Upverter (YC W11): the perfect tool for open-source hardware
venturebeat.com
venturebeat.com
There is a lot more to your average EE CAD package than just wiring up some parts. That's the easy bit. Rules checks, physics, placement, packages, PCB design, (auto)routing, Gerber files and all the other 'goodies' are required before you can take on the big guys.
Schematics is an important element in the whole chain of developing hardware, but it really needs a lot more than that.
That's a large number of man-years, and by going the software-as-a-service route they gain some advantages but are also stuck with a number of disadvantages, the largest of which is that they won't be seeing the benefits of user contributions at a level where it will move the needle. Of course there will be contributions to the parts library that can be shared, which is a step.
The 'perfect tool for open-source hardware' would benefit from being open-source itself.
Open source hardware is no good if I can't manufacture it on a circuit board. Yes, breadboarding is handy for small stuff, but 256 pin BGAs don't fit on breadboards. And repeatability isn't so grand.
With a nice, reasonable to use layout tool (even with just simple routing assistant type features [push and shove]) and tutorials on how to use it, what DRCs are and how they're used, and maybe videos showing people who know how to do this stuff teaching, upverter could be really cool. Heck, for a full-ish featured EDA package in the cloud targeted at open hardware, I'd even pay to use it (provided we're not talking Cadence or Mentor money, more like GitHub paid account money).
The hope is to also add a "print button" to connect users designs with rapid manufacturing suppliers.
Best of luck to you guys.
Agreed. Open Hardware needs fewer closed tools, not more.
But I think they're coming at the wrong end. Their schematic capture was painfully slow when I tried it a couple weeks ago, and I (and I suspect nearly every other electrical designer) rues having to change tools. If someone wants to become the 'github of electronics', I strongly recommend working on getting some sensible versioning and sharing going without also tackling the problem of creating EDA tools.
Having said that, a shared crowdsourced library is a big help, and some others have already noticed the good it can provide: https://github.com/adafruit/Adafruit-Eagle-Library
We just pushed a major update publicly yesterday that had some incredible performance gains (50 - 100x faster). We have also done a bunch of selective rendering updates which make it feel a ton faster to use.
You should give it a try again and let me know what you think! Hopefully we are a lot closer!
Most EDA software is written by companies who have no interest in letting you see how their proprietary formats work.
Oh boy, I know it, and it's been a pain in my side. However, EAGLE (lots of hobbiests use it) is migrating to a fully-documented XML-based format [1], and there are two open-source projects for schematic capture, kicad and geda, which would give someone a head start on parsing their design files, at least.
Other than that, this looks really awesome!
Diffing has been talked about internally and it would be pretty awesome.
We're not aiming to replace the tools you already use, just make it easier to share your work on the web. I do love the idea of a web based EDA tool, but I'm yet to believe it can be done with performance characteristics that will work for any reasonably complex circuit.
Bonus points for parameterising the component selection and adapting to current design rules!
Second, separate the parts into different libraries! Other EDA packages have thousands of parts, it would be a mess to have all in one list, even with filtering. It seems you are loading all the parts when creating the library dialog? Probably a different approach to make it manageable...
Preview for the parts, including PCB footprint and variations. What do you plan to do with parts with different blocks, like logical gates, power pins, etc ?
When adding nets, allow the user to move parts around.
I have a few more, but I'll wait when you get layout, drc/erc , auto routing and pcb webgl 3d view working ;) Besides this, keep the good work!
You can try to look at the mailing list archives of the open source packages gEDA, PCB and kicad. There's a trove of nice info and discussions there. I contributed long ago (1998?) to gEDA and use these , eagle and others more or less frequently.
I don't see any support for hierarchy in a design. For scaling things you need that. You advertise pages that are as large as you want, but that's more a hindrance than a help, I think, because you can't print an infinite-sized page. And how do you print? How do you get the data out at all - is that possible? In order to do layout you'll need to be able to produce a netlist, and the community-provided parts library is not so great if it doesn't carry over into the physical world of PCB layout. It would also be good to be able to do a simulation, but that is going to really be tough on your servers.
It would be really cool if you didn't have to sign up to try it, that stopped me in my tracks even if it did look like an easy form to fill out.
It'd be cool to go from the end of a search via octopart ui to a new symbol in upvert without cutting and mis-pasting part numbers.
I'd be interested in learning more about this if you wanted to share.
I'm a Altium user, can't really use it on a iPad. Upverter can go almost everywhere!
U-pervert-er