-layer, 4-layer or higher, and or materials (e.g. single-sided MCPCB). It would also be nice to understand how well it handles the manufacturing aspects - do users get their BOMs, CPLs right for the manufacturing house - so if assembly and parts are included, that would be even sweeter. Perhaps this last part can be a one-off batch of 5, but the others could be board-only for 2-layer, board-only for 4-layer, board-only for MCPCB (perhaps with a choice of colour and thickness, just in case someone wants to make a connectorless USB gadget).

As a result, it would become more of a "project" RoadTest ... the costs may vary per participant due to shipping and duties as well, but I can't see the software being given a proper show without bringing it all the way from desk to finalised item(s).

- Gough

With the recent loss of Cadsoft / Autodesk Eagle I think we should start reviewing, roadtesting and assessing what's out there and KiCAD seems to be a good, obvious start.

I'm also sorting out the Eagle Libraries which will help support this (you can import to KiCAD, though I'm aware that there are other libraries / already some in KiCAD, still, figured I'd mention - some people are still using old versions of Eagle that have a freeware license). I'm certainly wanting to bolster our PCB conversations and support on the site.

robogary said:

Maybe a bit off the wall, but wouldnt a nice part of a road test include actually get a 5 pack of their fabs made for participants ? 

I've thought about this, there're logistical hoops for that, especially discussions of "will they be populated maybe?" and what cost that'll involve. Definitely something I'd like to do, but depends on what's designed 

I'd be interested in seeing designs whether they were simple or complicated, or what a 'torture test' for an EDA would look like.

We used to have a deal with them where people could click through for a discount PCB and we'd get a cut of the sale, that went on for a long time though not many took them up on the offer.

shabaz said:

* Build an add-on, or Build a dev-board
This is just extending the Raspberry Pi HAT idea, perhaps to allow the RoadTesters to choose Pi or Pico. The add-on could be a HAT, or perhaps build an add-on or dev-board for the Pi Pico. This might also help in case Raspberry Pi availability is low at the time of the RoadTest.

RP2040 breakout board is also something I'd find interesting, or "<microcontroller> breakout board" a lot of 'dev kits' certainly don't give you access to everything.

Gough Lui said:

I'd like to see multiple board manufacturing runs being paid for as part of the deal

That sounds more like a roadtest of the board manufacturers.

Gough Lui said:

As a result, it would become more of a "project" RoadTest

I call that feature creep ;)

When I proposed this idea, one of the comments back was "well this sounds like a design challenge" and it can easily explode to be one, and I had put it forward as being one (design a raspberry pi hat, since we have a raspberry pi design challenge each year) but there's concern over doing that in being able to 'pull in' the right people, and get the right people's attention to have the knowledge to enter that competition (I'll openly admit, that with our current facilities and methods we struggle to get reach for applicants for the really technical competitions and that makes the team uneasy).

So this would focus more on the software side, with maybe fabrication depending on it, but we wouldn't want to put people off and feel they need to make a completed product. I don't believe testing software should have that as a requirement.

Also I'm going to call ya'all out and say if you're going to respond to one poll, please respond to them all. One poll has 17 responses and one has 20, if you're saying what kit you'd like to have for the roadtest at least respond to whether or not you'd do the roadtest ;P

cstanton said:

roadtest of the board manufacturers

Well .. if it all comes from the same manufacturer, then perhaps it's a bit of a sampler of the capabilities. But alas, the way the EDA handles and exports outputs for each board type may be subtly different. What works fine for 2-layer may become tricky for 6-layer. Regardless, I think there should be an allowance for mistakes that can't be "bodge-wired" away ... it happens to me often enough to admit it ...

cstanton said:

"well this sounds like a design challenge"

Well, it's in the name, EDA, the "design" part. Wouldn't be a proper use of EDA if it wasn't to design something ...

cstanton said:

focus more on the software side

I still feel that the purpose of an EDA is to build the files necessary to generate a physical item. If you don't do the final step, you may never know that you have to hand-massage filenames, tick a myriad of cryptic boxes to make it all work ... I see the physical form as "proof of success". Without it, given the software is free, what is the incentive for the RoadTester?

- Gough

Gough Lui said:

I still feel that the purpose of an EDA is to build the files necessary to generate a physical item.

I agree with you that getting to that stage would be a part of testing the software. I partly replied to robogary about this here  RE: Would You Be Interested in Roadtesting KiCad EDA? 

Gough Lui said:

Without it, given the software is free, what is the incentive for the RoadTester?

That's why we have the question of what would be expected in a 'kit' for this RoadTest, that would be a part of it.

I've also considered that we would offer discount code(s) for the online store, or we would order a basket value of parts from the online store.

And as mentioned, actually fabricating the PCBs. Then there's also being given the platform to share your opinion about the software and for that to be shared.

Ahh.. a RP2040 board would be very neat. Although currently each time I end up using the Pico since it's fairly compact anyway, but it would definitely be a fun project to at least attempt a RP2040 board sometime, and in particular resolve the 1.8V GPIO thing (i.e. allow the GPIO to be 1.8V, which is not possible on the Pico board without a level converter, which isn't easy when you're quickly prototyping and would rather just have it on the board already). A RP2040 with adjustable capability (1.8V and 3.3V) would very useful. Also with a reset button on it, and space for another RP2040 with the PicoProbe!  I can't live without that feature now, ever since Jan Cumps  suggested it for the Pico Eurocard.

Incidentally I'm currently making a Renesas RA4M1 board, because the ready-made one is a bit awkward for some things (particular LCD wiring). So this board will simplify that a bit. I'm still working on the RA4M1 symbol, since it's a 100-pin part:

   Hell yes!! This is my daily ecad driver.  This is a drawing of my Fire Switch Interconnection 3 fire switches to one DB25, I have modified the plan to use Molex Connectors to the fire switches not the interface. Then I need a 4-pin Molex for Lamp Power and Lamp Ground

Each Fire Switch has 4  SPDT switches  - 1 common 8 contacts, Lamp Power & Ground.

I really only care when the switch is on in which case the switch's NO pins will go to ground.