Glad it is working! Unfortunately, I have heard your story before. I can’t explain what exactly is going on, There may be some issues with flashing process. What those are, I can’t say. But, sometimes it takes a few times to get it to take.
Gerald
One thing I’ve noticed is that, while flashing an image without the 5v/2A DC PSU resulted in random failures, I’ve not had a single failure when the 10W PSU was used. Have a feeling there are power requirement spikes during the flash process, where the 500mA (max via USB alone) could have been insufficient.
Just an empirical conclusion… Maybe hooking up through a DMM could share more light…
Wondering, are there any events/interrupts handled (and logged) by the s/w when there are “power failures like these”. May shed some light.
/venkat
Thanks for the information! I am happy to take all the information I can get. Can I have the make and model numbers of the power supply?
We have talked about lowering the clock speed during the flashing process when on USB. That will lower the power a little but it would also increase the write time.
Any fluctuations in power would be hard to track.
Gerald
The Atmel docs say that a single ATMega with the interrupts all firing can
spike to 400 mA. The PS caps are supposed to be able to handle it but that
nevertheless supports the idea that it could be a power issue.
Britton
Since I got my bbb @adafruit, I also got their recommended (worldwide use) 5V/2A psu:
http://www.adafruit.com/products/276
at the same time. So far, works fine. fwiw, even tried flashing with a fully charged 5V/2A battery pack:
http://www.amazon.com/Dynamo-On-RP-PB04-Charger-Powered-Cells-Includes-Thunderbolt/dp/B008FV350I
and the single attempt did succeed. This was with the latest 05/08 image.
/venkat