Hi. I'm Nick, and I'm the guy at Geppetto Electronics. I build little electronic things to solve problems. The big issue is that sometimes I come up with things that I think have a wide applicability, but it's too costly to build a bunch of widgets up front, where scale can result in greater efficiency.
The problem at hand is the Raspberry Pi power problem. The Raspberry Pi comes with a surface-mount micro-USB socket. You're supposed to connect up a high amperage 5 volt supply to that. There are a variety of problems with this. Among them, the surface mount connector is relatively weak. It's also fiddly (can't rotate the plug), and good quality, high current USB supplies are not always easy to come by (by contrast, crappy ones are common).
Pi Power is a switching DC-DC converter that will accept any DC input voltage from 6 to 15 volts and convert it to up to 2 amps at 5 volts. It connects to the Raspberry Pi's GPIO connector, and comes with a stacking header so you can "piggyback" other things (and power them) as well.
Pi Power works with all current models of Raspberry Pis. On the model B units, you can make use of the extra power for USB peripherals. On model A units, to use high powered USB peripherals you still need to either modify your Pi or use a powered hub. On B+ models with the 40 pin GPIO header, you can add a 14 pin stacking header to the bottom of the GPIO header next to Pi Power to raise the complete GPIO header up to the same level, or you can use the 24 pins above PiPower with peripherals that don't use the extra pins.
What We Need & What You Get
I want to fund a volume purchase and build of Pi Powers. At this point, I have a working prototype. In fact, I sell them for $15. But each one of those I have to hand-assemble. If you include my time, I'm definitely losing money doing it that way.
Here's what building 1000 of them would cost:
- DigiKey BOM: $3900 (T&R or DigiReel).
- PCB Fab: $1600 (OSHPark medium-run order for 100 panels of 10 boards)
- Stacking headers: $370 (4uconnector)
- Assembly: $1410 (Small Batch Assembly)
There's really only one perk on offer here - a Pi Power for $15. Anyone who contributes that or more will get one. They'll be supplied just as they come off the panels, plus a stacking header for you to solder on yourself. If and when 486 people do that, the campaign will be fully funded. From there, Pi Power should be self-sustaining at the thousand-unit per batch level.
My initial estimate on the timing is that from the moment I shout "Go!" initial deliveries of the completed product should be within 3 months. I believe the PCB fab requires 4 weeks lead time, and that the assembly does as well. That leaves a month of slack for delays.
I believe Pi Power is a good product. The major reason I am going through this is that I believe large numbers of Raspberry Pi users would buy a Pi Power if they knew about it, and the visibility of the project and the efficiency gains from volume purchasing are the goals here.
With a total investment of $7280, I could build 1000 of them and purchase the parts in whole-reel quantities for maximum savings, and have parts left over on some of the reels to reduce the per-unit costs of the next batch even more.
Risks & Challenges
There's very little risk. I'm building these one-at-a-time right now, I'm just not very efficient at it. If it all falls through, then I just go back to making them by hand in my garage.
Other Ways You Can Help
This campaign really only works if we reach - or at least asymptotically approach - the goal of building and selling 1000 Pi Powers. The more people hear about this campaign, the more likely we are to meet or exceed that goal.