I don’t know about the current draw, but you are correct about using the gnd/vin pins on the edge of the board, and in that case, the jumper doesn’t even come into play.
The VIN/GND pads can take 3.3-17v, they are just not reverse-voltage protected as they are on the pack pads. They are just fine to use, though.
The current you can get to VIN from USB is limited by the max current of the connector (1A?) and the 0.1" header (I wouldn’t put more than a couple of amps through there).
The current you can get from P+/P- is limited by the header AND the reverse voltage transistor which will be better than 135mohms at 5v (spec is 135 at 4.5v). Say you had 2A, then dissipation in this package would be IIR = 220.135 = 0.54W. The package can dissipate 1.3W at 25C, so that’s within spec.
Realistically though, the traces on the PCB aren’t likely to be good to supply more than 2A. The imp itself uses an average of under 100mA… what’s the other current going to?
I am building a weather display which uses a string of around 300 Neopixels. I was going to power them from the output of the April board and power everything from the P+/P- pads. The Neopixels are RGBW so if all 4 portions of all 300 lights (rare) are set to maximum illumination, they could draw nearly 10A.
I’ve since decided to power the Neopixels and the board in parallel but was trying to find an easy way to connect to my completed board. Luckily I can power the VIN/GND bus using the blue connector which was originally only intended to connect to the Neopixel string. Now it can also reverse feed into the April board and everything should be fine. I’ll be sure to double check the polarity before powering it on.
