P3 turned off and fell from the sky

[ParsnipHysorter]

After reading a lot of threads on the new firmware upgrade and prior "sky fall" incidents, I think that cold makes the voltage drop of the battery worse when it is under heavy load such as when revving up the engines by throttling upward aggressively. When batteries are older or lower on charge the voltage also drops severely when under heavy loads. So what DJI seemed to do is monitor battery temperature and calculated voltage to change throttle load limits (decrease them) in times of low temperature and lower battery life (60% and lower perhaps). This would decrease the voltage drop and hopefully keep it above 12 volts (3 per each Lipo cell X 4 = 12v). I also hope they erred on the side of safety and eliminated the battery shutoff. Who cares about damage to a battery over time; this better than a crash that bricks a phantom and its battery!

Probably, but how they are factoring in battery temperature confuses me. Even in so-called cold-weather conditions the battery internal temperature starts at the starting (room) temperature and does nothing but increase, and increase fairly quickly, over the course of the flight. Below is a graph of my battery temperature from my tests today in 30-something weather -- the decrease in performance started around 5 minutes into the flight, and temperature of the battery was over 80 degrees F by then. By the end of the flight when performance was at its worse the battery temperature was over 100 degrees F. Even when I've flown in sub-zero (F) weather I get a similar temperature curve.

Oh and let me add that even at end of flight, about 30% remaining during this flight, battery voltage was around 3.5 to 3.6 V -- a little ridiculous to be throttling so much when the cells are well above 3V still.

 

[ParsnipHysorter]

So when my battery when charged to 100% gets to 50% than I need to pack it up and quit flying?!?! I can't believe all of you find this is normal!!!!!

That's a bit silly; I've never done that, unless there is a reason like I'm just done for some reason. Always landed around 30% or so; sometimes a bit less, sometimes a bit more.

 

[jaked_902]

That's known as "critical low battery". At that stage, the Phantom is supposed to auto land at its current location. It seems whatever these pilots are doing is causing that to be skipped somehow.

Or, perhaps the software is failing all on it's own, likely due to faulty battery chemistry, and the pilots aren't doing anything other than flying under the circumstances DJI recommends.

 

[Carmelsteve]

I agree completely. I've flown and owned several other quads and never had to be so careful with voltage for fear of a motor failure. If you wanted to fly down to 3 volts you may damage or decrease lifespan of the battery, but it wouldn't just shut down. I commonly flew to 10-15% without ever having a problem. I still say it would be nice if DJI stated whether this was a software or a hardware issue.

 

[msinger]

Or, perhaps the software is failing all on it's own, likely due to faulty battery chemistry, and the pilots aren't doing anything other than flying under the circumstances DJI recommends.

It seems most pilots are taking off with partially charged batteries -- which DJI does not recommend. So, they aren't exactly flying under DJI's recommended circumstances.

But, yes, there could be an issue with the software that is causing the critical low battery detection to be skipped. And, by the pilots doing something, I didn't mean it was necessarily their fault. They might just be hitting an edge case that DJI did not test.

 

[jaked_902]

The .dat file starts as soon as power-up, so from that you could plot the voltage until takeoff.

Jake,

When you play back the flight in the DJI go app what battery levels does it show? That log also starts before takeoff.

Voltage is not an accurate way to determine SOC....and it is also all that is available when you start. It can be accurate at full charge. I always check that battery voltage is at >17V before flight....anything less and I would definitely not fly....

I saw 98% on the app when I connected. The logs show the battery at 58% for the entire flight, which obviously isn't correct. It's plausible that having just disconnected this battery from the charger I saw 98% when it was instead 58%. However via my experience as a professional IT tech, I would also argue it's plausible that the battery or the battery chip malfunctioned and that app did indeed display 98%.

I'm also curious if the battery or the Phantom has a temperature sensor. If so, it seems logical that the battery capacity would be calculated differently after being brought outside where the temperature was lower.

I did wait with props turning / hovered for 30 seconds before moving.

I actually re-read the user manual, but I can't find anywhere that DJI recommends not using full throttle, or warns that the battery could turn off mid-flight for any reason.

 

[Dirty Bird]

Obviously most of my experience is with P2 batteries. I have ran my batteries down to 10% (or less) on many flights. Probably most of my flights in fact. When the batteries were new over the first summer (got the P2+ in April '14) they performed flawlessly & I never had an issue. As temperatures began to drop in the late fall I began noticing under max power, some cells would signal a low battery alert. That's how my bird wound up spending a night in a tree! Back off the throttle & the alert would normally stop. By late Fall all 4 packs had close to 100 cycles each so it was probably a combination of aging packs & decreasing temperatures causing the issue.

Throughout it all, however, I don't recall hearing stories about P2 packs just shutting down mid-flight other than one time I deliberately hovered mine about just above the ground all the way to 0% & then some until the pack literally shut off (she dropped & stayed on her skids!) I have had a pack get to the auto-land point & the aircraft stops & initiates auto-land. It seems these P3 packs would be better to lower the shutdown threshold & risk damage to the battery rather than simply powering down & crashing both aircraft & battery.

 

[Richard W]

Reading above about the experience with P2 batteries, and reading many recent posts about birds falling down due to shutdown related to P3 battery behavior does not add up.

First, an idea of protecting a battery by shutting down the motors in flight that in all likelihood will crash the bird, and the battery in it, is insane. Not sure if that is truly the DJI logic intentionally implemented in the P3 ware but that is what was implied many times in many posts on these forums. Hope it is not, as that would be, as already demonstrated, a counterproductive approach, and class action lawyers would have a field day with it.

The more likely explanation is either hardware (battery) sudden failure under certain conditions, or software logic error somewhere. I wish we knew which one ... it would be much easier to work around the problem.

 

[Oso]

It seems these P3 packs would be better to lower the shutdown threshold & risk damage to the battery rather than simply powering down & crashing both aircraft & battery.

I'm pretty sure that everyone agrees with you. AFAIK, all that we have been told is the software has been miscalculating remaining capacity. This was a defect that was recently understood but it did already cause lots of crashes on both P3 and Inspire 1. Hopefully 1.6 fixes the bug in battery management at least for the time being.

 

[kirk2579]

So when my battery when charged to 100% gets to 50% than I need to pack it up and quit flying?!?! I can't believe all of you find this is normal!!!!!


Sent from my iPhone using PhantomPilots mobile app

not what I said, But once you get to 50 or less its time to start towards home, keep flying as you wish .

I was just responding to someone that did not believe once they get low that its possible to over tax them. The other point is fly easy vs hard and fast when the battery gets lower. The battery will last longer and not over tax the available current.
as far as normal --that is why the new FW , to address some of the experiences.

I don't run my car gas to empty when out on the highway hoping....
I stop and top off at about a quarter tank. that's my RESERVE I think that's all anyone is saying!
Be aware batteries are a bit finicky and I would rather land with 30% reserve battery time and time again vs run it to 10% and maybe run into an issue.

Ill give up some fly time today, to make sure I have something to fly tomorrow.

good luck and have fun flying!

 

[1st_time_flyer]

It seems most pilots are taking off with partially charged batteries -- which DJI does not recommend. So, they aren't exactly flying under DJI's recommended circumstances.

But, yes, there could be an issue with the software that is causing the critical low battery detection to be skipped. And, by the pilots doing something, I didn't mean it was necessarily their fault. They might just be hitting an edge case that DJI did not test.

How do you explain the chart that is above when it comes to the temperature of the battery increasing than?

 

[msinger]

How do you explain the chart that is above when it comes to the temperature of the battery increasing than?

How is the battery temperature related to my comment? What exactly should I be explaining?

 

[JamesBandes]

Jaked,

When was the 58% battery charged last? To me it seems, from all the similar posts that once the battery enters self discharge, the firmware doesn't read or report the correct voltage. Many machines shutting off in FFF (fast forward flight). You wouldn't have had 58% really left in the battery. DJI just released new FW to address this 2 days ago.

Another cause could be high internal resistance in the battery cells to have a resting charge of 3.83v per cell and pull down to 2.98. What is the battery life reading and how many cycles on the failed battery? I still think it's the firmware after a self discharge. Again, when was the last charge on the battery that took a nap?

Sorry about the crash.

SD

New firmware 1.6 is supposed to fix this but I am also resetting my battery discharge back to the default of 10 days

 

[JustinJ]

New firmware 1.6 is supposed to fix this but I am also resetting my battery discharge back to the default of 10 days

I don't think the discharge timescale is the problem here...

It's more so the fact that you need to ensure the batteries have received sufficient charge prior to use.

 

[JustinJ]

How do you explain the chart that is above when it comes to the temperature of the battery increasing than?

What's your point?

What do you need explaining?..

That the graph illustrates the data that the battery gets warmer the longer it's in use...

 

[1st_time_flyer]

How is the battery temperature related to my comment? What exactly should I be explaining?

you have been insisting all along that the battery is losing voltage during cold climates and that the battery should be charged to 100% before ever flying....if the battery is in fact getting warmer than why should we worry about voltage drops in cold climates with less that 100% battery life?

 

[Mario_SB]

jaked_902, your logs show the following:

- You took off with a battery that was only 58% charged.
- At takeoff, the battery voltage was 15.35V (about 3.83V per cell).
- When the battery shut off, the voltage was 11.95V (about 2.98V per cell).

The battery will auto shut off when the voltage drops below 3V per cell. That is by design to prevent the battery from damaging itself. See more details here.

Since you climbed at such a rapid pace, it caused the voltage to drop to a very low level since your Phantom was drawing a lot of power to achieve that feat. You would have been fine if you had taken off with a fully charged battery.

Busted! Another one bites the dust.... 100% battery charge and nothing less folks.

 

[Mario_SB]

Forget about emailing DJI, either call them or use their chat feature. Calling is your best bet.

I agree. They actually have a very decent customer service team that will handle your questions over the phone. Out of the three times I've called (P3 cracks / follow up) I've only had to wait no more than 2-3 minutes.

 

[dirtybum]

My p3 adv. had total power failure 2 sec. After takeoff on firmware 1.5.3 and i just received this back from dji, which is bull****! So any talks about dont upgrade and your still under warranty, better look at this!!!!

 

[dirtybum]

I updated the day 1.5 came out and this power failure happened on the 9th dec.