FED3 power consumption measurements

[Lex Kravitz]

I purchased a cool power meter called the Otti Arc and did some power measurements with FED3.  This may give some insights as to how FED3 consumes power and how much power can be saved with sleep modes.

Here is the power consumption while FED3 boots up with sleep modes enabled.  Given that FED3 spends the majority of its time idle, the most important number is the idle current draw, which is 8.1mA.  FED3 has a 4400mAh battery, which should theoretically give FED3 an idle standby time of ~3 weeks.  With normal use the battery life is much lower than that because FED3 burns a lot of power being awake, moving the motor, etc.  But we often get ~7-10 days of battery life from FED3 between charging.

Here is the same recording with sleep disabled.  Now the idle current draw is 20.6mA, which works out to ~9 days of idle time.  Without sleep I tend to get ~3-4 days of battery life.

So using sleep modes should at least double battery life.

Nik Hayes asked about a super low power FED3 that would turn everything unnecessary off.  I think this would require a redesign of the electronics to truly make an impact. Of the 8.1mA being used at idle state, ~7.8mA are from the FED3 PCB and 0.4mA are from the Adalogger board.  The 7.8mA are used because the FED3 PCB has three IR LED beam break sensors that are always on and active, and the Neopixels have microcontrollers in them that burn a couple mA even when off.  So messing with the code cannot impact these unfortunately.  You might be able to share a few microamps with code tweaks but it is not likely to produce further improvements to battery life.

Hope this analysis is interesting/useful!

[Nik Hayes]

Interesting, thanks for the info Lex! Maybe I'm misunderstanding, but in Free Feed mode sleep goes on after 1minute of a pellet not being taken? If so, I guess the screen stays on but there's no way to "turn it off" in the meantime?

Thanks,

Nik

[Lex]

You're welcome, I love thinking about this stuff!  

The screen is a Sharp memory LCD display that uses very little power, in the microamp range. So turning off the screen won't change much.  The main power drain comes from those three IR beam breaks (both pokes and the pellet well) and the neopixels, which together burn ~7mA when idle.  Unfortunately these cannot be turned off via code alone.  We also need the IR beams active to wake the FED when the pellet retrieval or poke happens, which is why they are always on and always sensing.

Eventually, if we redesigned the PCB from scratch we could do some clever things to save power without losing functionality, but I don't have any immediate plans to redesign it right now.

--
You received this message because you are subscribed to the Google Groups "FEDforum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fedforum+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fedforum/ecfc5883-4aa6-4ed5-8edd-9d3291cfdcf7n%40googlegroups.com.

[Nikolas Hayes]

Hmm maybe a dumb question... but my lab's only real immediate desire is to use the Free Feeding... what would be the result if I just didn't solder the neopixel and nosepoke IR sensors onto the PCB? At the moment I've been leaving the neopixels off of our Fed3's except for my initial one.

[Lex]

Leaving the Neopixels off should be fine (I guess you're already doing this?).  We may need to give you a small patch for the library if you don't populate the poke sensors, to not check the non-existent sensors. But it should be fine to leave them off.

I have an old board in lab, if I find time tomorrow I can try pulling these parts off and do some power measurements to give you an idea of the impact of doing that on power consumption!