Welcome Inspire Pilots!
Join our free DJI Inspire community today!
Sign up

Battery Mod Info

Status
Not open for further replies.
That is the crux of trying to do the whole "parallel battery mod". You are limited by not only the stock battery management system, but you also run the real risk of putting the TB47/TB48 into charge mode based on potential differences between the batteries.
 
it finally worked properly with the two 3s batteries in the series I only have 3 minutes extra time, even though the batteries were still 39%. The dji program struck because the main battery had dropped to 20% .... Conclusion? batteries with the same discharge rate as the dji battery, and also with a larger capacity. My battery has a capacity of 4139mha ...
Unfortunately nothing is obvious on this territory since the system is dealing with two circuits of different properties and only one circuit is controllable/monitored by firmware/software. God only knows how a given combination will perform during flight, which and when one branch is discharging faster, voltage drops and current starts to flow between main and aux battery rather than toward motors. Only a new pack of specific LiPo cells in 6S configuration and connected with proprietary DJI "smart" circuitry may deliver. And this is what some of us is trying to determine, with mixed rate of success by far ...
 
Unfortunately nothing is obvious on this territory since the system is dealing with two circuits of different properties and only one circuit is controllable/monitored by firmware/software. God only knows how a given combination will perform during flight, which and when one branch is discharging faster, voltage drops and current starts to flow between main and aux battery rather than toward motors. Only a new pack of specific LiPo cells in 6S configuration and connected with proprietary DJI "smart" circuitry may deliver. And this is what some of us is trying to determine, with mixed rate of success by far ...
I hope a solution will be found soon
 
I hope a solution will be found soon
Doubt it. Even if so, no solution will be simple, cheap and easy. The main reason is the availability of raw LiPo cells, which are fabricated and sold in China only without exception. To source them, purchase and receive exactly what you've ordered is the task I gave up already ...
 
this battery is not the same? Has anyone tried to change the lid?
 

Attachments

  • Screenshot_2021-04-21-23-14-48-562_com.android.chrome.jpg
    Screenshot_2021-04-21-23-14-48-562_com.android.chrome.jpg
    509.5 KB · Views: 8
Only the battery pack is identical, the electronics are different. Therefore the whole top, contact board and temp probe must be removed and replaced with original from Inspire TB47 battery.
I would like to add, that every single battery is also self-calibrated and set to ideal parameter management for that set of cells inside that battery. So if you switch the cells out they are no longer calibrated and can run into major failure conditions and the pack would then lock out, being no longer useable at all.
 
  • Like
Reactions: Akis1981
I would like to add, that every single battery is also self-calibrated and set to ideal parameter management for that set of cells inside that battery. So if you switch the cells out they are no longer calibrated and can run into major failure conditions and the pack would then lock out, being no longer useable at all.
Now, you're frightening me ... I would imagine that individual cells are selected and matched for capacity/IR, but calibrating the assembled pack is a new thing I would never think about ... I was under impression that the balancing procedure takes care of calibration issue ...
 
Now, you're frightening me ... I would imagine that individual cells are selected and matched for capacity/IR, but calibrating the assembled pack is a new thing I would never think about ... I was under impression that the balancing procedure takes care of calibration issue ...
Actually the smart board (or BMS as they are referred to) is a VERY complex and complete piece of engineering that has the ability to calibrate for battery characteristics and for the ability to determine, based on programmed parameters, when and if a cell will fail, BEFORE it fails. It tracks dozens of parameters, knows the offsets between batteries for balancing, knows tons of stuff about how the cells react and even trends tha data over the lifecycle of the pack to keep it useable long after most packs would be considered "damaged". Ie, the DJI packs "age" very well.

There is a lot of stuff that happens during manufacturing and there is some stuff you can do "after the fact", but one does not equal the other in regards to being able to "repurpose" packs (ie, load new cells in an already existing pack). On assembly they tell the pack to self-calibrate for the cells installed. That way it knows what is installed and can know exactly how they need to operate based upon formulation.

I am still working with a number of smart boards and getting a handle on the technical side of things (I have all the specs for all the chips used, their protocols and I am mapping out the I2C protocol as well as the chip-level configurations and capabilities), so it may be anywhere between a week or a month before I am able to get to a point where I can say EXACTLY what can be repurposed and what cannot.

But, if anyone thinks it is just a battery meter and balance system that keeps the voltages the same, they are way off... It is a very complex system that affords for variances within cells as well as tracking discharge and charge characteristics to keep everything in "balance". Quite frankly, almost all of the DJI drone battery systems use just as a complex system. Most other higher-end drones do as well. Then you have failures, like Parrot Anafi that won't show the voltages of their batteries and constantly run them into the "damage zone" on the low end. They "fixed" it in their USA offerings by increasing the size of the cell packs and changing the parameters, but they did not backward engineer it into their previous offerings thus if you aren't aware you are killing your batteries by not bringing the drone back before voltages get too low. This is all well documented and known on the other forums.

So, to sum up, if you just put a new set of cells into an existing DJI battery, it is not calibrated for those cells. It could be WAY OFF, and you won't know until you get a power fail out of the blue. Also if the proper formulation is not configured on the smart board it will not be monitoring the configured cells properly. If it is a DJI pack to DJI pack transplant it is pretty safe to say they are the same, but ALL of the alternative cell transplants are running a risk without knowing EXACTLY the characteristics of the cells being used (which is why you have to go to the manufacturer of the cells like I did to get all of that information to program the system appropriately).

Also, for those that are still looking at the addition of extra external packs, the issue there is you change the entire characteristics of how the smart board monitors power usage and end up messing with its ability to allow the pack to provide proper voltage and current. Unless certain electrical things are taken care of (that require changes to the TB47/TB48 packs themselves) you will always run the risk of tripping the charge mode and having the aircraft fall from the sky.
 
Actually the smart board (or BMS as they are referred to) is a VERY complex and complete piece of engineering that has the ability to calibrate for battery characteristics and for the ability to determine, based on programmed parameters, when and if a cell will fail, BEFORE it fails. It tracks dozens of parameters, knows the offsets between batteries for balancing, knows tons of stuff about how the cells react and even trends tha data over the lifecycle of the pack to keep it useable long after most packs would be considered "damaged". Ie, the DJI packs "age" very well.

There is a lot of stuff that happens during manufacturing and there is some stuff you can do "after the fact", but one does not equal the other in regards to being able to "repurpose" packs (ie, load new cells in an already existing pack). On assembly they tell the pack to self-calibrate for the cells installed. That way it knows what is installed and can know exactly how they need to operate based upon formulation.

I am still working with a number of smart boards and getting a handle on the technical side of things (I have all the specs for all the chips used, their protocols and I am mapping out the I2C protocol as well as the chip-level configurations and capabilities), so it may be anywhere between a week or a month before I am able to get to a point where I can say EXACTLY what can be repurposed and what cannot.

But, if anyone thinks it is just a battery meter and balance system that keeps the voltages the same, they are way off... It is a very complex system that affords for variances within cells as well as tracking discharge and charge characteristics to keep everything in "balance". Quite frankly, almost all of the DJI drone battery systems use just as a complex system. Most other higher-end drones do as well. Then you have failures, like Parrot Anafi that won't show the voltages of their batteries and constantly run them into the "damage zone" on the low end. They "fixed" it in their USA offerings by increasing the size of the cell packs and changing the parameters, but they did not backward engineer it into their previous offerings thus if you aren't aware you are killing your batteries by not bringing the drone back before voltages get too low. This is all well documented and known on the other forums.

So, to sum up, if you just put a new set of cells into an existing DJI battery, it is not calibrated for those cells. It could be WAY OFF, and you won't know until you get a power fail out of the blue. Also if the proper formulation is not configured on the smart board it will not be monitoring the configured cells properly. If it is a DJI pack to DJI pack transplant it is pretty safe to say they are the same, but ALL of the alternative cell transplants are running a risk without knowing EXACTLY the characteristics of the cells being used (which is why you have to go to the manufacturer of the cells like I did to get all of that information to program the system appropriately).

Also, for those that are still looking at the addition of extra external packs, the issue there is you change the entire characteristics of how the smart board monitors power usage and end up messing with its ability to allow the pack to provide proper voltage and current. Unless certain electrical things are taken care of (that require changes to the TB47/TB48 packs themselves) you will always run the risk of tripping the charge mode and having the aircraft fall from the sky.
I always knew that DJI's intelligent battery management is of a very complex and advanced nature, years ahead competition. I also know that transplanting exotic cells into the system NOT prepared to accept them is a risky move, a recipe for a disaster. So, until somebody finally write the software and access fine and proper tuning of every and each battery with new replacement pack, I will remain skeptical ...

As for now my two batteries with fresh TB48S pack transplant seems to deliver the juice. The charges count and total capacity are still displaying old battery values from before transplant, but cells voltage readings during flight remains on solid green, throttle regardless. Inspire 1 is not a long runner, but if I can pull 15 minutes with X5R camera from TB48, I'm OK ...
 
I always knew that DJI's intelligent battery management is of a very complex and advanced nature, years ahead competition. I also know that transplanting exotic cells into the system NOT prepared to accept them is a risky move, a recipe for a disaster. So, until somebody finally write the software and access fine and proper tuning of every and each battery with new replacement pack, I will remain skeptical ...

As for now my two batteries with fresh TB48S pack transplant seems to deliver the juice. The charges count and total capacity are still displaying old battery values from before transplant, but cells voltage readings during flight remains on solid green, throttle regardless. Inspire 1 is not a long runner, but if I can pull 15 minutes with X5R camera from TB48, I'm OK ...
Yeah, believe it or not, but the majority of the tech actually came from laptop batteries. That is the major consumer of the BMS board technology besides drones. The drones just found the perfect application for real-time high-current consumption usage.
 
In my opinion, the DJI batteries are way over engineered, hence the issues we currently have when DJI decides, we're not going to make these anymore, that and no 3rd parties willing to pay the high startup costs to make batteries. A lot of RC and other drones use the approach of having the BMS in the charger, not the battery, i.e. the batteries are dumb but the charger is smart. While you don't get the detailed cell by cell voltage and cycle telemetry DJI provides, it's not really necessary. Every time you buy a DJI battery you're basically buying a smart charger with it, adding weight.

With proper design, as long as the discharge rating of the cells exceeds the draw, temperature is of little concern while in use, you will never over tax the batteries with the exception of over discharge. A temperature probe is really only beneficial while charging. I use the touch the cell approach and look to see if the input Amps are declining or a cell is taking too long to balance.

The batteries in my big electric helis, never get hot.

Using any of these (smart or dumb) batteries safely, requires active participation while in use. Complacency is the enemy.

As far as cycles, the rated cycles are an average. Some get more some less. I'm not concerned with cycles, I'm concerned with internal resistance. A full cycle is a complete charge and discharge, anything less is a fraction of a cycle. Some batteries should be replaced with very few cycles. And yes, at some point with age, the internal resistance increases, the run time fades and the batteries will need to be replaced, all of which can be determined while charging before use. Most of the information I get from a smart charger, I can't get from DJI batteries, we have to trust the BMS. Most people that have issues, try running batteries that should have been replaced long ago, continuing to use them like they were new. I'm not faulting them, given the high cost, and few replacement options it's not surprising. High cost, "being smart", forced obsolescence, basically encourages bad behavior, like over use, attempted personal rebuilding, cheap jury-rigging, etc., all negating any perceived safety benefits.

Voltage and charge remaining can still be provided through telemetry with so called dumb batteries.

Batteries shouldn't be left alone while charging anyway. Having the BMS separate from the battery reduces weight and cost. Personally I prefer to use a balance charger/tester of my choosing on the balance leads. Having started early in this hobby with LiPo's, I was forced to learn and make some expensive mistakes.

How many Tesla's crashed because with autopilot, drivers abdicated their responsibility of being the driver, or automation in a cockpit leading to softened pilot skills, etc.

https://batteryuniversity.com/ is a great resource regarding all sorts of cell and battery chemistries.

All that said, DJI batteries have caught fire, just like all the others using different approaches, are they statistically any safer? I don't know, but to me just more expensive and something else to break.
 
  • Like
Reactions: The Editor
In my opinion, the DJI batteries are way over engineered, hence the issues we currently have when DJI decides, we're not going to make these anymore, that and no 3rd parties willing to pay the high startup costs to make batteries. A lot of RC and other drones use the approach of having the BMS in the charger, not the battery, i.e. the batteries are dumb but the charger is smart. While you don't get the detailed cell by cell voltage and cycle telemetry DJI provides, it's not really necessary. Every time you buy a DJI battery you're basically buying a smart charger with it, adding weight.

With proper design, as long as the discharge rating of the cells exceeds the draw, temperature is of little concern while in use, you will never over tax the batteries with the exception of over discharge. A temperature probe is really only beneficial while charging. I use the touch the cell approach and look to see if the input Amps are declining or a cell is taking too long to balance.

The batteries in my big electric helis, never get hot.

Using any of these (smart or dumb) batteries safely, requires active participation while in use. Complacency is the enemy.

As far as cycles, the rated cycles are an average. Some get more some less. I'm not concerned with cycles, I'm concerned with internal resistance. A full cycle is a complete charge and discharge, anything less is a fraction of a cycle. Some batteries should be replaced with very few cycles. And yes, at some point with age, the internal resistance increases, the run time fades and the batteries will need to be replaced, all of which can be determined while charging before use. Most of the information I get from a smart charger, I can't get from DJI batteries, we have to trust the BMS. Most people that have issues, try running batteries that should have been replaced long ago, continuing to use them like they were new. I'm not faulting them, given the high cost, and few replacement options it's not surprising. High cost, "being smart", forced obsolescence, basically encourages bad behavior, like over use, attempted personal rebuilding, cheap jury-rigging, etc., all negating any perceived safety benefits.

Voltage and charge remaining can still be provided through telemetry with so called dumb batteries.

Batteries shouldn't be left alone while charging anyway. Having the BMS separate from the battery reduces weight and cost. Personally I prefer to use a balance charger/tester of my choosing on the balance leads. Having started early in this hobby with LiPo's, I was forced to learn and make some expensive mistakes.

How many Tesla's crashed because with autopilot, drivers abdicated their responsibility of being the driver, or automation in a cockpit leading to softened pilot skills, etc.

https://batteryuniversity.com/ is a great resource regarding all sorts of cell and battery chemistries.

All that said, DJI batteries have caught fire, just like all the others using different approaches, are they statistically any safer? I don't know, but to me just more expensive and something else to break.
But, quite frankly, to put it bluntly, NONE of this matters.

We have what we have with this product. Great breakdown of your opinion, but it does not change what we have been talking about, what we have to deal with and solutions that need to be found to allow this platform to work long-term into the future.

There is only two ways of doing just this:

1) The protocol of the aircraft to battery is ascertained and disseminated and equivalent new tech is put into place to "emulate" a batteries comms to pacify the built in firmware. This method allows any "dumb" battery to be used and would be only useful if it tracked the voltage during flight. Any "smart" features could be picked-and chosen as to which would be supported depending on needs/technology implemented.

2) The BMS is figured out in detail and is reprogrammed from scratch with new cell information on a new buildup. This would allow for custom battery construction and the ability to change the capacities. All the rest of the "smart" features would be retained.
 
Well, from what I recall, one of CruzRoy's contacts in China had already cracked the BMS to make the 12,000MaH battery he tested, resetting the cycle count and setting the higher capacity. Either that, or they may have been a former DJI employee with the keys to unlock the BMS.

Personally, an easier approach may be to just salvage the air frame, replace the flight controller with a PixHawk or other, and add a custom 3 axis gimbal and camera, and the radio of your choice like a Frsky Horus x10. Once done batteries would no longer be an issue. Basically any DIY quad kit, less the air frame, just use the Inspire 1 as a donor.
 
Actually, those larger batteries are not calibrated. I already have the inside info on those and they are no longer available. Basically, quickly hacked tech, not a fully-engineered solution.

As far as using the Inspire 1 airframe, again, that is NOT what is being discussed. This is an Inspire 1 forum and a discussion about battery options. You are not discussing that.
 
In my opinion, the DJI batteries are way over engineered, hence the issues we currently have ...
From an avid hobbyist's point of view you're certainly right about over engineered DJI's batteries. But you don't see the whole picture in global scale, what actually happened during last 6-7 years since DJI surprised millions with the Phantom concept ... A drone for the masses! A flying rock in hands of an army of weekend warriors, happy ignorants and beer drinkers with absolutely no idea about Lipo battery things ... Do you see these masses worrying about IR, BMS or voltages drop? That's why "smart" board is taking care of these things, warning when something is not right before total failure. Well, most of the time ...
 
Last edited:
  • Like
Reactions: Brian Fist
Reece28Battery on Facebook as of March is no longer doing Inspire 1 battery mods. I should have jumped with Vincent Ng when I had the chance.

Heli Engadin Inspire 1 products seems to be the only commercially viable battery option at the moment.

If anyone is interested, there's been some recent discussion here on reprogramming the TB47 BMS
Battery firmware · Issue #17 · o-gs/dji-firmware-tools

I may try the re-celling option when I need it using a TB48S.

Yes there will be risk with any of these options but the only safe option is to do nothing and accept the Inspire 1 becoming an expensive paper weight or shelf queen.
 
Status
Not open for further replies.

Members online

No members online now.

Forum statistics

Threads
22,277
Messages
210,655
Members
34,333
Latest member
marve84