GPIO0 trough 2 shorted to ground

Andy Green andy at
Fri Feb 22 13:51:14 CET 2008

Hash: SHA1

Somebody in the thread at some point said:
> When playing with the "gpio" tool, I noticed that the GPIOs driving
> the LED transistors are effectively shorted to ground.
> This causes a few curious effects, for instance that the pins read
> back as zero even if set to one. This in turn means that the usual
> way of modifying bits, e.g.,
> gpiob |= 1;
> will clear all the LED GPIOs that are not explicitly set by it.
> This shows by the corresponding u-boot command acting strangely.
> (So anyone testing the LEDs with u-boot should be aware of this
> issue.)

Ha nice gotcha.

> I'm not sure if all this causes any real problems (except for
> drawing a lot of current, e.g., about 50mA per lit LED instead
> of roughly 10mA, bringing idle current from ~120mA without LEDs
> to ~265mA with all three LEDs lit), but it's certainly not pretty.

What's notable is we might as well save current by throwing out the NPNs
and drive the LEDs directly then :-)

A fix we might be able to do is change the NPNs into N-Channel MOSFET in
same package.  Then we will have pretty much zero drive current at the
CPU and no rework.

- -Andy
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Fedora -


More information about the openmoko-kernel mailing list