Remote Sensor with Solar Powered Li-Po Charging

Discussion in 'The Projects Forum' started by Marquis, Jun 28, 2015.

  1. Marquis

    Thread Starter New Member

    Jan 18, 2015
    16
    0
    Hi,

    I'm working on some remote sensors that need to need to be solar powered. I've got some questions that are probably pretty basic.. And I'm looking for feedback.

    I want to use a small solar cell to charge a 3.7v lipo battery. Since my circuit is also drawing from the battery, what type cycling will happen? How often will the cell be topped up? What should the duty cycle be? When the sun is shining does the panel power the circuit directly? I plan to power a 3.3 volt atmel directly from the 3.7v lipo. I'm planning to use this charger. https://learn.adafruit.com/usb-dc-and-solar-lipoly-charger

    Now to the trickier part, I want to log when the charger is operating, and when it isn't. Thoughts?

    To the simple questions. I am trying to measure the power from the solar panel. What happens to power from a solar panel when the charger isn't operating? Can I measure that without the charger operating? I take it that a panel doesn't create power when the circuit doesn't draw current? So do I need to find a way to draw power to say a fan, just to be able to measure the power? Suggestions?

    Some people have mentioned a diode is needed to prevent current from leaking back to a solar panel? What does the panel do with the current?
    Thanks for answering any of the above questions..
     
  2. Lectraplayer

    Member

    Jan 2, 2015
    77
    5
    I'm not knowledgeable on a solar panel, but I would size it at least three times what the battery size is, maybe four, and battery for about 48 hours total load. I also would use a charging regulator, or you could overcharge and wreck the battery. To monitor the charger, I would use a comparator to tell if solar voltage is higher than battery voltage. It also sounds convenient to run all loads off that battery.
     
  3. wayneh

    Expert

    Sep 9, 2010
    12,153
    3,059
    If the panel is powerful enough, it will indeed power your load and charge the battery at the same time. As the panel capacity drops, there will be a point where the battery has the higher EMF, and thus charging will stop and the load will be supplied from the battery alone.

    You'll need something capable of logging data. Lot of options and I'm sure the recommendation will depend on a lot of factors you haven't mentioned, such as familiarity with one microprocessor or another.

    Power depends on the load. No load, no power. Too much load (such as a short), the power is very hard to measure because there is a tiny voltage drop across the load.
    You cannot measure power with no load, but you could measure open-circuit voltage which will, for a given PV panel, indicate the availability of power.

    Correct; no load - no power. But again for a given PV panel, the open-circuit voltage will be a measure of illumination and thus the power available. A short-circuit current might be more useful, if you have a way to measure that. Best of all is to load the panel to drop the voltage to ~80% of the open-circuit voltage. This will be close to the peak power point for the panel.

    A blocking diode is the most common method to prevent discharge of the battery into a darkened panel. If the current is blocked and there is no load, no current flows in either direction in the panel. Think of it as a battery not connected to anything.
     
    uwed likes this.
  4. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,176
    397
    It appears that the selected charge controller has builtin reverse current protection. Reverse current con be measured when solar panel, controller, & battery are all connected.
     
  5. Marquis

    Thread Starter New Member

    Jan 18, 2015
    16
    0
    Thanks for the great replies. I should clarify a few things. I am familiar with micro processors and the logging data will be no problem for me. My issues mostly involve understanding solar panels and charging of lipo batteries.

    It sounds like ideally I want to measure two things, the open-circuit voltage of the panel and a panel loaded to 80% of the load. If I wanted to measure this intermittently, I basically need a dummy load (large resistor) that I check the voltage drop across to see what kind of current it is making.. Sound about right?

    As for sizing the panel, my logger will use a pretty small draw. I'm just working at prototyping that so I can measure my current consumption and see what size panel I need.
     
  6. wayneh

    Expert

    Sep 9, 2010
    12,153
    3,059
    Yup, that's how I would check the health of a PV panel – slap a load across it and see what it can do. Note that performance is inversely affected by temperature. The effect is not huge, but easily detected.

    I question whether you need to do this on an ongoing basis. Once you know the I-V curve for your panel (which you can guesstimate from the specs or from the OC voltage and short circuit current, or you can map it out precisely), it should be stable for a long time. It will degrade only slowly over time, maybe a few percentage points per year.
     
  7. Lectraplayer

    Member

    Jan 2, 2015
    77
    5
    I know on an amorphus silicon panel I just got, they say it may degrade a bit in the first six months of use. ...though I expect TiO2 based cells or cells overlaid with Lexan or anything but glass to go downhill more quickly and for longer. You may also want to monitor the panel to determine when it needs to be cleaned.
     
Loading...