Welcome to PhantomPilots.com

Sign up for a weekly email of the latest drone news & information

LiPo Battery Charge Table

Discussion in 'Other DJI Multi-Rotors' started by ladykate, May 25, 2014.

  1. ladykate

    Joined:
    Jan 7, 2014
    Messages:
    994
    Likes Received:
    7
    Location:
    Bellevue, NE
    This topic is meant to discuss LiPo charge tables and maybe best practices - not which brand should or should not be used. It is also meant to be for discussion, not necessarily for reference. I would like to hear your thoughts. I am not a battery expert but I have learned a bit in the last few months.

    To gird your loins

    Here is some general information about LiPo care and feeding for us RC types: http://www.rcgroups.com/forums/showthread.php?t=209187 (don't overcharge, discard batteries that have swollen, max charge rates, recommended charge rates, LiPos are dangerous, etc. This is a post that had dozens of references but mostly other posts on forums. Still - his guidelines rang true to me)

    Here is a wiki about LiPo and what the complete discharge state is (3 volts per cell): http://en.wikipedia.org/wiki/Lithium_polymer_battery (it also has a lot of general information that is good background)

    In a Google quest to find a charge table for LiPos, I found a bit of conflicting information (as always). I found charge tables that represented 'Fuel Table' that did not seem to follow the default recommendations of DJI in their assistant software. I also found information on LiPos that explain what represents a completely discharged battery (and therefore a damaged battery in most cases) but that wasn't even close to the values we normally work with.

    In an endurance test for my 810, it fell out of the sky (no graceful/slow landing) when the battery state got to 3.15 volts per cell. In researching this, I found that the particular ESC had a low voltage cutoff (LVC) of 3.15 volts per cell so... duh...

    ****... that means whatever the tables and information out there are - I must work with what is MY system's restrictions. A nominal '20% battery level' figure may be something quite different on my machines - and it may not be a good difference. I could be ready to land upside down if I push the envelope.

    So... and to the point of this post... I made up a charge table for MY machines. Each machine will have a different table if the ESC LVC is different but I now have a figure that at least makes sense to me... because I know the source data and logic that went into it. I know that 50% charged does not mean I've used half the battery (hope you can see why). And I also know what I'm doing when I change the values in the administrative assistant.

    Here is a table I made using the nominal 3.15 volts per cell. Again, I used 3.15 volts per cell because that is drop dead (literally) for my ESCs on one of my machines.
     

    Attached Files:

  2. doug86

    Joined:
    Apr 26, 2014
    Messages:
    290
    Likes Received:
    1
    Location:
    Bi-Polar
    Is that what you meant?
     
  3. ladykate

    Joined:
    Jan 7, 2014
    Messages:
    994
    Likes Received:
    7
    Location:
    Bellevue, NE
    Wow...

    Yes. The subject can be tedious but useful so it is worth the battle. Read the references and the battle is easier. Is that your only concern with my post? :lol:
     
  4. ElGuano

    Joined:
    Sep 4, 2013
    Messages:
    2,219
    Likes Received:
    4
    Nice chart! One to note: 4.2v is unloaded, and 3.15v is loaded, so if you're calculating percentage based on that, the numbers will be a bit different if you correct for both values being loaded or unloaded. Also, the drain off a lipo is largely non-linear (and depends greatly on load and type of battery), so your real-world ending point probably won't match with the spreadsheet numbers too well. For example, 20% drain on a 3S is nowhere near 10.08v, once you're there you're already falling off the voltage drop cliff. It's closer to 10.9v loaded/11.1v resting. It'll match up a bit better as you go 4-6S because there's less line loss as you increase in series.
     
  5. ladykate

    Joined:
    Jan 7, 2014
    Messages:
    994
    Likes Received:
    7
    Location:
    Bellevue, NE
    I agree with your points but this is a charge table (I know... I blurred the lines a bit). I think that is a hazard when trying to kill two stones with one bird (load vs no-load values). Loading will drop the 4.2 value but the 3.15 value will be the same no matter what. However, I'm not sure how to correct for it (or even if I need to). 3.15 is fixed (loaded or unloaded) since it is an ESC issue. How would you adjust for loading? It seems it would depend on the situation (though I'm only guessing at that). A smaller rated battery would drop more (more internal resistance) but upon charging, it would show up with more juice than was usable. The chart would show the amount of charge left (above 3.15) but wouldn't tell you how much time you have left on the particular application. I don't know how to make a general chart for that.

    I do know that this chart more closely represents what I'm getting when I test the batteries. Wouldn't mind some ideas on tweaks, though.
     
  6. ElGuano

    Joined:
    Sep 4, 2013
    Messages:
    2,219
    Likes Received:
    4
    You're right, your ESCs will never see an unloaded 3.15v/cell, but you'll most likely never really measure that either (unless you intuit from a low voltage alarm or read straight off an OSD).

    It gets hard for me to suggest many changes because many of my batteries behave differently from each other. A couple of my packs will drop down to 3.3v/cell loaded fairly slowly because they sag like crazy and that's not really the end of their life. I would never be able to accurately hit 3.3v on other packs because after 3.45v they drop like rock.

    Foe charging purposes I would personally go with unloaded because the goal is hitting 4.20v/cell and you take out the voltage sag variable of loaded voltage, but it's still rather battery-dependent and the important thing is that it's clearly noted I think.
     
  7. ladykate

    Joined:
    Jan 7, 2014
    Messages:
    994
    Likes Received:
    7
    Location:
    Bellevue, NE
    Maybe simply say: "Note: battery levels are unloaded values"

    Once you start to thoroughly explain the logic, you start drinking from the spittoon and can't stop.
     
  8. ElGuano

    Joined:
    Sep 4, 2013
    Messages:
    2,219
    Likes Received:
    4
    Yes, when every answer you give starts with "Well, that really depends..." then you know it's time to stop and take a step back :D
     
  9. dcoski

    Joined:
    Jul 21, 2013
    Messages:
    144
    Likes Received:
    8
    Aw! This is going to be a good post and just what the Doctor ordered. I get and have a lot of information every time I handle one of my batteries. Just my charger alone is giving me ton of information not to mention my battery tester, and the Phantom itself, but I have admit this is just blackmagic to me, and I remain clueless to what they are really telling me.

    So thank you for starting this thread. I am sure I will, and already have, learn a lot of valuable information.
     
  10. OI Photography

    Joined:
    Sep 23, 2013
    Messages:
    5,542
    Likes Received:
    17
    Location:
    Clarksville, TN
    If you have some time, this is a great primer on everything you need to know about lipo's: http://www.rchelicopterfun.com/rc-lipo-batteries.html
     
  11. dcoski

    Joined:
    Jul 21, 2013
    Messages:
    144
    Likes Received:
    8
    I agree it is a very good read. Hence after reading it last year I went and bought a ICharger and power supply last summer. I guess what I am trying to say is batteries are the life blood of our hobby. They are what make our Phantom/muli go. I have some batteries that approaching a year old and are working just as good as when they were brand new. I hope threads and topics like this continues. Who know maybe they (Batteries/Chargers) could be a general topic in this forum. We all use them.
     
  12. ladykate

    Joined:
    Jan 7, 2014
    Messages:
    994
    Likes Received:
    7
    Location:
    Bellevue, NE
    I'm dwelling on this a bit. I'm not satisfied with my first table. As was pointed out, 3.15 under load is the dead end for the ESC. If measured under load, the actual voltage would be 3.15 but under no load it would be higher.

    If I artificially raised the bottom voltage to, say, 3.6 volts (the intent of which is to represent zero effective power left), then the percentages come out more realistic and in line with what Ol Photography linked. That is, a battery with 3.72 volts per cell is 80% discharged. That also makes the bottom figure come from the same logic as the top figure (both no load).
     

    Attached Files:

  13. ElGuano

    Joined:
    Sep 4, 2013
    Messages:
    2,219
    Likes Received:
    4
    There we go :)

    10.8v unloaded is really close to the 3.3v/cell cliff that you want to avoid. And 3.72 unloaded is precisely where you'd find 80% discharge on a 3S (between 11.1 - 11.2v).

    And keeping the chart to unloaded values saves you from having to explain weirdness such as why 3.95v/cell is more than 40% discharged but in reality is what a freshly charged battery will display the instant load is applied.