Group Actions
Engagement
  • Author Author: alanmcdonley
  • Date Created: Date Created
  • Views 677 views
  • Likes 3 likes
  • Comments 3 comments
Related
Recommended

Rug Warrior Pi: Life in 10 Hour Segments

alanmcdonley

I have tested my bot's battery life a few times in the last days to find the safe operating limit to avoid memory card corruption.

 

RWPi Battery Life Tests

 

Configuration ***:

EBL 5000mAh Ni-MH C Batteries: $25 15Dec2014

Raspberry Pi B+ 512MB

Mikronauts.com Pi Droid Alpha interface card (MCP3008 Analog Digital Conversion)

Pololu S18V20F5 5v 2A Step-Up/Step-Down voltage regulator

Pololu Pushbutton power switch

Fair Top 5A Current Sensor Module ACS712

 

*************

Test 17Jun2016

 

(Batteries are 18 months old)

 

Analysis ***:

 

  •   Safety Shutdown Voltage: (10 readings) of 6.7v leaves 10 minutes remaining charge.
  •   Voltage knee starts at 7v, 10.25 hours run after removal of trickle charge
  •   Starting at removal from the charger, robot will sit quietly

         10 hours 45 minutes (6.7v) and then should shutdown for safety.

  •   0.9v per cell reached at 10.9h,

          less than one minute before voltage plummeted precipitously

 

 

Results ***:

 

  • The sometimes recommended 0.9v per cell discharge limit ( 5.4v ) occurred at less than one minute of juice left before shutdown.
  • Voltage Knee began at 7v, 10.25 hours
  • Precipitous voltage drop at 1v per cell, 6v was reached at 10.9 hours.
  • Recharge, took about 4 hours to replenish the roughly 3000 mAh run

 

 

Method ***:

 

  • Battery was charged fully overnight before the test.
  • Test was started promptly after removal from trickle charge.
  • center.py was executed to initialize tilt/pan servos holding at centers.
  • Test began with the command: sudo python battery_life.py
  • battery_life.py measured the 7v2 unregulated power through a 2:1 15k Ohm resister divider to ADC6, roughly once per second, printing to a WiFi connected remote ssh shell, until 10 readings of 5.3v were seen, then the system executed sudo shutdown -h now.

 

image

  • Interesting.

    I think perhaps it would be "better" to use LiPo batteries.

    Any chance of a Circuit diagram etc?

     

    I want to power a Rpi 3 with a 20,000 Ma Power pack or LiPo and really need to

    1) have a shutdown switch

    2) Automatic battery low detection and auto shutdown - preferably over an USB port sice I have a sensor hat on board.

     

    Due to a handicap I cannot build kits or solder myself so I am forced to use build and tested products - OR get them made for me.

  • in reply to clem57

    I found that interesting also, but then in retrospect I think I understand why:

     

    I don't know at what voltage the step-up/step-down regulator changes from step-down mode to step-up mode, but after that point, the current draw starts going up just to supply (load) power at the regulated voltage:

     

    If the transition point is 7v, the 260ma draw was using 1.8 watts (7v * .26A) to supply roughly 1.5 watts (290-305 mA at 5v) to the Pi, servos, interface board, and sensors.

     

    1.5/1.8 = 83% efficiency at the transition point.

     

    Below the 7v knee (transition point) the step-up process kicks in, and the power draw starts to go up.

     

    At 6v the current draw went up to 310 mA or 1.86 watts, giving 80% efficiency.

     

    At 3.6v it was sucking 600mA, which means the efficiency went down to 74 percent and the current draw was almost 2.5 times prior nine hours of draw down.

     

    Just when the batteries are tired, everybody wants more!

  • I have down a lot of graphs like this and that is one of the sharpest knees I have seen in real live analysis. I came close after many tries in as many months of work to find one of these.

    Clem