Hi,
I wanted to test Tonys 4GB-RAM patches on various PCBs I have here, however, I couldn't get them to boot using only the console port to power them.
I tried the following setups:
1. PC connected to console port, no battery: Switches on, shows U-Boot message on terminal and immediately switches off again.
2. PC connected to console port, with battery: Switches on and boots.
3. AC Adaptor connected to console port, no battery: Switches on and immediately switches off again. (so same as 1.)
4. AC Adaptor connected to USB OTG port, no battery: Switches on and probably boots (keyboard backlights stay on) (so same as 2.)
5. AC Adaptor connected to USB OTG port, PC connected to console port, no battery: Same as 1.
6. AC Adaptor connected to USB OTG port, no battery. Switching on: Boots. Connecting the PC to the USB OTG port: switches off again immediately.
So conclusion: As soon as something is connecting to the console port (PC or AC Adaptor), the unit switches off immediately again if there's no battery connected. And the console port seems to have a higher priority than the USB OTG port.
This makes testing PCBs without a battery attached impossible.
The question is: Why is this happening?
It was definitely working with the previous prototype revision (as that's what Nikolaus and me worked on when me met an afternoon at GC), so where is that regression coming from?
notaz also tested it with the latest U-Boot from Letux (http://download .goldelico.com/letux-u-boot/Pyra+LC15/) with a 2GB board, but the same thing happens.
Is this a regression in hardware or something that's wrong in software?