New to this forum and have a newbie question. Bought a Panda II a while back and build a project that I am quite pleased with. Showed it to a couple of business type folks and there is interest in possibly taking it to market. Problem is my Proof-of-Concept (POC) looks like a typical “spider” hobbist project/mess.
The project uses surprisingly little of the base Panda II core capabilities
the USBizi chip (of course)
some I/Os and the
the clock
a couple of UARTs
etc.,
But that is about it. I don’t for example need:
USB support
SD card reader
A shield (and therefore a shield connector)
etc.,
So my question is how does one go about taking this project and “reducing” it to find out:
a) How small it can be made? (I have very tight encloser constraints)
b) How cheaply it can be made (i.e.; minimal parts)?
c) Create a minimalistic PBC protoype layout for possible manufacturing?
I have been reading this forum all weekend and have learned many things including the existance of services like the “Fritzing” and “DorkBotPDX” services. But these strike me as addressing only c) above. I think I first need to address a) and b) by “reducing” project to its bear minimum requirements.
As I come from the software world I don’t have the skill set to deal with the electrionics myself (which is why I was attracted to the Panda II in the first place), but I don’t want that to stand in the way of making progress here. So I am looking for information on how to best proceed here. For example should I:
Take a soldering iron and start “removing” stuff from my Panda II board until I break the project (half joking here )
Get a breadboard and an USBizi chip and start building from the ground up (how much electronics knowledge would one need to do this??? - Its not clear to me how much of the functionality I need is already implemented directly on the chip v.s. implemented in circuitry on the Panda II board)
Find someone here in the forums to “reduce” my project for me for a fee
Contact GHI to see if they offer these kinds of services
Thanks for the tip on Cerb40. Looks very interesting.
I read the “software is not ready”. I assume that means the drivers?? When they are completed can I assume that my Panda II program would run on it as well (just simple i/o stuff)?
One of the key items you need to understand is potential market. Are you making this device in tens, hundreds, thousands, or what? That has a significant impact on how “rewarding” this all is.
Design:
The Panda and Panda II design files are available in Eagle format, so you can take them and strip out the connections you don’t need and make a credible start of a design. DorkBotPDX could then be used for short run production of circuit boards, and larger places could potentially be looked at for production runs.
Bill Of Materials (BoM) and cost to manufacture:
The standard Panda BoM is in the design files. You can price your own costs based on sourcing the USBizi chip from GHI then the rest of the parts from any other large electronics supplier.
Assuming you have a demand worthy of the cost, you can take that BoM and your board and approach board assemblers. They’ll then figure out how much it is to assemble and perhaps test/load the software you want.
For your sample boards you can do all the assembly at home, you could even do that for short runs as well.
Finally, of course GHI are the people who know the Panda II design so you could engage them to redesign to your spec, they would have sources for appropriate components, and their new equipment line can potentially do the assembly etc. So depending on how comfortable you are with the tools and processes above, you could reduce the risk by talking to GHI.
I have read about the Cerb40 which is “open source”. Then you say that the Panda II has Eagle files available. Are these “free”? Does that mean the Panda II is “open source” (I assume not!). I understand what this term means in the software world but am new to it in the context of hardware and ICs. Can you explain what “open source” hardware is?
Finally, if I had an Eagle file, could I bring it into something the “Fritzing” tool and begin to “reduce” it to create a minimalistic circuit?
Panda II design files are “open” and GHI provide them - you can check out any licensing restrictions in the downloads, but GHI have always supported people taking those files and enhancing / customising the design to their own needs.
The Panda II uses USBizi 100 chip, and the netmf port that this contains is not open, it is GHI’s IP. The Cerb port has been opened to the community, but as you see it has not been “completed” and it is likely it will not contain all the features of the GHI port on USBizi for a while, as there were some significant features GHI delivered in only their firmware
Fritzing is not really what you need here. Fritzing would help you show how your current Panda II was connected to real-world devices, but my expectation on what you would need to do is to take the Eagle files, strip out what you don’t want, add in what you need (for instance UART connectors direct on the circuit board or perhaps RS232 drivers if you use that) and then make the board from that. Fritzing is a visualisation tool, not a design tool.
If I were actually going to market with a NETMF device, right now, I’d almost certainly do it with USBizi, and if I weren’t an EE (and I’m not…) I’d probably spend a bit of money having GHI help out with my hardware design.
They’ve done a lot of this, and they know how to make it solid.
Cerb40 is SO young, I’d hesitate to base a commercial product on it quite yet, unless it was fairly small-run. We don’t even know if Cerb40 has IFU yet (does that require TinyBooter?), so you might have to go around shorting BOOT0 pins and using ST’s DFU tool to update the thing.
