Battery Mod Info

[Phatzo]

ok. I did some more pcb porn - sorry Damon had to take it...
I made a microscopic video from top left to bottom right and an edge longer side to side. youtube video is up in few min. will post in this post.

as promissed:

they are a bit long so don't be shy and fast forward if to slow or boring. please keep in mind it fell from about 6-8 meters. so any damage that could occur from that should be taken into account please.

 

[Adelberg aerial imaging]

Cool

 

[damoncooper]

Again, that does make me feel better about the connection board but the images Phatzo posted still leave me questioning the reliability and quality of the smart PCB and possible bonding issues/solder joints.

Ahhh. Reminds me of the good ol DJI Phantom 2 ESC cold solder point failure days.

Waxing nostalgic over here....

 

[Phatzo]

posted two posts above

 

[damoncooper]

Bloody hell...that board is a disaster.

EDIT: I really dont need to see any further images. Despite having no issues with my new Inspire I am not flying again without backup power or re-soldering those pins. If these images are representative of the boards that put our birds life in their hands, no thanks. That is crap quality.

Food for thought:

Dropgate – Why Are So Many DJI Inspires and Phantoms Dropping Out of the Sky?

 

[MrDRC]

Food for thought:

Dropgate – Why Are So Many DJI Inspires and Phantoms Dropping Out of the Sky?

Its probably safe to assume there are a lot more AC in the air these days. On a percentage basis, my guess is the overall failure rate is still somewhat low versus how many flights are made each day. It does seem power problems own a higher percentage of failures then ESC or CSC though. From looking at the smart battery PCB's they are bad quality. I've seen better quality "dumb" hobby boards at radio shack for $5.

Having backup power is a good idea. There are too many potential points of failure in DJI's smart battery system. When you add cold weather to the equation it gets even worse.

 

[Phatzo]

Its probably safe to assume there are a lot more AC in the air these days. On a percentage basis, my guess is the overall failure rate is still somewhat low versus how many flights are made each day. It does seem power problems own a higher percentage of failures then ESC or CSC though. From looking at the smart battery PCB's they are bad quality. I've seen better quality "dumb" hobby boards at radio shack for $5.

Having backup power is a good idea. There are too many potential points of failure in DJI's smart battery system. When you add cold weather to the equation it gets even worse.

is this also true with FW 1.4.00.10+??
i do not have the budget yet to get TP and charger from over the pond. small country = small market and the need for a strong contact base, specially in an ex sozializm country... democray does not really work here yet...
i am strongly working on the task though
are the videos i posted any good for use and/or info you guys were after? if anyone is in the need of only pics, let me know and i will make some more.
before i forget i only microscoped only one side. any one cares for the other?
let me know and i will be happy to help
cheers
m



Sent from my iPhone using Tapatalk

 

[damoncooper]

is this also true with FW 1.4.00.10+??
i do not have the budget yet to get TP and charger from over the pond. small country = small market and the need for a strong contact base, specially in an ex sozializm country... democray does not really work here yet...
i am strongly working on the task though
are the videos i posted any good for use and/or info you guys were after? if anyone is in the need of only pics, let me know and i will make some more.
before i forget i only microscoped only one side. any one cares for the other?
let me know and i will be happy to help
cheers
m



Sent from my iPhone using Tapatalk

Not sure if FW version is firmly implicated in failures at this point but there is at least one case of "falling from the sky" that is due to an ESC error. Hauntingly, there were DJI Phantom 2's falling from the sky in early 2014 due to poor quality soldering on ESC's as well.

Now we have evidence of poor quality soldering on the battery PCB.

 

[damoncooper]

Flytrex Live 3G installed with the Neil Watts X5/R aux battery mount on the Inspire Pro Black Edition.

 

[Phatzo]

Flytrex Live 3G installed with the Neil Watts X5/R aux battery mount on the Inspire Pro Black Edition.View attachment 6585

i have to get at least 2 sets of 3500TP the charger and the aux battery holder from Neil... ****
kind of scared to fly now or at least with a not so confident feeling after what you guys observed...
thanks for the answer Damon. will be super ultra cautious on flying and treat my first unit with corrosion block in case it decsides to fall into a lake i used to film. really hope this gets resolved soon. nice pics - what is the flytrax for ? is it like the trackimo or marco polo that was discussed in the beginning of times when I1 was bust released? is it any good ?
cheers


Sent from my iPhone using Tapatalk

 

[Phatzo]

Not sure if FW version is firmly implicated in failures at this point but there is at least one case of "falling from the sky" that is due to an ESC error. Hauntingly, there were DJI Phantom 2's falling from the sky in early 2014 due to poor quality soldering on ESC's as well.

Now we have evidence of poor quality soldering on the battery PCB.

what is the workaround apart from the battery mod? is there any?
worrying conditions coming up ...


Sent from my iPhone using Tapatalk

 

[unyks]

Flytrex Live 3G installed with the Neil Watts X5/R aux battery mount on the Inspire Pro Black Edition.View attachment 6585

Hi,

maybe the answer is somewhere in this thread and i overlooked it, do you have some footage how strong the battery insurance is visible in the air especially when you turn around the camera? How much of the flexibility is lost with the additional batteries when you mount them this way?

Thanks in advance..

Chris

 

[MrDRC]

are the videos i posted any good for use and/or info you guys were after?

The video's are good for identifying the P/N, chip set, board layout, etc but the resolution wasnt high enough to inspect the solder joints and traces. The photos you posted were good for looking at general board quality (which sucks). A full crop image from a macro lens would really show the detail. Regardless, there is enough evidence in the photos to indicate its suspect and battery redundancy is a wise move.

is this also true with FW 1.4.00.10+??Sent from my iPhone using Tapatalk

Im on current FW so I dont know. I havent had any probs. I dont feel like its a much of a FW problem as it is a hardware/battery problem across several layers of DJI tech. FW is probably part of it but from what we've seen hardware is probably a big contributing factor.

 

[Phatzo]

The video's are good for identifying the P/N, chip set, board layout, etc but the resolution wasnt high enough to inspect the solder joints and traces. The photos you posted were good for looking at general board quality (which sucks). A full crop image from a macro lens would really show the detail. Regardless, there is enough evidence in the photos to indicate its suspect and battery redundancy is a wise move.

the video microscope we have is from 1994 I should have mentioned it lol... nevertheless I shall do some more pics with the newer video microscope and post closer pics
I haven't had any problem yet apart from DJI braking an arm and sending my shop a 120 euro invoice and sending it back (shop payed) without resolving my issue which I clearly did not do (one antenna cable was not connected and also wrongly threaded (it went through the frame instead where it should have - outside like the other cables). after having to send it back 3 times I finally got the cable problem solved but upon receiving it I noticed that one leg (the part where the antenna is under the motor) is broken. my shop got me a spare leg which I have to replace now since I did not want to send it back again and wait almost 2 months, - AGAIN -
apart from that my first units flies nice and good. I am planning to treat it with corrosion block to have a piece of mind while over water shots.
this reminds me to share this - sorry to jack the thread, just a little link- of an alabama company startup waterproofing ANY drone. seems they have done the step I am planning to in my country. I do not know what they use and how it works but smells like liquipel or similar. anyways, here is the link: Tech One Plus - Splash Right In - for the ones flying over water and want to protect their investment or just simlply fly in rain or snowfall LOL

Im on current FW so I dont know. I havent had any probs. I dont feel like its a much of a FW problem as it is a hardware/battery problem across several layers of DJI tech. FW is probably part of it but from what we've seen hardware is probably a big contributing factor.

I really hope the problem gets resolved and DJI is taking off their blindfolds to tackle the QC issue. keeping in mind that service in germany hinted to assume I did something kind of pissed me off since I had or at least I thought I had a good relationship with them bearing in mind I speak and write german fluently (mother tongue next to english and hungarian) but I was proven wrong an went through my shop to solve this. which thankfully is a closed case now, just waiting to be replaced once I get the time or need to do so. I will stick to the 1.4.00.10+ and not upgrade until 1.7 is out and confirmed that this FW is worth putting up due to the enhancements and bugs removed.

 

[Mad_angler1]

Iv been looking into these packs heavily the last few days the pictures posted on here confirmed my initial thoughts, a huge thankyou for posting them fella [emoji3], also managed to get my hands on a dead one as well.

DJI have an interesting setup and power looks to be fed from the cells to its output via 2 pairs of FET's, the load is shared across a single pair but both pairs have to be active to supply power, this setup means the same FET's can be used to control charge current to the main battery or prevent power reaching the collective connector instead being pushed to each cell via the balance port.

It's rather a neat setup but as your introducing controls between the main battery and the craft there is room for a number of things to cause the FET's to shut down and power to stop.

When it comes to pack shut downs what we still don't know is if it's FW cutting power due to the LVC, some other reason or if it's FET failure or overload, the first thing we need to know if the pack will turn on after the shut down happens as that will pretty much confirm if it's an blown FET issue or not, but then if just one of a pair fails the pack could turn on but fail again under load , one another possibility is as pack voltage sags current draw will go through the roof, this could exceed the maximum rating of the FET's and they would tolerate burst current for a period but then shutdown as they exceed their thermal limits.

The truth is we really don't know and all these reasons are just speculation, there could be a huge number of things causing a shut down and every incident may be different, pack age, craft hours ect.

When you have a digital switch between the battery and motors there always will be the possibility of failure but that's no more than you already have with the ESC, these have their own LVC normally and again consist of FET's controlling power, there is no reason to think the FET's in the battery are any less reliable than in the ESC's

One very simple mod though could be done on a pack to bypass the FET's meaning regardless of what they are told to do power would still make it to the craft, we already know thanks to Damon's tests that if the battery does shut down it will still fly if provided with power, I need to do a few tests but a bypass can be easily added, this would be done only after turning the pack on in the craft and it would not affect the voltage or capacity readings as this is still being taken as before.

Something like an XT60 exiting the battery with a bridged one you can connect to bypass the switching would work well.

This would prevent the power being cutoff from the craft but this will not prevent cell issues or possible ESC shutdown.


Sent from my iPhone using Tapatalk

 

[Phatzo]

Iv been looking into these packs heavily the last few days the pictures posted on here confirmed my initial thoughts, a huge thankyou for posting them fella [emoji3], also managed to get my hands on a dead one as well.

DJI have an interesting setup and power looks to be fed from the cells to its output via 2 pairs of FET's, the load is shared across a single pair but both pairs have to be active to supply power, this setup means the same FET's can be used to control charge current to the main battery or prevent power reaching the collective connector instead being pushed to each cell via the balance port.

It's rather a neat setup but as your introducing controls between the main battery and the craft there is room for a number of things to cause the FET's to shut down and power to stop.

When it comes to pack shut downs what we still don't know is if it's FW cutting power due to the LVC, some other reason or if it's FET failure or overload, the first thing we need to know if the pack will turn on after the shut down happens as that will pretty much confirm if it's an blown FET issue or not, but then if just one of a pair fails the pack could turn on but fail again under load , one another possibility is as pack voltage sags current draw will go through the roof, this could exceed the maximum rating of the FET's and they would tolerate burst current for a period but then shutdown as they exceed their thermal limits.

The truth is we really don't know and all these reasons are just speculation, there could be a huge number of things causing a shut down and every incident may be different, pack age, craft hours ect.

When you have a digital switch between the battery and motors there always will be the possibility of failure but that's no more than you already have with the ESC, these have their own LVC normally and again consist of FET's controlling power, there is no reason to think the FET's in the battery are any less reliable than in the ESC's

One very simple mod though could be done on a pack to bypass the FET's meaning regardless of what they are told to do power would still make it to the craft, we already know thanks to Damon's tests that if the battery does shut down it will still fly if provided with power, I need to do a few tests but a bypass can be easily added, this would be done only after turning the pack on in the craft and it would not affect the voltage or capacity readings as this is still being taken as before.

Something like an XT60 exiting the battery with a bridged one you can connect to bypass the switching would work well.

This would prevent the power being cutoff from the craft but this will not prevent cell issues or possible ESC shutdown.


Sent from my iPhone using Tapatalk

I am happy to help. Damon did a great job at getting all the test done and sharing all the knowledge with us. - thanks man. -
I personally like to help, specially when it is a community like you guys and if I can be of any further assistance let me know. I could probably have it x-rayed if needed for further investigation. please do keep in mind that it did crash and bent so the lines and connections may be broken due to the fall out of approx 8 meters onto the battery like I mentioned in an earlier post. - please keep that in mind. but I can and plan to do the same (not drop it) but make pictures from a healthy PCB in hope it does not look so whack... lol
nevertheless if the x-ray is needed let me know and I check if I can still have it made. in case you need more closeups - that is possible. just let me know
you are very welcome. glad I am in this forum guys.
cheers.

 

[damoncooper]

I agree - we don't know why a pack might stop powering the aircraft - sag, solder or switch failure, AC FW, battery FW, etc.

I think doing a local bypass is an interesting idea but it would only potentially help protect against FW or possibly a PCB failure. Cell failure voltage sags and other types failures can only be protected against with a redundant power source (or preferably two).

For the effort to do such a local mod, I think you might as well add full protection against any failure and just add the external batteries.

My 2c

 

[Raphael Dahan]

I agree - we don't know why a pack might stop powering the aircraft - sag, solder or switch failure, AC FW, battery FW, etc.

I think doing a local bypass is an interesting idea but it would only potentially help protect against FW or possibly a PCB failure. Cell failure voltage sags and other types failures can only be protected against with a redundant power source (or preferably two).

For the effort to do such a local mod, I think you might as well add full protection against any failure and just add the external batteries.

My 2c

Damon, i think that for you and for all the pro in this thread, it will be interesting to remunufecture the PCB, by coping one to one and make proffetional soldering and contacts. I know it won't be easy project , but at least will show somethings to DJI how the things sould be done.

 

[Raphael Dahan]

Iv been looking into these packs heavily the last few days the pictures posted on here confirmed my initial thoughts, a huge thankyou for posting them fella [emoji3], also managed to get my hands on a dead one as well.

DJI have an interesting setup and power looks to be fed from the cells to its output via 2 pairs of FET's, the load is shared across a single pair but both pairs have to be active to supply power, this setup means the same FET's can be used to control charge current to the main battery or prevent power reaching the collective connector instead being pushed to each cell via the balance port.

It's rather a neat setup but as your introducing controls between the main battery and the craft there is room for a number of things to cause the FET's to shut down and power to stop.

When it comes to pack shut downs what we still don't know is if it's FW cutting power due to the LVC, some other reason or if it's FET failure or overload, the first thing we need to know if the pack will turn on after the shut down happens as that will pretty much confirm if it's an blown FET issue or not, but then if just one of a pair fails the pack could turn on but fail again under load , one another possibility is as pack voltage sags current draw will go through the roof, this could exceed the maximum rating of the FET's and they would tolerate burst current for a period but then shutdown as they exceed their thermal limits.

The truth is we really don't know and all these reasons are just speculation, there could be a huge number of things causing a shut down and every incident may be different, pack age, craft hours ect.

When you have a digital switch between the battery and motors there always will be the possibility of failure but that's no more than you already have with the ESC, these have their own LVC normally and again consist of FET's controlling power, there is no reason to think the FET's in the battery are any less reliable than in the ESC's

One very simple mod though could be done on a pack to bypass the FET's meaning regardless of what they are told to do power would still make it to the craft, we already know thanks to Damon's tests that if the battery does shut down it will still fly if provided with power, I need to do a few tests but a bypass can be easily added, this would be done only after turning the pack on in the craft and it would not affect the voltage or capacity readings as this is still being taken as before.

Something like an XT60 exiting the battery with a bridged one you can connect to bypass the switching would work well.

This would prevent the power being cutoff from the craft but this will not prevent cell issues or possible ESC shutdown.


Sent from my iPhone using Tapatalk

Sorry but if you can explain what is:
FET, LVC
Thanks

Raf

 

[Phatzo]

Damon, i think that for you and for all the pro in this thread, it will be interesting to remunufecture the PCB, by coping one to one and make proffetional soldering and contacts. I know it won't be easy project , but at least will show somethings to DJI how the things sould be done.

that will be very difficult because the program for communication is unknown... but to rearchitect - rebuild /plan the pcb would surley be a step forward. if R/D at dji would think about considering and give benefits to the ones that planned it better than dji itself would show gratitude i think. not to mention the credit that should be mentioned for the ones that ha e put in money and energy to improve the product.


Sent from my iPhone using Tapatalk