Welcome to the MIDI Sprout forum

Author Topic: Battery Life  (Read 714 times)

Offline sam

  • Sam
  • Moderator
  • Full Member
  • *****
  • Posts: 147
  • Understand How Your Tools Work!
    • View Profile
    • Electricity for Progress
Battery Life
« on: March 21, 2015, 04:46:40 PM »
MIDIsprout Battery Life Testing

When experiencing low battery power, the MIDIsprout goes into a 'Low Power Mode' in order to ensure uninterrupted performance.  Low Power mode turns off the LED Light show, saving power normally used as a part of the display.  At low power levels, it is increasingly possible for microcontrollers to produce MIDI data errors, by conserving power we can ensure that the MIDIsprout functions for many additional hours.

When using the sprout, if MIDI data is seen streaming out of the device (to a computer or synth)  but the LED light show does not illuminate, this indicates that the batteries are running low and that the sprout has kicked into Low Power mode.  Utilizing 3 AA batteries, full power provides 4.5V to the sprout. This voltage varies under load as the sprout detects microcurrent fluctations, outputs MIDI, and performs the light show.  The MIDIsprout can self detect its internal voltage (very cool SecretVoltmeter), when the system detects a sag below 2.7V under load the Low Power mode is engaged shutting off the LED Lights.

Normal MIDIsprout Usage: Attaching a MIDI sprout to a plant's leaves (Snake Plant, Philodentron, etc) using electrodes and connecting the sprout to a computer or synthesizer is considered 'Normal Usage'.  While MIDIsprout probes can be attached to people, plants, apples, or anything that conducts, the sprout is tuned specifically for sonifiying biodata from plants.  Battery life will vary depending on your application and usage. 

All values listed in this exploration of Battery Life refer to 'Normal' usage consisting of a Snake Plant and USB MIDI interface to a laptop.  All times are listed in Hours.

(0.00) MIDIsprout connected to snake plant using a pair of gel Tens electrodes and macbook pro with USB-MIDI interface.  I am monitoring MIDI data using a custom build of MIDIsprout_Tools, keeping an eye on the activity of note data and watching out for any MIDI data errors.  The USB-MIDI interface utilized doesn't appear to be ground isolated, which may present some noise and loading issues.  I've also seen artifacts in the past present in the MIDI data when 'touching' the body of a macbook due to non-optoisolated MIDI interfaces (particularly during periods of low plant activity).

(64.4)  The interface has been rock solid over the first few days of spring, with good temperatures and sunshine.  No MIDI data errors have been identified.  The plant has shown normal high/low activity shifts from day thru evening.  Touching the plant elicits the normal 'flourish' response.

(86.5) The system has now been operational for over 3 days of uninterrupted activity!

(90.3) Power Saving Mode ... engaged
In the later evening of day 4, power saving mode finally kicks in.  While the plant had been showing 'low' activity levels (appearing as few infrequent notes - a flat line), upon activation of Low Power Mode I am also noting an odd reaction when touching the laptop (USB MIDI grounding).  Unsure if this is at all related to Low Power mode, or if the plant is simply sensitive to the grounding change in the later evening hours.

As this was the first detection of low power, I decided to cycle the power (turn it off and back on).  The LED light show turns back on and activity is seen, still relatively low activity (normal for evening)

(91.56) Power Saving Mode ... Again!
After about two hours from the first activiation of Power Saving Mode, the sprout has again detected a critical voltage level and shut off the LED light show.  At this point, its rather late at night, and i noticed the data was producing a very small number of notes tightly clustered. When touching the laptop body, flourshes are seen in the note data.  I got up to check/adjust the electrodes, and saw normal activity before I touched the plant.
    - I am thinking that since the voltage has steadily decreased over days with the leds on, the voltage was sagging way low on average.  Then when the LEDs are turned off the average available voltage increases.  This would result in a net increase in voltage to the electrodes.  The plant may react to this sudden change and result in low activity. It has become obvious, from touching the laptop and seeing a reaction in the data that the USB MIDI interface is not completely isolated.

(92.21) power saving mode
    flatline same as above,rest again - leds illuminate, normal activity
(95.96) still in low power, normal activity ...
   -obviously a change when in low power mode that the plant reacts to, gets used to...

(96.56) a return to moderation, touching laptop body results in diverse signal.  note i had been using the keyboard for a while.  perhaps my touching creates a disturbance in the current.  again MIDI in ground must be coupled to usb ground ... should test!

(109.68) after all night alone in low power mode, normal activity seen in morning.  no errors.  amazing!

... not much activity, its the first day of spring, and its snowing ...

(161.0) the plant has been flatlined for about two days.  MIDI still flows and there have been no errors.  when touching the plant, normal excitation is seen, suggesting that the MIDIsprout is still working correctly.  I have noticed a steady increase in ground sensitivity, touching the macbook body does not consistently make an effect.  unplugging the laptop increases sensitivity, but the plant still shows flat.  I have decided to remove the electrodes and take the plant outside for some spring sun. 

this experiment is concluded.

    after leaving the plant outside for the first day of spring, i have again attached the trodes to the plant and usb interface, using the same low level batteries.  some activity with the laptop unplugged.  flatline slow notes when charging.  touching plugged in laptop has minimal effect.
     now i tried new batteries to see how the plant reacts with the difference in charge: lots of activity and lightshow.  Perhaps the charge, the galvanometer, and the threshold have a relationship.  i should test low power mode (no leds) with full power batteries to see if the change in activity is related to charge, or perhaps pumping from the LEDs introduces noise (tickle).

Battery Life - Low Power Mode and isolation considerations
« Last Edit: March 24, 2015, 12:06:57 PM by sam »