20 Seconds in the air, my 6 kg oktocopter crashed with two Gens Ace batteries aboard (4S / 6750 mAh), both batteries were considered in very good condition and fully charged. The voltage suddenly dropped from above 15V to below 8V, in less than 10 seconds. Discharge graph attached. I would have never imagined that such a massive power loss might happen so quickly, using first class batteries, less than a year old. The details: - 2 x TATTU batteries 4S / 6750 mAh, connected IN PARALLEL - both batteries less than a year old, 18 cycles, no problems whatsoever - both batteries considered fully charged (6 weeks ago) -- however, this cannot be verified - after crash (motors off), both batteries recover to more than 4.0V per cell - same scenario was repeated on the ground: same batteries not recharged / 20 seconds delivering ~55 Amps, then, within seconds, the within seconds, voltage drops below 8 Volt The questions: - what has happened? - why can a battery show > 4.10V per cell (without load), and, under normal load (4C), drop below 2.0V per cell? - what might have gone wrong before the flight? - what would have happened, if one battery was not charged (20% remaining cap.) while the other one was fully charged, both connected in parallel? Diagram attached shows A and V of battery #1 over time (seconds) - no sensor data for battery #2.
Reto: It looks the battery chemistry was compromised (not that you haven't figured that out yourself). At what voltage did you keep the batteries stored? You mentioned that they were fully charged six weeks ago? Andy.
As Andy asked, were these sitting fully-charged for six weeks? That is generally considered an abuse of a LiPo. Most of us will not leave a battery in anything but "storage" voltage for that long. It's widely believed a fully-charged LiPo can become compromised if left fully charged for a long time (and by long most of us believe that's more than a couple days). No conclusive measurements are available (that I'm aware of), but you might have sadly produced a test case for that belief.
Thanks a lot for your valuable input. I must confess that the two batteries were left charged at about 4.15V per cell for 5 weeks, unfortunately. But that happened to much cheaper (Zippy) batteries quite some times before. What I observed was a slow but steadily degradation in terms of voltage stability, but never something like that. The strange thing is, I´m using two batteries in parallel, and both batteries have failed exactly at the same time? That´s hard to believe. And even worth: They delivered energy for 20 seconds without any noticeable deficit and then both(?) power sources collapsed within the same 5 seconds? After that unfortunate event, I measured both batteries: They showed a voltage of 4.10 V per cell, no drift between the cells. I repeated the failure on the ground, this time just one battery at a time. Both batteries gave up after 20-25 seconds, exactly as during the fatal flight. Similar voltage/current diagrams as above, without recharging in between. Really strange. In the rather unlikely case, that I forgot to charge one battery, do you think such a scenario could have happened? Or what will happen if I connect a fully charged LiPo to an almost empty one, in parallel? The full battery starts to charge the empty one? One last thing about LiPos: I can hardly imagine that anybody truly adheres to all the dos and don´ts about LiPos. Or do you really monitor all your batteries during charging, all the time ...? But I start to realize that keeping the batteries charged for more than 2-3 days does not just shorten their live, but is also very dangerous.
A few comments: A LiPo will often exhibit far different voltage under various loads than at rest. So while a resting check of a LiPo is helpful if you don't remember if you charged it, it won't tell you much more than that. Proper charging, use, discharging, and storage is the difference between catastrophic situations like what you experienced and safe operation. I asked an industry veteran (who shall remain nameless, but you all know who he is) if he's concerned about running LiPos in parallel because of the potential for failure. He said in thousands of hours of flying in ten plus years he's never had a battery fail in the air. Ever. He's had plenty that exhibited charging anomalies, and many that were damaged in crashes or other mishaps. And in every one of those cases, the batteries in question were retired to either bench duty, or the scrap heap (after carefully being rendered inert). So while I freely admit to have left batteries charged for a few days at a time (longer than the generally-held orthodoxy of 'fly them or store them within 24 hours'), I take LiPo safety very, very seriously. I appreciate your candor and willingness to share your story. I'm really sorry about your crash, and I hope you can recover from it. And I also hope that you (and everybody who reads this thread) will appreciate the complexity of our friend (sic) the Lithium Polymer battery.
One more thing I forgot to mention so far: After the fatal flight, which "successfully" pulled 25-30A out of the defective packs for at least 20 seconds, I repeated this scenario on the ground. I again pulled 27A out of each battery (one after the other) for about the same time, before each of the two batteries collapsed again. After that, all 4 cells of both batteries still showed more than 4.0V per cell, without load. I therefore discharged the two batteries with my charger at only 2A - 5A and I managed to get about 3000 mAh (half of the capacity) out of each battery. That means that 1. the batteries were far away from being empty when they collapsed under "high" load of about 5C. 2. the batteries were still able to deliver low current of less than 5A for many minutes without dropping below 3.50V.
I had a very similar problem: http://forum.freeflysystems.com/index.php?threads/voltage-16-8-to-14-in-30-sec-flat.2505/ My health is very up and down, and in the past I've not worried too much about leaving my batteries fully charged, while my health improves. I have found, over the last 4 years that this is a big problem. I've had 4 month old batteries fail after 5 flights, and the only reason I can figure out is because they sat fully charged. Fortunately most of them were cheap Zippy 4S 4500mah for my smaller Quads. (Phantom, etc) The Tenergy and Zippy's were the worst. The Gensace and Tattu hold up much better, but with these larger more expensive batteries I've tried to be a lot more careful. I now only charge them if I am sure I'm going to fly. And if I don't fly, no matter how crappy I feel, I drag myself to my charger and discharge them. Heck I even had two Tenergy's puff before I used them, charged for about 6 months, $50 down the drain.
Yup. Fully charged for "too long" (and nobody seems to know that means exactly) or over-discharged (below 3.0v/cell) is a real Lipo Killer. The Hyperion SuperDuo 720 (which I have) and, I suspect other similar chargers (which I don't have), have a "take to storage voltage" which either charges up to, or discharges down to, storage voltage. And that's the only way to keep Lipos. I take the "too long" threshold for keeping a LiPo fully charged as 48 hours. Why? Absolutely no basis whatsoever. It's the Goldilocks Effect. Less than that seems too short, longer than that seems too long, and 48 hours seems just right. FWIW the new so-called "Intelligent" batteries from DJI (and I have a deep, deep distrust of anything with microcomputers and firmware that uses "Intelligent" in the name) have a built-in discharge to storage feature that kicks in with a default value of 240 hours, but can be set down to 24 hours if you want. One other alleged fact (I read it on the Internet, so it must be true....) is that some folks take their Lipos to storage voltage and then put them in the fridge. I've not done this because it would mean I have to take the beer out of the garage fridge and there is something, oh, I don't know, fundamentally wrong with that idea... Finally, as Steve says, the "no load" voltage of a LiPo is usually quite different from the "loaded" voltage. Remember voltage is the potential difference and it says nothing about the ability of the LiPo to deliver power (volts x amps) -- that will depend on such factors as the health of the electrolyte, the internal connections, the internal resistance of the cells, and, of course, the phase of the moon and whether I'm the guy trying to use the LiPo for that last flight of the day when the sunlight is magic. Andy
Thanks a lot folks for your valuable input. What I learned so far is: 1. Keep batteries fully charged for more than 2 days is a bad idea, I knew that but I underestimated the importance. 2. Hovering for about 10-20 seconds not far up while monitoring the voltage seems to be a good idea, will do so. 3. High quality batteries are better; I did use hq batteries when I crashed - with my cheap batts I never had such a problem. Hmmm ... But what´s still a myst is the fact that two batteries, always used in parallel, always charged separately, worked perfectly for 18 cycles, and then, on the same day, fail within seconds. And by failing I don´t mean a voltage drop of 0.5 Volt per cell over some minutes. Failing means the voltage was suddenly cut in half, within seconds. That´s really odd and leaves a very bitter taste, after all. How can I ever be sure that the batteries I´m flying are really healthy (before I land) ?
I´m attaching both discharge curves of every single cell (of the two packs), under nominal "hover load" of about 30A (per pack). I can confirm that both batteries have been left fully charged for a bit more than 4 weeks (I know that was a fault). There are no other irregularities I´m aware of. I can exclude, however, that one battery was charged and the other was not. That´s for sure. Please note that the batteries were less than a year old and only used for 19 cycles.
Reto: Nice work. What's your experimental setup? I've been doing some discharge curves but just on the batteries total voltage not individual cells. I use four 0.5 ohm 1,000 Watt ceramic resistors as a dummy load and a precision Empro shunt that produces 100mV/Amp then feed the volts/amps data into a DataQ acquisition system and post process with custom software and use Gnuplot for graphing. Unfortunately it does look like that extended fully charged period has killed cells doesn't it? C3 and C4 in battery 1, and C2 and maybe C3 in battery 2. Andy.
I haven´t produced the curves myself, to be honest. I sent the batteries to a battery expert, a guy who builds very elaborate battery monitoring systems called IISI. This telemetry system can track single cell voltages among many other parameters and one can setup alarms for almost any value being monitored. This system warned me when the batteries went crazy but it was already too late to land. The voltage dropped so quickly, I would never have imagined that this can happen to a Lipo battery. And I had two of them, connected in parallel. For me it´s still a myst, although I know that I mistreated the batteries by keeping them charged for a couple of weeks.
Thanks. That was the info I needed. The actual URL is http://www.iisi-rc.com/IISI-RC/Home.html just in case anyone else is interested. Andy