Recreated "falling from the sky" issue

[James Stout]

I just tested what temperature version 1.6 will let you start the motors. I put my phantom in the garage to do a cold IMU calibration (45 degrees ) tried to start motors after completion. Battery temp was 10 degrees C (50f) got the too cold message. Waited around while the bird was powered up slowly warming the battery up. When the battery reached 15c (59f) the motors started.

 

[alokbhargava]

The key issue is that the P3 power demands are likely exceeding what the battery can provide when it is not in optimal conditions. The motors are demanding more current than the battery can safely provide which is why the voltage sags. This is most likely at full throttle.

Motors may demand more currents but it's the control system that should lower the P3 slowly to limit the currents and should not switch off controls totally as observed by others. You must have seen in cheap drones that when the battery weakens, aircraft automatically drops height and lands. No point on maintaining position of the aircraft if battery can't provide enough power to hold. DJI needs to work to improve the control system loops.

 

[Trackman1]

That was an excellent video. I completely understand it now. Thanks.

 

[TheRealNick]

This evening while doing my monthly battery testing I was able to "accidentally" recreate the falling out of the sky "dead" issue. I flew several different locations today and came home with some partially discharged batteries. The last day of the month I log all of my batteries and deep discharge any that are around the 20 cycle mark (20, 40, 60, 80 etc).

I had a battery that was at 47% and I wanted to run it down to less than 10% to initiate the Recalibration function of the Smart Battery. I was watching the actual battery statistics while hovering just above eye level. I wanted to see just how much difference putting a heavy load on this battery would do so I (don't repeat this at home it's not safe) took hold of the landing skid and applied more throttle. The motors revved hard but the battery indicator was barely showing red on one cell so I increased the throttle more. Tx immediately indicated "Critically Low Voltage" (still showing 45% capacity but voltage was red on every cell). Within about 2 seconds the aircraft completely powered down and had I not been holding the landing skid it would have tumbled to the ground.

So if you are trying to get a few more minutes by flying with a partially charge pack realize you could very easily cause a complete power-down.

Air temp was roughly 54degs so it was not a cold battery issue at all.

Good luck and fly safe.

How long had the battery been sitting around? I have been trying to recreate a false read of the battery, but my batteries only sit a couple of days max before I fly again...I am lucky...

 

[BigAl07]

How long had the battery been sitting around? I have been trying to recreate a false read of the battery, but my batteries only sit a couple of days max before I fly again...I am lucky...

I had been out flying that day previously. I flew all (4) batteries on Saturday and topped them all off Sunday morning prior to going to the job site.

 

[TheRealNick]

I had been out flying that day previously. I flew all (4) batteries on Saturday and topped them all off Sunday morning prior to going to the job site.

So you flew that battery down to 47% earlier that day and it correctly read as 47% when you installed it to further lower it?

 

[BigAl07]

So you flew that battery down to 47% earlier that day and it correctly read as 47% when you installed it to further lower it?

I flew for a job earlier in the day but didn't need to fly anymore so I landed and shut down with 47% still on the battery.

Later in the day (maybe 4 hours later) I flew it for a couple of minutes just hovering it about 7' AGL. Then I decided to try and see what putting a heavier load on the system would do (honestly I just wanted to run the battery down sooner) and around 45% I grabbed the skid and applied more power. One cell indicated red but the other 3 were still yellow so I applied more power. That's when everything went RED into Critically Low and the aircraft shut down like someone had pulled the battery from the aircraft.

I hope to duplicate these settings again in the next couple of days to see if I can reliably replicate the failure situation. Then I'll update the FW to confirm/deny it resolves the problem IF I can reliably recreate it.

 

[shockwave199]

I agree about running with 1.6. I do not subscribe to running these batteries down to extremes for recalibration, though. There's a massive thread here somewhere about opinions on that. I fly sensibility without putting undo stress on the bird and I'm a believer in that extreme battery rundown is worse, not better for these batteries. I typically fly them down to between 30 to 40%, sometimes only to 50. I charge them fully before flight. I haven't had any problems and have not deep discharged either of my three batteries. My opinion, anyway. Anyone, anyone who is running without the battery update is seriously risking their bird.

 

[shockwave199]

I just tested what temperature version 1.6 will let you start the motors. I put my phantom in the garage to do a cold IMU calibration (45 degrees ) tried to start motors after completion. Battery temp was 10 degrees C (50f) got the too cold message. Waited around while the bird was powered up slowly warming the battery up. When the battery reached 15c (59f) the motors started.

I think you need it colder for a better cold imu calibration. I don't think 45° will give the proper results. Mid 30's yes. But mid 40's, no.

 

[ianwood]

Motors may demand more currents but it's the control system that should lower the P3 slowly to limit the currents and should not switch off controls totally as observed by others. You must have seen in cheap drones that when the battery weakens, aircraft automatically drops height and lands. No point on maintaining position of the aircraft if battery can't provide enough power to hold. DJI needs to work to improve the control system loops.

Completely agree.

Right now the battery is protecting itself by killing the host. The existing protection circuit in the battery is cut too close to operational current. Transient loads can trip it. And it really shouldn't exist. At a minimum, it should only be in case of short circuit, i.e. a substantially larger than possible current.

And a better solution is to implement limiting factors in the flight control algorithm that reduces throttle on over-current. No need to land as the sag will recover once the load is reduced.

I agree about running with 1.6. I do not subscribe to running these batteries down to extremes for recalibration, though. There's a massive thread here somewhere about opinions on that. I fly sensibility without putting undo stress on the bird and I'm a believer in that extreme battery rundown is worse, not better for these batteries. I typically fly them down to between 30 to 40%, sometimes only to 50. I charge them fully before flight. I haven't had any problems and have not deep discharged either of my three batteries. My opinion, anyway. Anyone, anyone who is running without the battery update is seriously risking their bird.

Agreed. I have a bunch of Inspire batteries that are claiming they need to be deep cycled. I honestly don't want to do it. 8%, especially when it is so inaccurate, is really low and I can't see how damage isn't done.

 

[Jeff48920]

My opinion, anyway. Anyone, anyone who is running without the battery update is seriously risking their bird.

See my earlier post where the battery and the bird were both updated and the battery was over 68 degrees F and the battery went into Critical from 35%. I'm not convinced that DJI's patch protects against Critical on less than full batteries. I am now recharging to 100% before trusting the FW patches. YMMV but following DJI's recommendations seems the prudent course to follow.

 

[shockwave199]

See my earlier post where the battery and the bird were both updated and the battery was over 68 degrees F and the battery went into Critical from 35%. I'm not convinced that DJI's patch protects against Critical on less than full batteries. I am now recharging to 100% before trusting the FW patches. YMMV but following DJI's recommendations seems the prudent course to follow.

Yes, I agree. I only fly with fully charged batteries. Lifting off at 35% under any circumstance is not a good idea at all.

 

[Dirty Bird]

Like you I have 4 decades of RC experience & currently own only exclusively Li-po powered aircraft. I'd say the difference between most RC & the Phantom is a Li-po pack is just a pack. It has no intelligence & you can run it to empty. The Phantom packs have built-in management electronics & firmware. It is the management firmware shutting down the pack that seems to be causing the problem.

I should have left a few more details from the get-go.

As a side note: I've been flying R/C for 4 decades and working with LiPo for at least the last several years. I have over 40 different sizes and types of LiPo in my possession (I fly planks, helis and MRs) and normally fly in varied temps. I can say with almost certainty that flying a battery at 54 deg is not going to be a huge determining factor in LiPo performance. I've logged flights in temps as low as 18deg F just in the last month with zero battery issues. I do preheat my packs when I'm flying in colder temps but anything above 45 I don't fool with this. The battery comes up to operating temp very quickly once the motors are engaged.

Also the ambient temp (outside) was 54deg but the battery and aircraft had been inside my house for a couple of hours which remains 69/70 during normal waking hours. I am pretty sure the LiPo was not cool enough to be problem but the next opportunity I have I'll try and recreate it again. Also note that I had been hovering the aircraft and inspecting battery levels as opposed to powering up and immediately going into a high load climb. Battery would have been up to operating temperature very quickly even if it had not been inside at 70deg immediately prior to the flight.

Just for kicks-n-grins I'll try to get the Healthy Drones report on this incident tonight and when I do I'll upload that info here as well. This way it can be a valid learning experience and not just a hunch and a guess.

 

[TheRealNick]

I agree about running with 1.6. I do not subscribe to running these batteries down to extremes for recalibration, though. There's a massive thread here somewhere about opinions on that. I fly sensibility without putting undo stress on the bird and I'm a believer in that extreme battery rundown is worse, not better for these batteries. I typically fly them down to between 30 to 40%, sometimes only to 50. I charge them fully before flight. I haven't had any problems and have not deep discharged either of my three batteries. My opinion, anyway. Anyone, anyone who is running without the battery update is seriously risking their bird.

I agree with you, I try to land shortly after 30% and have no plans to ever discharge them further. I believe it will decrease their overall life. I may consider trying a deeper discharge if the cells become unbalanced, but I am under the impression that the battery has circuitry to balance them when charging....

 

[tcope]

I agree with you, I try to land shortly after 30% and have no plans to ever discharge them further. I believe it will decrease their overall life. I may consider trying a deeper discharge if the cells become unbalanced, but I am under the impression that the battery has circuitry to balance them when charging....

Balance charging is different. The deep discharge serves to make sure the firmware know the upper and lower limits of the battery. If the battery us always only taken down to 30% there is the possibility that the firmware will make an error and now think that 30% is (effectively) 0%. So you'd really be running the battery from 100% down to 60% and not 30%. Charging each cell up equally is something totally different.

 

[TheRealNick]

Balance charging is different. The deep discharge serves to make sure the firmware know the upper and lower limits of the battery. If the battery us always only taken down to 30% there is the possibility that the firmware will make an error and now think that 30% is (effectively) 0%. So you'd really be running the battery from 100% down to 60% and not 30%. Charging each cell up equally is something totally different.

I could be wrong, but I do not believe that the battery tracks the state of charge. I think it is tracked by the Phantom software. I have been able to almost exactly match the remaining battery percentage by calculating from the voltage and current data points in the .dat file and subtracting that from the initial capacity. In my days of working on vehicle SOC that was the most accurate way to track it from a fully charged battery. Most vehicle batteries only allow the battery to operate between 35% - 100% of capacity to increase the life of their battery....so, until I see a change in flight times or imbalance between the cells, my personal plan is to fly down to about 30% and land and not to intentionally go any lower. Once I have a problem with either reduced flight times or cell imbalance I will investigate further and report back.

 

[Mark The Droner]

I could be wrong, but I do not believe that the battery tracks the state of charge. ...

What do you mean by "state of charge?" Charge percentage? The battery can report to you a range of the charge percentage by the number of lights it shows on a button check. Also, I have a couple Venom aftermarket batteries, and they show the charge percentage in digits on a screen right there on the face of the battery. Finally, one of the reasons we do a deep discharge is to calibrate the reading so that the battery doesn't report bad information (i.e. inaccurate charge percentage) - as tcope explained above.

 

[TheRealNick]

What do you mean by "state of charge?" Charge percentage? The battery can report to you a range of the charge percentage by the number of lights it shows on a button check. Also, I have a couple Venom aftermarket batteries, and they show the charge percentage in digits on a screen right there on the face of the battery. Finally, one of the reasons we do a deep discharge is to calibrate the reading so that the battery doesn't report bad information (i.e. inaccurate charge percentage) - as tcope explained above.

Yes battery percentage, I think the lights on the battery are just based on voltage level but that the algorithm the Phantom uses is more complex.

 

[tcope]

I could be wrong, but I do not believe that the battery tracks the state of charge. I think it is tracked by the Phantom software.

You can fly without the app and the Phantom will know how much power is left and go into RTH and auto landing. So it's tracked by the battery, not the app.

 

[Mark The Droner]

You can fly without the app and the Phantom will know how much power is left and go into RTH and auto landing. So it's tracked by the battery, not the app.

Yes. This was discussed at length in another thread a while back, and that's what was concluded.