Yes I've tried changing both bits 0 and 1 to no effect.
I was following the recommendation on page 2204 of the TRM for disabling unused USB PHY and it only specifies setting bit 0. I'm still puzzled that changing the bit doesn't stick, why is that?
When I follow the recommendation on page 2204 the first 2 steps do show a ~10mA drop when applied to both USB devices.
Through trial and error I think I've determined that most of the clocks are already disabled, although I never got a response from you on the best way to go about disabling the clocks. Furthermore, I'm not entirely convinced that just turning off the clocks is enough. For instance I tried turning off the clock for the LCDC and that had no effect on power usage, it wasn't until I disabled the LCDC interface in the device tree that I saw a 50mA drop in current.
Ultimately I'm trying show that we can get to ~750mW total with just Ethernet, 1 UART, and a couple of GPIO pins enabled, currently I'm just under 1000mW, I'm measuring at the 5v DC power connector. The power estimation tool appears to indicated that I should be able to get down to ~250 mW although that probably doesn't include the Ethernet PHY and magnetic's which I believe consumes around 250mW. So there appears to be around 250mW being consumed by something(s) on the BBB, that or the Power Estimation tool is off by a factor of 2 or more.
Any suggestions?
Thanks
Matt S.
