I was considering it because N64 and PSX run so much better on it but now I'm not so certain now since the issues have been pointed out. So with that being said, I'm curious if anyone will be using the Pi 3 and if so, how do you plan on doing so? Share your thoughts below.wermy wrote:I really, *really* wouldn't recommend a Pi 3 for this. Those things use such a crazy amount of power and generate so much heat. Just sitting there on my desk, with a heatsink on, my Pi 3 shows the temperature warning image on the screen while compiling large codebases. When that happens it throttles the CPU. I literally had to point a desk fan at it to get it to stop. ಠ_ಠ
And then there are the power requirements. Under load the pi 3 can consume upwards of 700mA, compared to the Pi 2's ~400mA
http://www.pidramble.com/wiki/benchmark ... onsumption
Not to mention that this is JUST the Pi itself, no screen or amplifier or anything.
For reference the guts of my GBZ build top out around 450mA total.
If you went with the Pi 3 (assuming you solve the heat issues) you'd easily be exceeding 1000mA, I would bet, which would make it so even the 1000mA power boost probably couldn't keep up with charging and playing at the same time.
And your battery life would be awful. Sorry to be a naysayer but I just don't think it would be a good option here.
Pi 2 would be awesome though!
UPDATE: Adafruit confirmed https://forums.adafruit.com/viewtopic.p ... 2049786661 that the powerboost 1000c wouldn't work and that they have no plans to upgrade the current version.
They did say that two regular powerboosts can be used as an alternative and explained it belowThe PowerBoost 1000C can only supply 1A. The Pi2 needs ~600mA. the Pi3 needs more like 1.5A
The PowerBoost 1000C uses an MCP73871 LiPo charger which also does dynamic load switching. When an external supply is connected, power goes to the load first, then any current not used by the load goes to charging the LiPo. When you unplug the wall supply, the chip automatically starts drawing power from the battery.
That approach makes the MCP73871 good for devices that charge while operating. It also means you can plug in a device with an empty LiPo and have it boot with full power from the wall supply.
To make that happen in the PowerBoost 1000C, all current to the boost circuit has to flow through the MCP73871, and is subject to the limits of that device. The major limit is that the MCP73871's internal switch is rated for an absolute maximum input current of 1800mA.
The boost circuit uses about 150mA @ 3.7v of input power for every 100mA @ 5v it puts out. Scaling the MCP73871's 1800mA limit by that amount, the PowerBoost 1000C can only put out about 1A @ 5v before you start facing the risk that you'll kill the MCP73871. To avoid that, we designed the boost circuit to max out at 1A of output power.
The PowerBoost 1000 Basic doesn't have the MCP73871, so the boost circuit isn't limited by that internal switch. The TPS61030 has an internal switch rated for 4A, so it can deliver 2A @ 5v without pushing the limits of the hardware.
You'll need an external circuit to charge the LiPo though, and you won't get any of the advantages the MCP73871 offers, like load switching. If you're pulling 1.5A of output and have a LiPo charger with a 1A limit, the boost circuit will be pulling 2.25A of input current, depleting the LiPo even though you have the charger connected. If you want to charge the LiPos at 1000mA while running a v3 at 1.5A, you'd need a charger set to about 3.5A.
No load switching LiPo charger we know about can handle that much current. Building a load-switching circuit that can handle enough current would be a hard job, and would make the PowerBoost larger, more complex, and more expensive.