There have been a rash of Phantoms falling out of the sky lately.  Many caused by the battery voltage becoming critically low, especially when using a partially charged battery.  This video explores the phenomenon.

Very informative video and thanks for making it. That's why users should update to the latest version to help with the cold battery operations.
Also, really cool test setup you have there.

Very nice.   What it also points out is that you should ignore the time line and focus on the battery voltage.  It is clear that DJI really has not thought out how to present the information.  You can have two absolutely opposed indicators - a critical low voltage error and a 'good' battery timeline.  This should never happen.

DJI needs to sit back and think through the Go app.  It's really a mismash of data presented in a fairly hostile and confusing manner.  

This really should be a sticky....

rogergolub Posted at 2016-1-15 18:39
Very nice.   What it also points out is that you should ignore the time line and focus on the batter ...

Good point would help many to understand the complexities of battery temps PERHAPS DJI WILL CONSIDER IT AS A STICKY

DJI-Ken Posted at 2016-1-15 10:57
Very informative video and thanks for making it. That's why users should update to the latest versio ...

I am using the firmware update for cold batteries.  I was a beta tester.  What I didn't appreciate at the time was what was meant by "cold".  I did a lot of flights at -10°C but always started with a fully charged 20°C bat with no problems.  I then did a take-off with a 70% charged bat and the phantom did a forced landing because of low voltage.  That precipitated these testing video.
I was surprised by how fast the batteries heat up and by how much better they perform when warm.
I found it takes 2 minutes of flight to warm the battery to a safe temp.  After this a low voltage alarm will not occur (partly charge bat at a min temp of 20°C.

hummingbird.uav Posted at 2016-1-16 03:49
I am using the firmware update for cold batteries.  I was a beta tester.  What I didn't appreciate ...

Sorry, when  said user should be on the latest update, I was talking to other members who would read my post. I know you were on the firmware.
But yes, cold batteries do not work very well at all.

This is true of even lead acid car batteries.  A little known trick if you are stuck with a dead battery in the cold is to heat the battery.  Get it warm enough and it will probably start your car.  An even less well known trick,  and slightly dangerous as well as hard on the battery, is to dead short the battery for a couple of seconds.  This will put lots of heat in for a short time.  Start the car right away,  because this charge will not last long.  Probably best to call AAA if you can instead of doing this,  but it does work.

This is TOTALLY opposite to the laws of physics.
If you want to make a "red neck" super conductor, cool off the wires in liquid nitrogen, in this way there is no resistance in the metal opposed to electrons flow. As the wire warms up the resistance increases and flow decreases.
In DJI case, a cold battery should have better voltage power than when warm. I remember, when I was flying RC planes, the electric guys were bringing the Li-Po in a cooler with those blue gel packs to keep the batts very cool before flight increasing the power and flight time.
What I see here, a HOT battery from DJI is better than a cold one??? Has DJI flipped the law of physics?
I would like to know who is making the Li-Pos in the pack. Is that a lowest bid contract?
Also the thicker the wires, the better the current flow, less heating, more amps to user.
Has anybody taken the drone body apart and look at the power wire gauges? I know the electric RC planes use thick wires (???gauge) throughout the circuit starting from the Li-Po battery to ESC to motors.

Dacon Productio Posted at 2016-1-16 12:52
This is TOTALLY opposite to the laws of physics.
If you want to make a "red neck" super conductor,  ...

Dacon what you fail to realize is that the battery is a chemical device.  The chemical reaction is enhanced with added temperature.  As the battery is used the reaction generates heat, quite a lot actually.

This post is excellent thank you for taking your time doing this test.

rogergolub Posted at 2016-1-15 19:39
Very nice.   What it also points out is that you should ignore the time line and focus on the batter ...

I totally agree!

DJI: your power management needs an accurate and full review as soon as possible!!
The charge level of the battery, to which is linked the estimated time of flight, in many cases is not reliable!!
This leads and will lead to many crashes!

What we know is that in order to use the estimated available energy of a battery, a minimum temperature is required!
More, we know that as the energy percentage decreases, this minimum temperature of the battery that lets this energy actually available increases.
So, as I suggested in many other posts, immediately after the warmup of the P3, if the above conditions are not verified, give a warning message like this:

   Battery status is not suitable for a safe fly! Fly time is not reliable!
   Please use a full charged battery!
   More, change the status of the timeline (i.e. all red or flashing) or place a warning message on it!

The user will be correcly informed! If he decide for the takeoff, he will do so under his responsability!

amadeus Posted at 2016-1-17 06:45
I totally agree!

DJI: your power management needs an accurate and full review as soon as possible ...

Yesterday I was flying in +7C weather - not particular cold.  The batteries were showing 3.6V after about 17 minutes.  That's not bad but the idiot time line showed I had another 15 minutes to go before 'critical'.   Not a chance.  I bring the Phantom in at around 3.5 - 3.6V.  I'll put another battery in rather than trying to eck out another few minutes.

But I just ignore the time line.  It's nowhere near accurate enough to base flight decisions on.

Nice tests. Thanks for sharing.

That makes me wonder... wouldn't it help in the cold weather to cover the vent openings in the shell?

I didn't have time to documented yet the results but got issues when used the bird after been in the back of the car for two hours in a 330C day. Video is all weird and buttery warning as well.

Also, after the two last updates, I'm often getting some warning "limiting power to protect the battery". Not sure what triggers but I'm suspecting windy conditions.

Cheers.

Paniza Posted at 2016-1-17 06:57
Nice tests. Thanks for sharing.

That makes me wonder... wouldn't it help in the cold weather to cov ...

When you get the 'limiting power' warning, are  you pushing on the sticks hard?  

WetDog Posted at 2016-1-17 11:11
When you get the 'limiting power' warning, are  you pushing on the sticks hard?

I probably am.

Paniza Posted at 2016-1-17 07:41
I probably am.

I think that's another common thread in the Phantom-falls-from-the-sky and why it doesn't happen to some people.  Watching people fly on YouTube, there seem to be two different types of pilots - light sticks and heavy sticks.  Some people just bump the sticks a little bit and others try to get there as fast as they can.

When you really load the Phantom, voltage is going to go down.  When you load a cold Phantom voltage is going down further.  Couple with not watching the voltage (not the timeline) carefully and perhaps a gust or two of wind and you can get the battery in trouble.  

And apparently, at least until 1.6, the battery would shut down and crash the Phantom so as to save the $150 battery.  It seems that in 1.6 DJI limited the amount of battery that the ESC could ask for and / or let the battery go into a low voltage state in order to keep flying.

I tend to be light on the stick - it's not a competition drone and taking video while moving fast tends to trigger vertigo in viewers.  But it might be wise to lighten up a bit, especially in the winter.

WetDog Posted at 2016-1-17 12:14
I think that's another common thread in the Phantom-falls-from-the-sky and why it doesn't happen t ...

That makes sense and it is a good thing.

When rain is coming in a tropical country, forget about light sticks.

Anyway, I think the issue might by windy conditions... a bit against the windy and it is a "false" heavy stick situation.

I have included a graph which compares the performance of fully charged batteries at different temperatures.  Each line is an average of four batteries.  Quite a difference in voltage sag!

Voltage Sag

Great post I personally experienced this with a cold ¾ charged battery. I lifted of and about a minute later I got a warning that said my thrust would be limited to protect battery. About 2 minutes later my phantom advance went to critical error and landed.  I still had half a battery left. Has anybody else received the limit thrust error? I called the help line and they had never heard of the error or the phantom doing this and told me to send it in for repair. Thanks for your help!!!

tinner21 Posted at 2016-1-26 15:53
Great post I personally experienced this with a cold ¾ charged battery. I lifted of and about a min ...

Yes, another poster has mentioned receiving this message. Sounds logical to me. It could prevent a complete shutdown in flight.

Incredibly useful. I had no idea that 20 celsius (68F) would be considered "cold"! The video really makes the point of distinguishing between low voltage and low "power" (really, low charge), with auto-lands shown for each situation. I may have read this in the documentation, but it didn't penetrate my thick skull.

Paniza Posted at 2016-1-17 08:57
Nice tests. Thanks for sharing.

That makes me wonder... wouldn't it help in the cold weather to cov ...

The limit thrust warning happened to me as well, with no wind. I did, however start with a battery that was not fully charged. The ambient temp was +1c. After receiving the error the  aircraft went to critical and landed.

tinner21 Posted at 2016-1-26 10:57
The limit thrust warning happened to me as well, with no wind. I did, however start with a battery ...

Interesting.

What if... you charge the battery at cold ambient temp and fly from there?

Would it kill the sag?

DON'T BRING IT INSIDE OF THE HOUSE! ... the warmer conditions could explode the battery.

Paniza Posted at 2016-1-26 11:11
What if... you charge the battery at cold ambient temp and fly from there?

Would it kill the sag?

Sorry... I will start another post with the question above.

Thanks.

DJI-Ken Posted at 2016-1-15 11:57
Very informative video and thanks for making it. That's why users should update to the latest versio ...

It's too bad that DJI will not honor their warranty if your battery failed over water ---

I'm not sure if you fellow DJI customers were aware of this or not, but apparently flying a DJI Phantom Pro3 over any body of water will now void your warranty if it crashes into it, even if the failure was due to a DJI failure of some sort and not user error.  After weeks of emails and phone calls to get DJI to replace mine today they sent me the following explanation for why they will not honor their warranty.

"Dear Gino Barasa

By choosing to fly the unit over open bodies of water, you run the risk of anything that may happen to the unit, whether it be user error, product malfunction, or some force of nature, it will not be repairable, and considered a total loss. If this were over open land, as we suggest, this would have been a simple repair, possibly, and more than likely covered by warranty.

By choosing to still fly the unit, after noticing possible issues, over a lake, I believe you are at least partially responsible."

So, even if the unit crashed due to "product malfunction" they will not honor their warranty if it lands in water, which is strange because you won't find that anywhere on their own website. But you will find a ton of images and videos of DJI drones flying over rivers, and lakes, and oceans and back yard pools.
The "possible issues" that they mention was a "critical battery failure that appeared on the flight previous to the crash. I landed the Phantom when the warning came on and switched out the battery.  I then flew the bird over my head a for a minute to see if the warning would come back on with a new battery, which it didn't. I then sent it out of the lake and it immediately shut off and dropped out of the sky about 50 feet away. I retrieved the wet drone and dried it out over 3 days. It powers up but has issues and DJI will not warranty the item despite the logs showing no user error, because - yes - I flew it over water - which is apparently a HUGE no no for all DJI product users.

Does that sound awesome to all of you or what?

1138studios Posted at 2016-1-29 13:00
It's too bad that DJI will not honor their warranty if your battery failed over water ---

I'm not ...

I am very sorry what happened, whomever you spoke to through email is in the warranty department and what they say goes. I am only on the forum to give technical support.

Interesting video. I was reading the other day that the latest tech in smart batteries is internal (self) warming.

DJI-Ken Posted at 2016-1-28 23:17
I am very sorry what happened, whomever you spoke to through email is in the warranty department a ...

Oh I know. Nobody at DJI can actually help their customers. It's OK. We are DJI customers. So we understand that helping us out when we need you to is just not every going to happen.

1138studios Posted at 2016-1-29 15:35
Oh I know. Nobody at DJI can actually help their customers. It's OK. We are DJI customers. So we u ...

I don't understand what you mean. I was stating the fact that the customer sent their unit in and determined the cause and offered a discount.
I said I'm sorry, but I am here on the forums for technical support and if the warranty department made a decision, that is up to them.
How did my statement in your eyes mean "nobody at DJI can actually help their customers" ?

afiend69 Posted at 2016-1-28 23:01
Interesting video. I was reading the other day that the latest tech in smart batteries is internal ( ...

The DJI batteries do "self warm" but I think it is mainly due to the amperage being drawn.

Made a couple of cold weather (-2°C) flights today.  Batteries were 85% charged and their starting temp was 17°C.  Took off and hovered close to the ground.  Monitored the battery page and in just under 2 minutes the battery temp was 25°C at which time I climbed out and made the flight.  No voltage alarms encountered.

DJI-Ken Posted at 2016-1-29 09:26
I don't understand what you mean. I was stating the fact that the customer sent their unit in and  ...

Well Ken, if you were a DJI customer who was told that "flying over water is discretionary and not recommended and will void your warranty if it fails, even if it's our fault" then you'd understand what I mean.  DJI actively encourages their customers to fly over water by time and again showing their units flying over water, and they never tell anyone that your warranty will be voided by doing this, and expect the customer to pay for a new unit and want us to say thank you because they give you a "discount" and allow you to buy another unit after you've already spent $1400.
And no matter how many times you call or email and ask them to show you were DJI says they will not honor their warranty if your unit falls into Wtaer - no one will ever be able to do so, because this was never said and is never said.
DJI encourages their customers to carry on in an activity that they know will void their warranty, but they never actually say this until it happens.

So, hopefully that clears this up for you DJI guy.

This exact thing happened to me on January 30, 2016, 55 degrees F outside, battery at 53%, 3 minutes into flight "critical low voltage" , down through the trees she came. She is still under warranty, sent her back to DJI.  

Now my checklist has a new item ;-) Thank You for sharing. Luigi

Very informative, thanks for taking the time to do this!  A picture (video ) is worth a thousand words!

Hi, sorry for 'opening' this topic after quite a long time.
Thanks for the very informative video.

Could it happen to the batteries in my Mavic too?
I flew today with a partially charged battery (38%), right after take off I switched to sports mode which I presume takes a lot of 'sag' on the battery, after a 500 metres I got the critical low voltage error. Luckily I recoverd from the auto landing and made it back home.

Robert09 Posted at 2016-12-16 13:08
Hi, sorry for 'opening' this topic after quite a long time.
Thanks for the very informative video.

I would expect so because all lipos have degraded performance when they are cold.

I would expect that too. I'll try to reproduce the same circumstances and see if it will occur again.

Peut Posted at 2016-12-20 17:45
I would expect that too. I'll try to reproduce the same circumstances and see if it will occur again.

If its cold outside its good practice to take off and hover and then check the battery temperature which can be displayed on the controller.  If it's higher than 22°C you can fly without battery warnings.