Particle Electron PMIC Efficiency

Hi guys,

I did some testing with the Particle Electrons’ PMIC and got some results that differ from what I would expect from the PMIC datasheet. According to the datasheet of the BQ24195 it states the following efficiency chart (see attachement). It’s clearly visible that at a current draw from approx >1A , 5V input voltage is ± 5% more efficient compared to powering at 12V. Now I ran a power consumption test (see chart second attachment) and checked it both with a multimeter and power supply power draw. It seems that powering at 12V is almost 80% less efficient compared to 5V at low load?

My question: what’s the reason for this? A possible explanation could be that the PMIC draws more current at 5V compared to 12V and the lower the current draw (sub 50mA) ,the more power inefficient the PMIC becomes(operating at the extreme left part of the graph “gray zone”). But i’m not really sure if that’s the real reason. The reason I’m asking is I have 12V available but I want to power the electron as efficient as possible (without bypassing the PMIC). If this requires adding an extra switching regulator to 5V, so be it.

Test setup: Particle Electron connected to the cloud, no battery (load) connected, empty loop code running

Attachements:

@KyleG, this is one for the Electron engineers :wink:

I’ve looked at alot of DC to DC converter efficiency graphs while picking the most efficient options for some product designs which I want to be as efficient as possible and your hunch is correct.

The converters are usually designed to be very efficient in certain ranges but not all ranges and that’s the reason you’re seeing the differences in efficiency.

Plus the TI efficiency graph was created using a certain ideal circuit layout and components which may or may not be exactly what Particle choose in their design so the efficiency graph of the actual Electron layout may be different from what you see in TI’s datasheet.

You could choose a different converter with higher efficieny if you desire.

Thank you for your answer @RWB! @KyleG Can you confirm / elaborate on the answer of @RWB?

Hello,

Let me ping someone that might be able to help, @rickkas7 are you able to assist?

Kyle