Welcome Mavic Pilot!
Jump in and join our free DJI Mavic community today!
Sign up

That new battery shine

MavicFT1P

Well-Known Member
Joined
Apr 8, 2019
Messages
714
Reaction score
354
In flyers experience and/or opinion how many cycles of a Battery (Especially one that has been run done to <%10 every other flight) would you expect before you start to see a significant performance reduction. Flight time, or other?

I couldn't help but notice my new battery excelled recently.
The other (i only own two) had not seen a huge ammount of cycles but it was the more used, and the difference was noticable.

So to put it simply, is New Better, if only slightly?

And at 50-100 cycles are you going to see less time in mins or seconds.
For a long time i only owned one battery, recently two, so I cant call on personal experience.

I dont have any Super old batteries with 200 cycles to do test. No reason why ;)
 

WithTheBirds

Well-Known Member
Joined
Sep 2, 2018
Messages
900
Reaction score
497
To put it simply- yes..... Newer is better. Obviously.

Age, cycle count and care are all factors that contribute to capacity loss and performance when loaded.
 
  • Like
Reactions: MavicFT1P

Kilrah

Well-Known Member
Joined
Nov 12, 2016
Messages
5,513
Reaction score
2,820
Age
36
Loc
Switzerland
And a bit of luck. There's the occasional report of the 4 month old, <10 cycle battery that gets really bad anyway. Aka, somewhat unpredictable.
 

MavicFT1P

Well-Known Member
Joined
Apr 8, 2019
Messages
714
Reaction score
354
To put it simply- yes..... Newer is better. Obviously.

Age, cycle count and care are all factors that contribute to capacity loss and performance when loaded.
To the point. Works for me
Id be lying if some threads werent Alt-tab busy time
 

MavicFT1P

Well-Known Member
Joined
Apr 8, 2019
Messages
714
Reaction score
354
And a bit of luck. There's the occasional report of the 4 month old, <10 cycle battery that gets really bad anyway. Aka, somewhat unpredictable.
This is true. Unpredictable, towards the end can be a kick in the nutz
I keep an eye on voltage not percent with my older one
 

WithTheBirds

Well-Known Member
Joined
Sep 2, 2018
Messages
900
Reaction score
497
This is true. Unpredictable, towards the end can be a kick in the nutz
I keep an eye on voltage not percent with my older one
Time to empty and % available are calculated in real time having regard to measured battery parameters (including internal resistance) and average and instantaneous current draw. The depicted telemetry should be accurate. More accurate than any assessment you might make looking at voltages. The obvious exception being cell imbalance. You can track this by looking at flight logs. 100mv or more suggesting the pack should be monitored.
 

Raven Heights

New Member
Joined
Jun 29, 2019
Messages
1
Reaction score
0
i have a question i have a fstop charging hub and recently purchased a fourth battery the hub holds four and when returning home from a test flight to make sure gimbal was working properly the battery life on my phone for the mavic was 46-48 percent i brought the drone down turned it off like normal and drove home to put it on charging hub now for some reason the lower two leds flash as if starting to charge but after roughly 3 - 4 cycles of flashing they shut off and repeat this action until i remove it from charging hub or it stops itself i am trying to figure out why this is happening i can provide video to show whats happening if anyone has advice please help me this is the video of the issue im having i need help or advice as soon as possible please.
 

The Editor

Well-Known Member
Joined
Oct 12, 2016
Messages
535
Reaction score
302
Time to empty and % available are calculated in real time having regard to measured battery parameters (including internal resistance) and average and instantaneous current draw. The depicted telemetry should be accurate. More accurate than any assessment you might make looking at voltages. The obvious exception being cell imbalance. You can track this by looking at flight logs. 100mv or more suggesting the pack should be monitored.
Wrong.....
Voltage is absolute
Percentage is a guess/calculation/best guess via algorithm.

Voltage readout will always give the most accurate state of charge.
 

WithTheBirds

Well-Known Member
Joined
Sep 2, 2018
Messages
900
Reaction score
497
Wrong.....
Voltage is absolute
Percentage is a guess/calculation/best guess via algorithm.

Voltage readout will always give the most accurate state of charge.
I’m sorry but your assumption here cant be correct. Simply because an old/tired battery with increased internal resistance will likely still charge to 4.35V (HV LiPO) however won’t deliver anywhere near it’s rated usable capacity. Relying on cell voltages (both resting and loaded) will prove close to useless in gauging remaining capacity in use. The fuel gauging algorithms in our batteries will outperform your observations of cell voltage under all circumstances. Proven fact.
 

The Editor

Well-Known Member
Joined
Oct 12, 2016
Messages
535
Reaction score
302
I’m sorry but your assumption here cant be correct. Simply because an old/tired battery with increased internal resistance will likely still charge to 4.35V (HV LiPO) however won’t deliver anywhere near it’s rated usable capacity. Relying on cell voltages (both resting and loaded) will prove close to useless in gauging remaining capacity in use. The fuel gauging algorithms in our batteries will outperform your observations of cell voltage under all circumstances. Proven fact.
Oh right - 30 years flying everything RC monitoring voltage is my mistake then. Sorry.
 

WithTheBirds

Well-Known Member
Joined
Sep 2, 2018
Messages
900
Reaction score
497
Oh right - 30 years flying everything RC monitoring voltage is my mistake then. Sorry.
No need to be sorry. The amount of time you have done something says nothing in itself about whether it’s best practice.

Prior to the implementation of battery monitoring SOC in our RC models I was with you in watching voltages and flight times. That combined with keeping a track on how many Ah we put back into packs on the charger was the best we had.

To say cell voltage provides a more accurate determination of SOC than the remaining % depicted in DJI GO telemetry is simply wrong.
 

The Editor

Well-Known Member
Joined
Oct 12, 2016
Messages
535
Reaction score
302
No need to be sorry. The amount of time you have done something says nothing in itself about whether it’s best practice.

Prior to the implementation of battery monitoring SOC in our RC models I was with you in watching voltages and flight times. That combined with keeping a track on how many Ah we put back into packs on the charger was the best we had.

To say cell voltage provides a more accurate determination of SOC than the remaining % depicted in DJI GO telemetry is simply wrong.
I will continue to fly by voltage and always will.
There is absolutely no accurate correlation between voltage and an arbitrary percentage of 'fuel' left in a pack/cell.
There are a plethora of graphs and assumptions all over the internet and they all differ in trying to estimate a lipos voltage against remaining capacity.
The problem with DJI smart packs is that they are anything but - they were simply invented in an attempt to weed out the idiot gene.
The problem is, with an increasing IR of a pack, the effective 'end points' or full and empty changes over time (assuming trying to calculate percentage). Without periodic recalibration of 'full and empty' the algorithms are flawed.
DJI has tried over a number of years to hone their idiot packs but they still cause issues.
Flying by voltage will never bring surprises since simply going into the battery screen every minute or so will tell you instantly the per cell voltage under load and enable you to see a faster than normal drop in pack/cell voltage. An increasing internal resistance will simply show a larger voltage drop and flying by voltage means you will notice this and land.
Ignoring the daft percentage thing and flying down to around 3.4v per cell (and then monitoring cell rebound time) will give a far better indication of pack health.

Have always flown this way and I have never had a crash due to failed pack.
New does NOT always mean better.
It was simply a way to bring lipo technology to the masses and attempt to mitigate idiots prematurely destroying their packs.
 
  • Like
Reactions: RadioFlyerMan

WithTheBirds

Well-Known Member
Joined
Sep 2, 2018
Messages
900
Reaction score
497
I will continue to fly by voltage and always will.
There is absolutely no accurate correlation between voltage and an arbitrary percentage of 'fuel' left in a pack/cell.
There are a plethora of graphs and assumptions all over the internet and they all differ in trying to estimate a lipos voltage against remaining capacity.
The problem with DJI smart packs is that they are anything but - they were simply invented in an attempt to weed out the idiot gene.
The problem is, with an increasing IR of a pack, the effective 'end points' or full and empty changes over time (assuming trying to calculate percentage). Without periodic recalibration of 'full and empty' the algorithms are flawed.
DJI has tried over a number of years to hone their idiot packs but they still cause issues.
Flying by voltage will never bring surprises since simply going into the battery screen every minute or so will tell you instantly the per cell voltage under load and enable you to see a faster than normal drop in pack/cell voltage. An increasing internal resistance will simply show a larger voltage drop and flying by voltage means you will notice this and land.
Ignoring the daft percentage thing and flying down to around 3.4v per cell (and then monitoring cell rebound time) will give a far better indication of pack health.

Have always flown this way and I have never had a crash due to failed pack.
New does NOT always mean better.
It was simply a way to bring lipo technology to the masses and attempt to mitigate idiots prematurely destroying their packs.
Fair enough. Continue with whatever your comfortable with. I simply responded to your direct suggestion that what I was saying was wrong.

The current battery management algorithms are very good. Texas Instruments claim the time to empty and % capacity fuel gauging is within 1% accuracy. If your interested in understanding how it works you can search for Texas Instruments white papers on “Impedance track” and “fuel gauging”. You will see that increasing IR is measured directly and applied as one factor to inform the algorithm.
 
  • Like
Reactions: The Editor

The Editor

Well-Known Member
Joined
Oct 12, 2016
Messages
535
Reaction score
302
Fair enough. Continue with whatever your comfortable with. I simply responded to your direct suggestion that what I was saying was wrong.

The current battery management algorithms are very good. Texas Instruments claim the time to empty and % capacity fuel gauging is within 1% accuracy. If your interested in understanding how it works you can search for Texas Instruments white papers on “Impedance track” and “fuel gauging”. You will see that increasing IR is measured directly and applied as one factor to inform the algorithm.
Understood. Thumbswayup
 

MavicFT1P

Well-Known Member
Joined
Apr 8, 2019
Messages
714
Reaction score
354
I will continue to fly by voltage and always will.
There is absolutely no accurate correlation between voltage and an arbitrary percentage of 'fuel' left in a pack/cell.
There are a plethora of graphs and assumptions all over the internet and they all differ in trying to estimate a lipos voltage against remaining capacity.
The problem with DJI smart packs is that they are anything but - they were simply invented in an attempt to weed out the idiot gene.
The problem is, with an increasing IR of a pack, the effective 'end points' or full and empty changes over time (assuming trying to calculate percentage). Without periodic recalibration of 'full and empty' the algorithms are flawed.
DJI has tried over a number of years to hone their idiot packs but they still cause issues.
Flying by voltage will never bring surprises since simply going into the battery screen every minute or so will tell you instantly the per cell voltage under load and enable you to see a faster than normal drop in pack/cell voltage. An increasing internal resistance will simply show a larger voltage drop and flying by voltage means you will notice this and land.
Ignoring the daft percentage thing and flying down to around 3.4v per cell (and then monitoring cell rebound time) will give a far better indication of pack health.

Have always flown this way and I have never had a crash due to failed pack.
New does NOT always mean better.
It was simply a way to bring lipo technology to the masses and attempt to mitigate idiots prematurely destroying their packs.
I think your being a bit harsh.
I think DJI did pretty well given a 3S Lipo battery for 2016.
The fact none have caught fire, despite noob usage/charging is a testament
to them or the charger

Out of interest, How do you fly to 3.4v per cell?
Ive seen fluctuations below 3.5 not resulting in a landing, but 'using the non-dji battery option', I count
on my AC landing at 3.5V
 

kidroc

Well-Known Member
Joined
Aug 21, 2017
Messages
440
Reaction score
161
Age
43
I will continue to fly by voltage and always will.
There is absolutely no accurate correlation between voltage and an arbitrary percentage of 'fuel' left in a pack/cell.
There are a plethora of graphs and assumptions all over the internet and they all differ in trying to estimate a lipos voltage against remaining capacity.
The problem with DJI smart packs is that they are anything but - they were simply invented in an attempt to weed out the idiot gene.
The problem is, with an increasing IR of a pack, the effective 'end points' or full and empty changes over time (assuming trying to calculate percentage). Without periodic recalibration of 'full and empty' the algorithms are flawed.
DJI has tried over a number of years to hone their idiot packs but they still cause issues.
Flying by voltage will never bring surprises since simply going into the battery screen every minute or so will tell you instantly the per cell voltage under load and enable you to see a faster than normal drop in pack/cell voltage. An increasing internal resistance will simply show a larger voltage drop and flying by voltage means you will notice this and land.
Ignoring the daft percentage thing and flying down to around 3.4v per cell (and then monitoring cell rebound time) will give a far better indication of pack health.

Have always flown this way and I have never had a crash due to failed pack.
New does NOT always mean better.
It was simply a way to bring lipo technology to the masses and attempt to mitigate idiots prematurely destroying their packs.
This is how is have always done it it too.. For all rc not just quads voltage has always has and will always serve me well and is the only way to for me to feel safe when i fly i check voltage and cell balance every few minutes give you a nice reflection of your batts health in flight.
 

LINXXL

Well-Known Member
Joined
May 31, 2017
Messages
45
Reaction score
18
Age
52
No need to be sorry. The amount of time you have done something says nothing in itself about whether it’s best practice.

Prior to the implementation of battery monitoring SOC in our RC models I was with you in watching voltages and flight times. That combined with keeping a track on how many Ah we put back into packs on the charger was the best we had.

To say cell voltage provides a more accurate determination of SOC than the remaining % depicted in DJI GO telemetry is simply wrong.
Have you ever owned or flown an Inspire 1, If so you would definitely rethink your statement about percentage being more accurate than voltage. This especially applies to the TB 47 battery. The link to the Inspire section is at the bottom of the page, Take a look and see how many "Fell out of the air crashes" have happened with 100% at take off only to drop to nothing after a minute in the air. Data logs then show major discrepancies from percentage left to actual voltage.
 

WithTheBirds

Well-Known Member
Joined
Sep 2, 2018
Messages
900
Reaction score
497
Have you ever owned or flown an Inspire 1, If so you would definitely rethink your statement about percentage being more accurate than voltage. This especially applies to the TB 47 battery. The link to the Inspire section is at the bottom of the page, Take a look and see how many "Fell out of the air crashes" have happened with 100% at take off only to drop to nothing after a minute in the air. Data logs then show major discrepancies from percentage left to actual voltage.
Have you ever owned or flown an Inspire 1, If so you would definitely rethink your statement about percentage being more accurate than voltage. This especially applies to the TB 47 battery. The link to the Inspire section is at the bottom of the page, Take a look and see how many "Fell out of the air crashes" have happened with 100% at take off only to drop to nothing after a minute in the air. Data logs then show major discrepancies from percentage left to actual voltage.
Yes I have. I never had an issue either. Perhaps I was lucky- I am aware that some had problems. Perhaps they were similar issues to those with other DJI models where ensuring the packs are properly topped off close to when flown seems to be the cure.

You are perhaps making a flawed assumption here- DJI has no hand in the manufacture of the cells with the battery monitoring and management smarts being bought in from Texas Instruments. The TB47 issue was not Inspire or DJI specific.

The % remaining determined by the battery is, and must be, more accurate than determining capacity than looking at cell voltages. This is simply because the remarkably flat LiION discharge voltage profile makes the task, as you are proposing close to impossible. Just 100mv voltage difference = 10% SOC variation between 40 and 30% and thats in a region where voltage is dropping faster. Your proposed task is further complicated by the fact that voltage sag under load will make the SOC assessment almost meaningless.

If you understood how current battery monitoring SOC tech worked you would appreciate the SOC determination isn't the issue. The issue is cell behaviour when entering self or assisted discharge and the resultant imbalance of SOC between series connected cells.

Yes- I do look at cell voltages during the first minutes of flight- particularly in cold weather and with packs that haven't been charged recently (when in a remote area without a means to top off). If the cells are discharging unevenly it will present in the first several minutes- this has been evidenced by the flight logs where the failures you have described have occurred and with Phantom and Mavic AC also. Voltage imbalance is an important indicator of a likely issue- it is not a reliable means of determining SOC.
 

LINXXL

Well-Known Member
Joined
May 31, 2017
Messages
45
Reaction score
18
Age
52
Yes I have. I never had an issue either. Perhaps I was lucky- I am aware that some had problems. Perhaps they were similar issues to those with other DJI models where ensuring the packs are properly topped off close to when flown seems to be the cure.

You are perhaps making a flawed assumption here- DJI has no hand in the manufacture of the cells with the battery monitoring and management smarts being bought in from Texas Instruments. The TB47 issue was not Inspire or DJI specific.

The % remaining determined by the battery is, and must be, more accurate than determining capacity than looking at cell voltages. This is simply because the remarkably flat LiION discharge voltage profile makes the task, as you are proposing close to impossible. Just 100mv voltage difference = 10% SOC variation between 40 and 30% and thats in a region where voltage is dropping faster. Your proposed task is further complicated by the fact that voltage sag under load will make the SOC assessment almost meaningless.

If you understood how current battery monitoring SOC tech worked you would appreciate the SOC determination isn't the issue. The issue is cell behaviour when entering self or assisted discharge and the resultant imbalance of SOC between series connected cells.

Yes- I do look at cell voltages during the first minutes of flight- particularly in cold weather and with packs that haven't been charged recently (when in a remote area without a means to top off). If the cells are discharging unevenly it will present in the first several minutes- this has been evidenced by the flight logs where the failures you have described have occurred and with Phantom and Mavic AC also. Voltage imbalance is an important indicator of a likely issue- it is not a reliable means of determining SOC.
I have used the Inspire and TB 47 as a reference since it is easily found without having to search all over and I also have never had issues as I keep a good eye on voltage especially in temperature extremes. I never stated anything about it being specific to DJI. The flight times on the TB 47 tend to be more erratic due to the smaller capacity and heavy load (X5 camera) made a good example as there are multiple threads from the below links as well as the DJI sight. My FPV quads and my heli's do not display % of time left and only show voltage and I can tell you exactly when I am a minute or two from shutdown as these don't have return from home or auto land. I can also do the same with my DJI products as well. You bring up excellent points and I commend you on your knowledge but I know what I am capable of using and how it works for me. I thank you for your input and time taken and am grateful for your points and inputs as well as from others. That is what makes the Forum an awesome place for help
 

WithTheBirds

Well-Known Member
Joined
Sep 2, 2018
Messages
900
Reaction score
497
I have used the Inspire and TB 47 as a reference since it is easily found without having to search all over and I also have never had issues as I keep a good eye on voltage especially in temperature extremes. I never stated anything about it being specific to DJI. The flight times on the TB 47 tend to be more erratic due to the smaller capacity and heavy load (X5 camera) made a good example as there are multiple threads from the below links as well as the DJI sight. My FPV quads and my heli's do not display % of time left and only show voltage and I can tell you exactly when I am a minute or two from shutdown as these don't have return from home or auto land. I can also do the same with my DJI products as well. You bring up excellent points and I commend you on your knowledge but I know what I am capable of using and how it works for me. I thank you for your input and time taken and am grateful for your points and inputs as well as from others. That is what makes the Forum an awesome place for help
I have plenty of flying toys that run dumb LiPOs also and yes, monitoring the voltages is something I do in a religious like fashion. I also time the flights, keep a note of charge cycles, measure internal resistance, track the Ah put in while charging, measure landing temps, discharge full packs I didn’t fly etc. Btw- all the things I mentioned I do routinely and at considerable time investment our DJI smart flight packs do for us.

I’m not suggesting cell voltages in flight are not a good way to look for issues. I am saying the the battery SOC is far more capable than any of us might be at gauging time to empty and remaining capacity and when operated within limitations is exceptionally accurate and reliable.

The inspire battery issues you mention are also frequently observed with phantom and Mavic packs also. They rarely present with a pack that is close to fresh off the charge though. It’s a kn easy problem to avoid it seems.