Well, I assume the limitation is software, hence hackable, inasmuch as the process of negotiating how much power a device may use is a software process.

Basically, the device is permitted to use 100 mA when first plugged in. The device asks the host "May I use some more current?" in steps of 100 mA (I think), up to 500 mA. At any time, the host may say "No", and the device is not supposed to use any more power. Physically, no extra current is "switched on"; it's just a way to avoid devices using more current than is available.

So the question is whether more current may be drawn without either:

• Burning something up
• The voltage dropping too low
• The host noticing and shutting off power

Don't know if any of these happen, and at what point, with the N8x0. But I'd guess the driver could be hacked to let you try it... if you're into extreme sports.

Of course, this causes some problems for use of power-hungry peripherals. The simple, light-weight solution is a power-injector in the custom OTG cable that you pretty much have to make yourself. 9V battery + voltage regulator, and you're electrically good to go.
But the N8x0 will not permit devices to use more power, because it doesn't know about the power injector. You then have to use a powered hub. (Then the power is known to be coming from the hub, so the negotiation works.)

That's where a hacked driver could be useful, if used only in conjunction with a power injector.

I'm guessing this is an open-source section of the kernel? But it could be one of the proprietary binary-only modules, in which case hacking it is far more difficult.


Yes, I think a USB sound card should be plug-n-play. Only one way to find out though, and I don't have a USB sound card.