Good info and thanks for sharing. May be if you post the serial# of the battery, others with the same batch may consider buying new ones - hopefully with better soldering
Cart before the horse. We have no idea if even a single crash has been caused by this yet.
As mentioned Go4 but also if you fly Litchi you can subscribe for free to Airdata UAV and it gives a great battery breakdown per cell including anomalies.
That one the battery dropped to half and the conclusion is conjecture:
I would submit that dropping to half is NOT what would happen with a terminal failure or a battery popping loose. It really is an all or nothing proposition.
No, you're not reading correctly. This is the graph of what it looks like when the battery pops out of the drone or there is some kind of battery failure where all power from the battery is immediately lost. The motor volts drop slowly rather than instantly to 0, as they drain the capacitors.
It doesn't just drop to half - but when it gets below that value the logging stops because the FC shuts down.
Just go and test it if you really doubt the explanation.
There are no capacitors of sufficient capacity to retain voltage for any significant time. Those motors draw amperes of current.
Best guess from the experts is that the Mavic motors draw around 12 amps each.
R = V/I = 1 ohm resistance.
Voltage across the capacitor - we'll use half since that is the data file value - is the formula
Vc is the voltage across the capacitor, Vs is the source (12 V), t is the time in seconds, RC is the time constant of the resistor capacitor combination.
So the motor resistance is 1 ohm - let's assume the time from the graph is 20ms which is the time interval of logging (1/50 Hz)
A 1000 uf capacitor (.001 farad) would be a very large one for the circuit board so the resulting time constant with the motor is .001.
Solving for t as a value to result in half the supply voltage of 12 = 6 v gives us .0007 s as the time needed to discharge the capacitor to 6 volts.
I think we can agree that 7 ten thousands of a second is far far less than the 20 ms sample time - it's nowhere near fast enough to detect.
if you wanted to solve for having 6 volts at the end of the 20 ms sample window the necessary capacitor would be .02 farad - a physically huge capacitor that would likely be bigger than the battery.
So I stand by my assertion in that there's no capacitor large enough to keep the motors running for anywhere near 20ms.
There are no capacitors of sufficient capacity to retain voltage for any significant time. Those motors draw amperes of current.
Best guess from the experts is that the Mavic motors draw around 12 amps each.
R = V/I = 1 ohm resistance.
Voltage across the capacitor - we'll use half since that is the data file value - is the formula
Vc is the voltage across the capacitor, Vs is the source (12 V), t is the time in seconds, RC is the time constant of the resistor capacitor combination.
So the motor resistance is 1 ohm - let's assume the time from the graph is 20ms which is the time interval of logging (1/50 Hz)
A 1000 uf capacitor (.001 farad) would be a very large one for the circuit board so the resulting time constant with the motor is .001.
Solving for t as a value to result in half the supply voltage of 12 = 6 v gives us .0007 s as the time needed to discharge the capacitor to 6 volts.
I think we can agree that 7 ten thousands of a second is far far less than the 20 ms sample time - it's nowhere near fast enough to detect.
if you wanted to solve for having 6 volts at the end of the 20 ms sample window the necessary capacitor would be .02 farad - a physically huge capacitor that would likely be bigger than the battery.
So I stand by my assertion in that there's no capacitor large enough to keep the motors running for anywhere near 20ms.
Nice analysis, but you are assuming a purely resistive load. In this case the load is highly inductive and you have not taken into account inductively stored energy which may account for the difference. Time for an experiment - I'll do it if you are not planning to.
Hotelone
Well-Known Member
Since about two DJI AC "updates" ago AirData has been unable to pull specific data for individual batteries, instead lumping all as "Default Mavic Pro Battery". This was a very valuable feature for me, have you found a way to keep track now? (Sorry for the hijack...)
Nope - sorry but if the motor draws that much juice it appears as that load resistively since it is the ESC that is the load - not the motors. The ESC's converts the DC into stepped phases to drive the stepper motors.
I'd hope that you could see there's simply no way that much capacitance is in the system but as usual y'all prefer to argue about pet theories even when I show the maths.
(edited to correct EMF to ESC - lol, EMF is the inductive backlash from a collapsing inductive field)
Last edited:
Hmmm - mine still shows the individual cell data and battery metrics. I just screen captured this to make sure.
Note that this is under flight information->power tab, not the batteries which is a premium feature.
In my opinion your maths is too simplistic and based on assertions about the nature of the load to be regarded as definitive. You may still be correct, which of course leaves the previously observed data as a bit of a puzzle, but we'll find out soon enough by testing. I suggest we reconvene once we have some good data to look at.
Yes it is simple but it's a standard technique in blackbox analysis to reduce complexity to simple to analyze forms.
At it's core the ESC is a resistive load since it consumes current. The fact that it transforms the current into another form to be delivered to a reactive load is immaterial since the load is isolated and we only care about and measure the current being consumed.
At it's core the ESC is a resistive load since it consumes current. The fact that it transforms the current into another form to be delivered to a reactive load is immaterial since the load is isolated and we only care about and measure the current being consumed.
Before you guys get too far off in the weeds. Looking at the motorVolts has proven to be a good way to determine if either the battery came loose or there was a battery failure.
Agreed, but it does not always follow that a simplification is correct. Remember that the voltages that we are talking about monitoring here are the voltages at the motor (those are the ones that are sampled at 50 Hz - not the battery voltages directly which are only 1 Hz), so they are in the inductive side of the circuit where the inductively stored energy is available.
I agree, but given that, we should still be able figure out that enough energy is stored to permit that to happen. I cannot criticize @Brojon for wanting to verify that.
Callum
Well-Known Member
- Joined
- Oct 10, 2017
- Messages
- 536
- Reactions
- 265
- Age
- 40
Excuse my ignorance but going to throw something in and you can correct me if I'm way off the mark.
Am I right in thinking the mavic motors are vector controlled AC motors?
I usually work with 20kw+ induction motors and when power is lost the voltage decays and shifts a phase as it coasts to a stop.
Could this decay in motor voltage explain the drop you see?
Am I right in thinking the mavic motors are vector controlled AC motors?
I usually work with 20kw+ induction motors and when power is lost the voltage decays and shifts a phase as it coasts to a stop.
Could this decay in motor voltage explain the drop you see?
I wouldn't think so since the voltage is converted into pseudo sinusoidal waveforms for the motor windings. The drivers are usually high power stepper driver chips containing power MOSFETs which will suppress any EMF from the motors. The chips have feedback (usually via serial) corresponding to the integral of the power delivered - not the actual voltage since it varies by a sine approximation.
Here's a good overview of a typical ESC.
Similar threads
DJI Drone Deals
New Threads
-
-
-
-
Texan and Californian Visiting Australia August, 2025- Started by Hauptmann
- Replies: 2
-
Forum statistics
