I have not seen any complete information about the default power management with the latest image.
Occasionally and after several hours, the Edison seems to stop, perhaps going into sleep mode.
Is there any way to control that, using a startup script perhaps or programmatically from an application?
Did you notice this behavior with the latest image only or with previous images as well? How do you know that the Edison goes into sleep mode/stops? By this I mean if you're running a sketch or anything similar that lets you know when the Edison stops. Depending on your situation it should be possible to get it back to work with a script. We will appreciate your details on this.
I am using only the latest image (JUN-06). I am not entirely sure what is happening, I thought it could related to some power management feature, if I can disable it may prove that it is the cause. This is why I would like to know what power management features I could control on the Edison.
This is important as well as the other issues on other threads, as we will be deploying hundreds of Edison based devices on the network and I need to resolve these issues or at least understand them so we can address them accordingly.
Any information you could provide would be greatly appreciated. Since we will be using many of these devices I would also like to have a direct contact with some of your engineers to help in dealing with such issues.
The Edison's PMIC is the SNB9024, provided by Texas Instruments, so there are not many power management options available as this is a non-distributor item.
However, if you're having issues with the Edison's UART operations it might be related to this known issue: when the Edison enters low power sleep, the UART internal FIFO and interface is powered down. This causes the two-wire UART (RX and TX only) will lose the first received character. To avoid this, you should use four-wire UART (RX, TX, CTS and RTS).
Sure! Check this link for further instructions /thread/61067 https://communities.intel.com/thread/61067. There was a known issue with this feature where you could only get it to work once, after that it wouldn't enter the sleep state anymore. I believe this has already been fixed in the latest images, but if you still find problems with this there's a patch for the driver. In the same thread you'll find instructions for its implementation.