DJI Mavic, Air and Mini Drones
Friendly, Helpful & Knowledgeable Community
Join Us Now

Tripling battery life span

Hi @DIY_Quad I was the one asking you to come over here instead of DJI forums, welcome. I'm glad you got sorted out and can start posting.

I too have some doubts on whether it will really triple the battery life time expectancy, by reducing the maximum charging stress. One thing I am guessing (Stress - Guessing) is that the life cycle limit is also likely limited or mostly affected by the thermal stress from discharge.

Nonetheless a cool mod. Can I know how does your mod works actually? Are you just detecting the battery voltage and applying a hard disconnect to the affected port? Or are you communicating to the charging IC and instructing a stop at your desired voltage? For the purpose of storage and actual flight for max performance, can we have some presets for like 3.9V and 4.4V per cell?
Hey, thanks for showing me this forum, now I can freely discuss custom mods.

I understand your doubt and there is nothing wrong with having a doubt, cuz that leads to experiments and better understanding of the subject matter.
I am convinced that reducing max voltage stress will increase number of discharge cycles, but it remains to be seen exactly how close it would get to the projected 3X increase in discharge cycles.
Even if it gets me 2X increase, I will take that tradeoff any day of the week.

Thermal stress is definitely another factor that affects number of discharge cycles.
Temperature and max voltage are 2 independent variables.
In this particular study, they are fixing temperature variable and only sweeping max voltage variable.
In other words, wear and tear by thermal stress is equally applied to all of their testing while sweeping max voltage.
Such that they can focus on effect of max voltage on discharge cycles only.

With that said, let's get to some technical details of the mod.
The battery has 10 pins.
Left 4 are ground pins, next 2 are communication bus and the last 4 are +ve terminal.
The charging hub uses pin #4 as a sensing pin to see whether or not a battery is plugged in to a given port.
It's normally 3.3V and when pulled low by inserting a battery, the MC (micro controller) in the hub knows there is a battery on that particular port.
There is a resistor between pin#4 and the hub-MC.
I disconnected that resistor and connected the pin side and hub-MC side to my own MC (my-MC).
So that my-MC can sense presence of a battery on a given port and be able to tell the hub-MC whether or not a battery is present on that port.
Basically, my-MC is intercepting signal on pin#4 and let the hub-MC know when to start/stop charging on each port.
When a battery is plugged in to a given port, my-MC lets the hub-MC know that there is a battery present and the hub-MC will automatically start charging that battery.
When that happens, my-MC reads charging voltage going into that battery.
There is a voltage divider that the hub-MC reads to determine voltage of the charger (current limited source) and my-MC is sensing that same voltage divider output to know charging voltage across the battery.
This lets me determine state of charge (SoC) for that battery pack.
I compare that voltage level to my preset level and let it continue charging if below my preset level and stop charging if preset level is reached by tricking the hub-MC to think that the battery has been pulled out from the port.
I do that for all 3 ports and any battery connected to the hub will be charged to a preset level and then stop.
I have programmed 3 preset levels/modes to choose from; ~50% (3.85V/cell), ~90% (4.25V/cell) and 100% (4.4V/cell).
With the 100% mode, I let the hub-MC completely take over sensing of pin#4 by simply passing whatever my-MC senses on pin#4 directly to the hub-MC.
In this mode, the hub functions exactly like it normally would without the mod.
The added LED shows currently selected preset charge level or mode.
For example, if currently selected mode is 50% charging, it blinks with 50% duty cycle.
For 90% mode, the LED blinks with 90% duty cycle and so on.

Pls feel free to ask questions!
 
Last edited:
Not at all, I never said that wasn't a trade for LiHV, what I said is that the 3x claim doesn't hold. With this battery type it could be 1.5x or 5x, but you said "tripling" and nothing you have shown backs this up for that particular battery chemistry. It's just clickbait.

This is what you wrote;
"Make a small difference maybe, but not even enough to bother."

And now you are saying "could be 1.5x or 5x"??
Even "1.5x or 5x" is not considered small.
What happened to your "not even enough to bother" claim?
When you don't even know if it's going to be "1.5x or 5x", how can you claim it's "not even enough to bother"?
If you had said something along the lines of "it remains to be seen" or "it could be more or it could be less than 3x" that would have made more sense.
But instease, you had made a definitive statement that it is "not even enough to bother" without providing any valid reason or data to back your claim.

Now, I don't know exactly how many times this mod will increase discharge cycles myself cuz I am still in the process of finding out for myself.
That's why I am careful with what I say that it is "supposedly like 3X longer" or "about 3X" and provided the source.
This is different from making a definitive claim of 3X increase.
I understand now you are trying cling on to the word "tripling" but if you read my post correctly, there is no definitive claim of exactly 3X increase in discharge cycle.
That is very different from your definitive claim.
I will quote what you wrote here once again in case you have forgot already.
"Make a small difference maybe, but not even enough to bother."
 
But instease, you had made a definitive statement that it is "not even enough to bother" without providing any valid reason or data to back your claim.
Simple, I've been following and using all DJI aircraft since they introduced their smart batteries and reports of dead batteries due to number of cycles are basically nonexistent.
So even having it 5x better doesn't matter when you never reach "1x" before you either upgrade your aircraft or the batteries die of other causes.

The measures that DJI has put in place are simply enough already.
 
  • Haha
Reactions: Ex Coelis
Simple, I've been following and using all DJI aircraft since they introduced their smart batteries and reports of dead batteries due to number of cycles are basically nonexistent.
So even having it 5x better doesn't matter when you never reach "1x" before you either upgrade your aircraft or the batteries die of other causes.

The measures that DJI has put in place are simply enough already.
Haha, now you are saying DJI batteries have "nonexistent" or zero discharge cycles??
But guess what? DJI drones do fly with their batteries.
Meaning they do have non-zero discharge cycles for the vast majority of the cases.
In fact, significantly higher than zero discharge cycles to make it commercially viable...LOL
Well, if this is the level of discussion you are seeking.... I will let you claim victory...LOL
 
  • Like
Reactions: Ex Coelis
Reread the post. Reports of batteries failing due to cycle aging are almost nonexistent, not cycles...
 
Reread the post. Reports of batteries failing due to cycle aging are almost nonexistent, not cycles...
Too slow, I quoted your post before you had the chance to change it.

"Simple, I've been following and using all DJI aircraft since they introduced their smart batteries and reports of dead batteries due to number of cycles are basically nonexistent."
 
I didn't edit my post, and in English it's read as

"Simple, I've been following and using all DJI aircraft since they introduced their smart batteries and reports of dead batteries due to number of cycles are basically nonexistent."
 
Reread the post. Reports of batteries failing due to cycle aging are almost nonexistent, not cycles...
I see that you are trying to shift the discussion to how crappy DJI batteries are.
Instead of staying with your unfound claim that increase in battery life is "not even enough to bother."
 
I didn't edit my post, and in English it's read as

"Simple, I've been following and using all DJI aircraft since they introduced their smart batteries and reports of dead batteries due to number of cycles are basically nonexistent."
and what does it read in English?
"Make a small difference maybe, but not even enough to bother."
 
Simple, I've been following and using all DJI aircraft since they introduced their smart batteries and reports of dead batteries due to number of cycles are basically nonexistent.
So even having it 5x better doesn't matter when you never reach "1x" before you either upgrade your aircraft or the batteries die of other causes.

The measures that DJI has put in place are simply enough already.
Oh, my bad, I admit that I misread your post.
What you wrote is even more hilarious than I originally thought.
You are saying DJI batteries don't die of discharge cycles.
Wow, that's new....and funny at the same time....LOL
That would surely render my mod completely useless cuz DJI batteries never wear out from discharge cycles.
Now I understand where you are coming from.
If this is your level of understanding of Li-based batteries....well I'm just speechless now.
 
Last edited:
That data is as you mentioned but ignored for different chemistries than what DJI is currently using, assuming it's also valid for them in the same proportions is as good as nothing.

Also, the base cycle count quoted on BU already exceeds what most people will ever do with their drone batteries... With all of my aircraft and those of colleagues and friends I don't think I've ever seen a battery with more than 100 cycles.
Yet, this is what you wrote earlier;
Also, the base cycle count quoted on BU already exceeds what most people will ever do with their drone batteries... With all of my aircraft and those of colleagues and friends I don't think I've ever seen a battery with more than 100 cycles.

What now??? You have never seen a battery with more than 100 cycles??
But you just said, "reports of dead batteries due to number of cycles are basically nonexistent."
Aren't you tired of contradicting yourself??
If you want to continue this, at least learn to honor your own words.
 
Last edited:
@DIY_Quad and @Kilrah this whole thread is turning into a slanging match ,and becoming rather boring, gentlemen lets agree to disagree ,the life span of our batteries is open to so many variables ,that its inevitable that users are going to get many different cycle counts from their batteries ,lipos do have a finite lifespan ,based on how they have been used, stored ,charged ,etc
 
Correct - It's the high current draw of a flight that wears batteries.
The current demand (load) presented to the battery by the drone, while it might be reduced to some extent by flying style, is a non modifiable parameter. Factors that we have some control over- including storage voltage, temperature, depth of discharge, and maximum terminal voltage are all well demonstrated as impacting service life and performance.

I charge close to when I expect to fly, try and land with 20% remaining, keep the batteries in a cool area for storage and avoid flying in high ambient temps.

My batteries last longer than the devices they are powering in many cases (not just for drones).

The fact most don’t need the solution the OP has engineered doesn’t mean it won’t provide some benefit, it certainly adds convenience and useful features.
 
  • Like
Reactions: DIY_Quad
@DIY_Quad and @Kilrah this whole thread is turning into a slanging match ,and becoming rather boring, gentlemen lets agree to disagree ,the life span of our batteries is open to so many variables ,that its inevitable that users are going to get many different cycle counts from their batteries ,lipos do have a finite lifespan ,based on how they have been used, stored ,charged ,etc
Yes, I agree that's the general understanding.
Sorry for the rant, I couldn't help it after reading some unfounded statement.
 
  • Like
Reactions: old man mavic
What are your observations, if any, with respect to cell balance after charging to a final cell voltage below spec?

I’m wondering how the balancing constant voltage stage is managed by the BMS where the cut off is potentially implemented externally rather than by the BMS SOC algorithm.
 
this is no way to encourage a good technical discourse. To say there is no evidence of charging cycles causing failure is probably inaccurate. I have heard of a few batteries puffing up during charge, especially on Mavic Air 1.
 
  • Like
Reactions: DIY_Quad
What are your observations, if any, with respect to cell balance after charging to a final cell voltage below spec?

I’m wondering how the balancing constant voltage stage is managed by the BMS where the cut off is potentially implemented externally rather than by the BMS SOC algorithm.
balance charging is done by the battery internal BMS. All chargers that I know just supply a max Voltage and let the battery do its own balancing
 
balance charging is done by the battery internal BMS. All chargers that I know just supply a max Voltage and let the battery do its own balancing
Have you looked at the cell voltages after your limited charging?

What voltage are you supplying for a 90% charge?

What you are calling the “charger” is simply a regulated DC supply.

Is the BMS actually completing the charge routine or is it shutting down before charge completion?

If it’s the later then you might find the BMS might not have done a great job of balancing the cells.
 
The current demand (load) presented to the battery by the drone, while it might be reduced to some extent by flying style, is a non modifiable parameter. Factors that we have some control over- including storage voltage, temperature, depth of discharge, and maximum terminal voltage are all well demonstrated as impacting service life and performance.

I charge close to when I expect to fly, try and land with 20% remaining, keep the batteries in a cool area for storage and avoid flying in high ambient temps.

My batteries last longer than the devices they are powering in many cases (not just for drones).

The fact most don’t need the solution the OP has engineered doesn’t mean it won’t provide some benefit, it certainly adds convenience and useful features.
You wrote well.
That is also my understanding.

"The fact most don’t need the solution the OP has engineered doesn’t mean it won’t provide some benefit, it certainly adds convenience and useful features."

As for this part, I would think DJI's first priority is to achieve long flight time while life span of batteries would be a lower priority relatively speaking (within reason of course).
For example, DJI adverizes for x many minutes of flight time for their drones, but hardly for x many discharge cycles.

There wouldn't be much incentive for DJI to give people option to charge under 100% and make batteries last longer.
Cuz then people would be 1) not happy/impressed with reduced flight time and 2) buy less batteries.
 
Have you looked at the cell voltages after your limited charging?

What voltage are you supplying for a 90% charge?

What you are calling the “charger” is simply a regulated DC supply.

Is the BMS actually completing the charge routine or is it shutting down before charge completion?

If it’s the later then you might find the BMS might not have done a great job of balancing the cells.
Yes, the quoted V/cell are what I am reading after stopping at 50% or 90%.
Voltage applied before cutting off for 90% charge is about 13.03V.
Once that level is reached and current source is disconnected, per cell voltage drops to around 4.25V.
"Charger" is the brick.
I believe it is a current-limited source with maximum voltage of 13.2V.
That's usual for charging Li cells with constant current.
The batteries are disconnected from the source once predefined level is reached.
I am hoping BMS or more specifically the balancer is active throughout charging, not necessarily kicks in towards the very end of charging.
For that, I am constantly keeping an eye on balance of the cells until I am confident that balancing is happening without having to reach 100%.
You ask good questions.
 
Lycus Tech Mavic Air 3 Case

DJI Drone Deals

New Threads

Members online

Forum statistics

Threads
130,585
Messages
1,554,092
Members
159,585
Latest member
maniac2000