Am 03.09.2014 um 14:04 schrieb Michael Mrozek:
Am Wed, 3 Sep 2014 12:16:15 +0200 hat "Dr. H. Nikolaus Schaller" hns@goldelico.com geschrieben:
Hi,
well, not only moved... it seems like the material is not yet milky enough, so it doesn't spread the light properly.
I'll ask if that can be improved.
Fine!
Here another (mostly hardware related) feature has been tested:
The TPS2505, the "wonky" chip which has heavily inspired the community :)
http://boards.openpandora.org/topic/16728-tilted-chip-t-shirt-design/#entry3...
is working. It's purpose is a DC/DC converter to supply 5V DC to both USB host ports and the HDMI (and drive the VBUS line of the modem's USB interface).
I can see 5V DC on the VBUS lines (although the whole board is powered from the 3.7V VSYS from EVM expansion connector). And it appears that they can be controlled by software (it is a little difficult to test because the reg-virt-consumer driver is not yet Device Tree compatible).
I could even use a GTA04 connected to the USB host port to monitor the VBUS voltage (through the TWL4030 in the GTA04) which gives a very strange power train:
Openmoko USB charger -> adapter (powerplug) -> OMAP5432EVM -> 3.7V regulator -> 100-pin adapter board -> flex-cable -> main board -> TPS2505 -> USB-A socket -> cable -> GTA04
One important lesson I have learned for the next PCB version is that I need to add some pull-down resistors to the TCA6424A which enables the DC/DC. Since it's reset state is "input", the "enable" lines are floating until the kernel is initialized and that makes the DC/DC converters start right after power-on.
And a new idea: the FAULT outputs (if a short circuit is applied to the USB host) can easily be connected to the tri-color indicator LEDs near the POWER button. So they will light if someone draws too much current. Another option would be to feed them to the CPU - which allows a more flexible handling (e.g. a popup window) but eats up two more GPIOs. Although it is possible to read the LED driver pins and check if the LED was turned on. So if we save GPIOs we still can detect FAULT by software (in parallel to the hardware LED indicator). So this needs some more thinking until a final decision can be made.
What I have not yet done is a stress test of the TPS2505. I.e. checking if it really can supply 1 Ampere on each output.
Am 03.09.2014 um 10:16 schrieb Dr. H. Nikolaus Schaller:
Hi, some good news from testing the Pyra main board:
- BMP180 (altimeter) is working out of the box
- shoulder and nub buttons are working
- hall sensor for lid detection is working
- I have fixed the as5013 driver (for 3.17 compatibility) and now
the nubs are working :) http://git.goldelico.com/?p=gta04-kernel.git;a=commit;h=683bf386827537cde41a...
Note: this is just 10% of what has to be done...
Next thing made working: Keyboard backlight
-- Mit freundlichen Grüßen,
Michael Mrozek
OpenPandora GmbH Geschäftsführer: Michael Mrozek
Schäffbräustr. 11 85049 Ingolstadt Deutschland Tel.: 0841 / 13 177 33 http://www.openpandora.de/ HRB 4879, Amtsgericht Ingolstadt
ICQ: 15691903 MSN: EvilDragon1717@hotmail.com eMail: info@openpandora.de _______________________________________________ Kernel mailing list Kernel@pyra-handheld.com http://pyra-handheld.com/cgi-bin/mailman/listinfo/kernel