Mavic 2 crasched

[EricJT]

Basing a % strictly on voltage only works if remaining capacity is linear to voltage, but it isn't. Not only is it not linear, but it's not fixed either. Age and temps are among the other variables that determines remaining capacity. As the battery ages, the bottom out voltage goes higher as the cell internal resistance increases. More power demand will suddenly drop the voltage when the internal resistance increases.

There's also the thought that if the battery is in the middle of self-discharge mode, the discharge circuit is still connected during use if not placed on charge first to turn it off. This will distort the calculations that the BMS is attempting to make.

Your post took me back to my early electricity studies :

Voltage (V) = Current (I) x Resistance (R)

Are you sure about that? I was under the impression that the self-discharge is via the individual cells bleed resistors that are used for balancing during charging. Either way, it's pretty clear from battery data that the cells become unbalanced during self-discharge.

All Li-Po cells are not created equal.

 

[sar104]

All Li-Po cells are not created equal.

Agreed, but I think that the point that I was replying to was that if it were significant cell differences at play then they should go equally out of balance and cause problems as the battery drops to 50% during flight.

 

[Mav1c]

I think it’s all relative to your current flight environment. Can you fly 5 minutes here and there on one battery throughout a day? Yeah if you were just flying in a field or taking RE shots for example. If something goes wrong you have the ability to come right down and swap your battery out. But it’s when pilots fly out over water downwind at 400ft with 60% battery that will get you in a tough situation. (Not saying you flew downwind but it’s a common theme among new pilots)

Using the Tesla example, you generally wouldn’t leave your electric car unplugged overnight with 40% battery having planned a road trip the following morning. It’s all about pre planning and pre flight checks.

 

[DanMan32]

Are you sure about that? I was under the impression that the self-discharge is via the individual cells bleed resistors that are used for balancing during charging. Either way, it's pretty clear from battery data that the cells become unbalanced during self-discharge.

That's what someone else pointed out in one of the other threads.

 

[sar104]

That's what someone else pointed out in one of the other threads.

I remember that post. I cannot vouch for it's accuracy or whether it is true for all models. Perhaps @everBit can elaborate.

Battery discharge cycle for 3-month maintenance...down to 10%? 8%? 0%?

There is no attempt to balance cells during auto discharge. Do you have proof of this?

mavicpilots.com

 

[Frustrated]

Its like having a bad battery in your car. You check it and your test says 12v, but it won't start the car. Its all about amps, capacity that the battery has.

 

[EricJT]

Agreed, but I think that the point that I was replying to was that if it were significant cell differences at play then they should go equally out of balance and cause problems as the battery drops to 50% during flight.

Sar104, given V=IR or R=V/I, if the battery terminals have aged and have more resistance over time, what would the effects on the voltage and current be compared to less resistive battery terminals?

 

[Capt KO]

Its like having a bad battery in your car. You check it and your test says 12v, but it won't start the car. Its all about amps, capacity that the battery has.

That’s the correct Correlation. Volts (%), with no relation to amperage. 12v to a useless 6v under load. The lipo voltage (%) shown is wrong during self discharge but with no warning until put under a load. Ok news for a car but not for a drone. Click click click click click....Dang, need a jump.

Useful thread, helpful warning.

 

[LukiDrone]

Sar104, given V=IR or R=V/I, if the battery terminals have aged and have more resistance over time, what would the effects on the voltage and current be compared less resistive battery terminals?

When testing the capacity by measure the voltage, you are reading it without load, where R tent to infinit and current tent to zero. This is the open loop voltage. Whe you start the motors Then you have a load there the voltage will follow ohms law and depend on the current that the battery can provide and the trotle you give to the motors. More throtle, less resistant (or if controlled by an ESC the power is controlled by PWM Pulse-width modulation - Wikipedia), if bat keep up the current the voltage kept constant, if bat can not keep up, voltage drop. This curve is not linear and depend on bat type and age.

 

[GadgetGuy]

It looks like you started with a partially discharged battery. If the battery has been self-discharging then the initial estimate of remaining capacity is very unreliable. As soon as significant current is drawn the cell voltages collapse, triggering the autoland and the recalculated battery %.
As for the crash - you got it back to dry land just fine, but flew it into the rock wall under full control:

Yes, the crash was completely avoidable, even under these dire conditions. Autoland due to low voltage, whether at 0%, or even up to 2 minutes after 0%, still gives the pilot full directional control, while delaying autoland, and even ascending during it, with full left stick. Been there. Done that. Not for the faint of heart! Only after any single cell reaches 3.0V does the voltage become so low that it can no longer sustain flight. "Tested" that, too!

 

[GadgetGuy]

Thanks so much for your analysis!
Now I know I should never start with a partially discharged battery.
I had map mode enabled to be able to quickly steer the drone back to me, standing on a 25 m high ledge. This meant that I did not see that the drone had lost so much height.
Thanks again!

Now that you know you still had full pilot control, despite 0% and autolanding, best to use the map view to get it within a couple hundred feet, then look up, and locate it in the sky, or hover it until you can, and then fly it back the rest of the way with VLOS only. Avoids overshooting the landing, and you keep it above you for a quick handcatch! The landing lights can also be activated, if set up on C1 or C2, to help locate it in the sky, especially if you briefly fly it backwards, while facing you, nose in, even in the dark, as the landing lights are in the rear, and point away from you otherwise, during rapid forward flight, making them harder to see. Post mortems make for great teachable moments!

 

[sar104]

Sar104, given V=IR or R=V/I, if the battery terminals have aged and have more resistance over time, what would the effects on the voltage and current be compared less resistive battery terminals?

Increased resistance anywhere in the circuit will cause an increased voltage drop at the load.

 

[EricJT]

Increased resistance anywhere in the circuit will cause an increased voltage drop at the load.

Thx Sar104....

 

[romko1]

Hello,
when you mention M2P batteries, I have approx. 17 cycles (charge) on two batteries and shows me a battery life of 98%. I take care of the batteries according to the instructions and recommendations from the forum. Do you have the same experience? (Battery Life). Thanks.

 

[JSM]

My Mavic 2:s battery capacity dropped from 60% to 0% and the drone started landing over Water. I managed to switch to Map View on the Phone to quickly see the fastest way back to me. During the way back it's altitude decreesed and it ended with a crasch into a rock which you can se here:

I have the DJI Care Refresh which made DJI to send me a new drone, but they never explained what that had gone wrong. I am curious to find out why the battery syddenly went from 60% to 0%.

Why didn't you just land on the edge of the Marina????? (Very lucky you didn't hit a car!!!!!!)

 

[bigswifty]

Interesting thread, I've learned a lot! I don't know if it helps (except with my OCD) but I've mapped one of the programable buttons to show the battery state and tend to check it from time to time, although usually I just check whether the cells are closely matching in voltage.

 

[MAvic_South_Oz]

(Very lucky you didn't hit a car!!!!!!)

Yep, there were quite a few following that white van type vehicle it just missed, the one(s) behind may have had to brake, very lucky.