Mavic Battery and Speed efficiency chart

[GooniesNeverDie]

I'm not doing Leaderboard, Yes Litchi is cheating, then again so is flying the Mavic anyway. Which is essentially already the modern day equivalent of Fly by Wire aircraft.

Try it yourself empirically, don't take my word for it. 30mph will definitely get you further total distance than 20 mph. At least on the Mavic. I never flown a Phantom so can't speak for it.

Geez. Opinions not welcomed. I agree 30mph will get you further, but going full throttle will get you further but not back home.

You can theorize all day long. The reality is the leaderboard is based on real world testing.

The testing shows in many cases- sport mode, everything else off CA etc. and maintaining a perfect speed there and back.

My testing with this is that altitude makes little difference in efficiency from 100' to 400'. Although I do both.

If you full throttle your car on a straight stretch of roadway, and then do the same maintaining a consistent speed you will go further, assuming no coasting and momentum.

There becomes a point that going as fast as you can will become inefficient, because your also depleting your battery at the maximum rate.

I'm no scientist, but been doing these drones for distance for a long time.

 

[adiru]

Well then here is one datapoint:

17 minutes total flight, landing with 31% battery left. Distance 6 miles. Cruising speed ~30mph.

YouTube video:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

Mission as flown:

Mission Hub - Litchi
Airdata UAV - Flight Data Analysis for Drones

The Airdata/healthydrone data is abridged so it gives false total mileage and max distance readings, since it stopped recording the moment RC lost signal. But total time is correct. If you export it to csv and use excel, examine rows:

2093
2094

Specifically examine column G (time stamp) To see when it last long contact and when it re-established.

The YouTube vid proves it made it all the way to the destination point as per the mission that it followed.

///////

Suffice it to say, had I flown slower, say cruising speed of 20 mph vs 30 mph, I wouldn't have made it back, much less back with 30% battery to spare.

I have flown countless missions at 20mph, earlier on, thinking it would be more fuel economical, but I was surprised to find out that flying at 30 and faster lead to much farther total distance.


///


Conclusion:

/////

Started with 100% battery, landing and shutdown with 31% battery, used total of 69% battery.
100% full bat is 3830 mAh, thus 69% of that is 2681 mAh
Total distance covered at 6 miles round trip is 31,686 ft
Thus, mAh/ft is 0.085 mAh
Total flight time in air is 17 minutes (16 minutes 58 seconds)
At average speed of 21.18 mph

That is of course accounting for the takeoff, landing, and recording in 4k 30fps, flying into the wind to take videos and get a better view of the mall once at destination, and the time spent just hovering in the air getting a GPS lock and waiting for Litchi to load and start the mission etc

But the ENTIRE trip as a whole averaged 21 mph with 0.085 mAh, which is significantly more efficent than the OP's 0.100 mAh.


///////

Let me re-measure with only the cruising segment. I elected to use a subset, from the point it lost signal to the point it reacquired signal.
Mission Hub - Litchi

Blackout duration was 9.536 minutes in length.
The total flight distance of this blackout was 4.1 miles
Thus average speed of this segment was 25.8 mph, measurably faster than the total entire overall trip speed of average of 21mph.

The question is should we expect mAH/feet to stay the same or decrease? (hence meaning faster speed is more efficient in terms of total length traveled)

Blackout period used up 47% of total battery, meaning 1800 mAH
That gets 0.083 mAh/feet

So again, average speed goes up, but average mAh/feet goes down.

At least between speed of 20 to 30, it is fair to say that faster is more fuel efficient.

 

[adiru]

Geez. Opinions not welcomed. I agree 30mph will get you further, but going full throttle will get you further but not back home.

You can theorize all day long. The reality is the leaderboard is based on real world testing.

The testing shows in many cases- sport mode, everything else off CA etc. and maintaining a perfect speed there and back.

My testing with this is that altitude makes little difference in efficiency from 100' to 400'. Although I do both.

If you full throttle your car on a straight stretch of roadway, and then do the same maintaining a consistent speed you will go further, assuming no coasting and momentum.

There becomes a point that going as fast as you can will become inefficient, because your also depleting your battery at the maximum rate.

I'm no scientist, but been doing these drones for distance for a long time.

But the point of contention is Mavic's top speed doesn't allow it to go fast enough for it to reach such point of diminishing returns that you speak of. DJI already put in nanny protocols for safety. Mavic's top speed could be more if it allowed larger pitch angles and even faster RPM, but like you mentioned, that would make it unsustainable and counter-efficient, which is probably why DJI capped it at 40mph anyway.

And OP is suggesting there is no difference between 20mph and 30mph, this is blatantly untrue.

 

[GooniesNeverDie]

But the point of contention is Mavic's top speed doesn't allow it to go fast enough for it to reach such point of diminishing returns that you speak of.
And OP is suggesting there is no difference between 20mph and 30mph, this is blatantly untrue.

I think we agree. 30 to 31 is about the best it can do.

Full throttle in sport isn't

 

[adiru]

I think we agree. 30 to 31 is about the best it can do.

Full throttle in sport isn't

Well to be fair I never did full throttle since my RC signal cuts at just over a mile and Litchi only supports top speed of 33 since it doesn't work in sports mode.
Once thing for sure, 20mph will get you much less mileage than 30mph.

 

[Andrew F]

The data is interesting but my experience is the same as above, i get more distance at 30mph than at full speed on a phantom 4
Haven.t done enough long distance runs in the mavic

Right. There are a lot of variables I didn't account for like the battery overheating on longer runs etc. I only measure 3,000 foot runs with a hover in between.

 

[Andrew F]

Well then here is one datapoint:

17 minutes total flight, landing with 31% battery left. Distance 6 miles. Cruising speed ~30mph.

YouTube video:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

Mission as flown:

Mission Hub - Litchi
Airdata UAV - Flight Data Analysis for Drones

The Airdata/healthydrone data is abridged so it gives false total mileage and max distance readings, since it stopped recording the moment RC lost signal. But total time is correct. If you export it to csv and use excel, examine rows:

2093
2094

Specifically examine column G (time stamp) To see when it last long contact and when it re-established.

The YouTube vid proves it made it all the way to the destination point as per the mission that it followed.

///////

Suffice it to say, had I flown slower, say cruising speed of 20 mph vs 30 mph, I wouldn't have made it back, much less back with 30% battery to spare.

I have flown countless missions at 20mph, earlier on, thinking it would be more fuel economical, but I was surprised to find out that flying at 30 and faster lead to much farther total distance.

So to make that "data" you'd need to run the exact same route at 20 MPH. However, as I said in another post, I was measuring 3,000 foot intervals, not 30,000. The battery drain may have non-linear results.

 

[adiru]

So to make that "data" you'd need to run the exact same route at 20 MPH. However, as I said in another post, I was measuring 3,000 foot intervals, not 30,000. The battery drain may have non-linear results.

This weekend the area is expected to have little to no wind. I'm going to be flying multiple missions, simple straight lines, same heading, point A to point B, long distance, direct point to point, over a calm lake, non-moving body of water, multiple runs each at 20mph, 25mph and 30mph, at same altitude, over a meaningful distance (at least a mile), and collect the data from the Mavic's soldered SDcard itself if/incase RC signal lost (but since over a lake, probably it will sustain a mile with no issues) with Litchi screen recording and 4k video recording... I think it will settle it once and for all... And yes I'll publish the waypoint flight, and the full flight data information.

(My flight to the mall wasn't in straight lines, I used curved turns with Mavic sometimes pointed in not the same angle as its track, thus slowing it down unnecessarily. Otherwise it would have been more efficient and traveled further or used less battery % )

Secondly, I'm not interested in short intervals. So if the data from short intervals doesn't interpolate and scale up to longer distance, it is of much less relevance to me for my purposes.

BTW, I added to my post without seeing you had already replied / here is addendum repeated:

Conclusion:

/////

Started with 100% battery, landing and shutdown with 31% battery, used total of 69% battery.
100% full bat is 3830 mAh, thus 69% of that is 2681 mAh
Total distance covered at 6 miles round trip is 31,686 ft
Thus, mAh/ft is 0.085 mAh
Total flight time in air is 17 minutes (16 minutes 58 seconds)
At average speed of 21.18 mph

That is of course accounting for the takeoff, landing, and recording in 4k 30fps, flying into the wind to take videos and get a better view of the mall once at destination, and the time spent just hovering in the air getting a GPS lock and waiting for Litchi to load and start the mission etc

But the ENTIRE trip as a whole averaged 21 mph with 0.085 mAh, which is significantly more efficent than the OP's 0.100 mAh.


///////

Let me re-measure with only the cruising segment. I elected to use a subset, from the point it lost signal to the point it reacquired signal.
Mission Hub - Litchi

Blackout duration was 9.536 minutes in length.
The total flight distance of this blackout was 4.1 miles
Thus average speed of this segment was 25.8 mph, measurably faster than the total entire overall trip speed of average of 21mph.

The question is should we expect mAH/feet to stay the same or decrease? (hence meaning faster speed is more efficient in terms of total length traveled)

Blackout period used up 47% of total battery, meaning 1800 mAH
That gets 0.083 mAh/feet

So again, average speed goes up, but average mAh/feet goes down.

At least between speed of 20 to 30, it is fair to say that faster is more fuel efficient.

If this correlation holds, then we can interpolate that 30mph would yield 0.080 mAh/feet
Essentially going from 20mph to 30mph would expect to get you extra 20% battery efficiency.

 

[Andrew F]

Cool. Note the before and after mAH readings and time the flight manually to verify real speed. But "once and for all" won't happen. Just another data set. My data is 100% valid. If we get significantly different foot per mAH results, there must be other variables at play.

 

[adiru]

Cool. Note the before and after mAH readings and time the flight manually to verify real speed. But "once and for all" won't happen. Just another data set. My data is 100% valid. If we get significantly different foot per mAH results, there must be other variables at play.

Yes, like perhaps DJI not being honest and fudging the numbers. (perhaps on the mAH readings and/or the % battery left readings, or both) The problem is, no one is ever going to risk flying their Mavic down to absolute zero battery until it actually drops out of the sky.
But, big picture, under no wind conditions, I can routinely fly 6+ miles with more than 30% battery left. The distance doesn't lie, so that only leaves the battery level. I suppose I could hover the rest of that 30% down to 10% to make sure that 30% left was actually there... but I haven't done that yet. Assuming the battery % is accurate enough, So its long distance mAh is less than 0.100 and that isn't taking account of non-efficient stages of flight like takeoff, landing, hovering waiting for Litchi to load and GPS lock, etc.

Regardless, I'm somewhat less interested in what is the true mAh/feet and more interested to know how speed affects efficiency over long distances for the Mavic...


Update: I looked at the own data and pulled a segment from time index 890440 to 935930, a period of 45 seconds or so, averaging over 30mph, and getting 0.055 mAH for that period. Of course 45 seconds is really short, but 0.05 is really significant.

This time, 57 seconds, aveerage 30mph, yielding 0.06 mAH

 

[Andrew F]

Yes, like perhaps DJI not being honest and fudging the numbers. (perhaps on the mAH readings and/or the % battery left readings, or both) The problem is, no one is ever going to risk flying their Mavic down to absolute zero battery until it actually drops out of the sky.
But, big picture, under no wind conditions, I can routinely fly 6+ miles with more than 30% battery left. The distance doesn't lie, so that only leaves the battery level. I suppose I could hover the rest of that 30% down to 10% to make sure that 30% left was actually there... but I haven't done that yet. Assuming the battery % is accurate enough, So its long distance mAh is less than 0.100 and that isn't taking account of non-efficient stages of flight like takeoff, landing, hovering waiting for Litchi to load and GPS lock, etc.

Regardless, I'm somewhat less interested in what is the true mAh/feet and more interested to know how speed affects efficiency over long distances for the Mavic...


Update: I looked at the own data and pulled a segment from time index 890440 to 935930, a period of 45 seconds or so, averaging over 30mph, and getting 0.055 mAH for that period. Of course 45 seconds is really short, but 0.05 is really significant.

This time, 57 seconds, aveerage 30mph, yielding 0.06 mAH

So do you have a hypothesis as to why I got 12 data points (12 flights) with such consistent results? It's very possible our birds are dissimilar. Perhaps you have more efficient motors. What is your date of manufacture? If your test shows a .06 then there must be a causal factor.

 

[adiru]

So do you have a hypothesis as to why I got 12 data points (12 flights) with such consistent results? It's very possible our birds are dissimilar. Perhaps you have more efficient motors. What is your date of manufacture? If your test shows a .06 then there must be a causal factor.

I got all excited hoping to prove that for a segment of 20mph, I would show that it was using up more mAh/feet but it acutally calculated to be 0.03... I don't know whats going on, something isn't right. Maybe my data is corrupted since I lost signal and a large chuck wasn't included. I should have pulled it from the Mavic but by now its overridden and I don't see it on the Mavic's internal storage anymore.

Anyway I'll do a controlled test this weekend if wind conditions permit. Not flying straight and increasing and decreasing in altitude etc doesn't really yield a good accurate result.

 

[adiru]

So do you have a hypothesis as to why I got 12 data points (12 flights) with such consistent results? It's very possible our birds are dissimilar. Perhaps you have more efficient motors. What is your date of manufacture? If your test shows a .06 then there must be a causal factor.

Airdata UAV - Flight Data Analysis for Drones
Based on airdata my serial is 08RDDB200104DV
DJI Mavic Pro Serial Number Decoder
I used lookup and it was
Your Mavic Pro Main Controller Board (08RDDB200104DV) was manufactured on Wednesday, November, 2nd 2016

 

[Andrew F]

I got all excited hoping to prove that for a segment of 20mph, I would show that it was using up more mAh/feet but it acutally calculated to be 0.03... I don't know whats going on, something isn't right. Maybe my data is corrupted since I lost signal and a large chuck wasn't included. I should have pulled it from the Mavic but by now its overridden and I don't see it on the Mavic's internal storage anymore.

Anyway I'll do a controlled test this weekend if wind conditions permit. Not flying straight and increasing and decreasing in altitude etc doesn't really yield a good accurate result.

I was in as controlled an environment as possible. Only variable was speed. 100 feet in air. Exact same route. Distance and time verified and speed recalculated. My numbers are accurate for my Mavic for a 3,000 foot run. Looking forward to seeing your numbers.

 

[Logger]

Your figures are not correct at least not the first one.

That would be because I was using OPs data as an example. The principle of wind effect I explained is correct though and will scale from a bumblebee to an Airbus A380.

A pilot doesn't fly at 40 mph which is stall speed for any airplane.

Hmmmm .

The Mavic shape generates lift at speed, easing the power required by the quad to stay in the air in the first place. the largest vector force component being gravity itself. Which isn't true for conventional fixed wing aircraft.

This effect is dubious at best and unproven.
Easy enough to test though. Attach a fairing atop the mavic flatten or even making it concave.. Fly it at high speed in sport mode, with and without the fairing and observer the change in power setting (if any) to achieve the same speed.



Personally, I am still a little doubtful about the idea that mAh per foot is contant > 20mph. However @Andrew F 's methodology looks sound and he has data to back it up. I think it was a very clever test. Like he said there may be some un-considered variable in play. Shorter sector length making acceleration period comprise a greater % of flight perhaps?

To date I have been flying at or just over 30mph in nil wind for max range. It is what people have always said is the way to go. It is also the RTH speed DJI use when in sport mode (with Smart RTH collision avoidance off). I suspect this is no accident.

BTW - Last longish flight I did was 6.6 miles and it resulted in a mAh per foot of 0.082. Flying out in to an 8mph headwind with a hi ~ 60%power setting and coasting back with a tailwind and holding the power back at 26% on the stick. I could have improved this by flying even slower on the return leg I guess.. Outbound into wind 0.110 mAh/ft and on the way back 0.56 mAh/ft. Not a data point. Just an anecdote.

 

[Andrew F]

The .1 (ish) mAH per foot is not the crucial data point. Remember I had to accelerate from a dead stop and I also had to perform a 180 degree turn so that number is overtaxed compared to a longer flight and the shorter the flight, the more energy used as a % of time flying. I personally believe a .085 mAH at 30 MPH for long distance is probably accurate. My data just says that it is effectively comparable from 20 on up. But the curve was still coming down at 20. As I think about it more, I agree that a flat line result given my testing parameters might equate to a continued decrease in mAH / foot rate. I could retest on a calm day going just one direction and starting at full speed. But for the purposes of my son's project (now concluded) the lack of a drag effect is what was being tested. The mAH / foot was just a byproduct.

 

[wsalopek]

Helicopter power vs speed...same principles as Mavic (discounting any "lifting body" effects)...in the chart, it's apparent that between about 40 and 80 mph, the total curve is pretty flat...which equates to the Mavic's curve between about 20 and 40. If the Mavic could go faster than 40, it's certain that it's curve would start to rise, just like the helicopter, due to wind resistance.

Any differences in the shapes of the helicopter curve vs the Mavic's curve are probably due to the ability or inability to measure more accurately (or again, perhaps because of other factors, like, perhaps, the "lifting body" effect):

 

[MichaelG]

My son did a middle school science project using the Mavic. See the attached graph. He (and I) expected a different result. I believe the results very clearly show that at 20 MPH and up to max speed, the battery use per foot traveled is equivalent. Therefore, there is no battery penalty for "coming home" fast (e.g. in sport mode) vs. slower. And, as has been discussed previously, below 20 MPH, you will actually use more battery per foot traveled. The "age old" question of the most efficient speed has been answered - it is 20 MPH or higher. View attachment 12817

Congratulations. Great project and useful information. Thanks

 

[TTFullTimer]

In all of this the percent of battery used is going to be dependent on how good the curve for voltage versus discharge that the Mavic uses in its estimation of battery usage. If it does indeed have a shunt to measure current draw and then uses the voltage integrating the watt hours drawn I would be surprised. I am not going to open my Mavic and go poking around trying to find a shunt. Maybe someone can do that. Or perhaps the shunt is in the battery itself. Who knows.

The amperage being used will affect the usable capacity of the battery ala Peukert. So the curves posted by the OP may be somewhat accurate as the Mavic goes further faster when the current used is reducing capacity.

If the distance flyers are finding a speed of 30 more or less MPH is optimum perhaps without any need for MAH knowledge we can say that they are probably more right than wrong.

And of course using the wind to advantage is a good thing if you are doing distance flying. That is just common sense.

On my Truggy batteries I have a charger that measures the total amp hours put back into the battery. It never agrees with the battery measuring device that reads a percentage of charge left.

JMHO based on 40 years of electric vehicle experience.

 

[Andrew F]

In all of this the percent of battery used is going to be dependent on how good the curve for voltage versus discharge that the Mavic uses in its estimation of battery usage. If it does indeed have a shunt to measure current draw and then uses the voltage integrating the watt hours drawn I would be surprised. I am not going to open my Mavic and go poking around trying to find a shunt. Maybe someone can do that. Or perhaps the shunt is in the battery itself. Who knows.

The amperage being used will affect the usable capacity of the battery ala Peukert. So the curves posted by the OP may be somewhat accurate as the Mavic goes further faster when the current used is reducing capacity.

If the distance flyers are finding a speed of 30 more or less MPH is optimum perhaps without any need for MAH knowledge we can say that they are probably more right than wrong.

And of course using the wind to advantage is a good thing if you are doing distance flying. That is just common sense.

On my Truggy batteries I have a charger that measures the total amp hours put back into the battery. It never agrees with the battery measuring device that reads a percentage of charge left.

JMHO based on 40 years of electric vehicle experience.

All great input.