Accident with Octocopter

[soler]

Hello,

ESc setings are totally wrong.
Battery should be NI and cutoff voltage low.

Regarding the choice for Ni or Lipo cutoff, how does the ESC know when to cut off as it only gets the full battery voltage, not the voltage from each cell. Surely it would need to know the cell count of the battery in order to work this out or does it just go on the starting voltage that is applied and assume that it is a fully charged battery?

 

[gtranquilla]

soler... that is very good question!

ESCs have to be connected to a charged battery pack when they are configured and will determine the total voltage from that. How it then determines the number of battery cells, I have no idea, unless we tell it what each cell nominal voltage is by means of selecting the cell type for LVC. We need to hear from someone on MRF who really understands the ESC programming......

Regarding the choice for Ni or Lipo cutoff, how does the ESC know when to cut off as it only gets the full battery voltage, not the voltage from each cell. Surely it would need to know the cell count of the battery in order to work this out or does it just go on the starting voltage that is applied and assume that it is a fully charged battery?

 

[kloner]

esc's self sense based on the voltage they see powering up, uses the total voltage to determine when to start shutting off, in a plane it's fine, glide a while, by then the voltage is recovered enough for one last hoorah at a landing and there down, multis drop an arm and fall like a rock when these start shutting down,,, that's the beeps when they boot usually. using nimh lets the voltage get lower than when it is set to lipo,,,,

when you use an esc in a multi, you are essentially shutting off any battery protection doing this and either relying on the flight controller to do it,, also one main voltage number, not cell volts only they try landing or at least warning ya, where an esc controlling it just throws it all down at the ground

 

[soler]

I don't like the idea they this is configured at start up and guesses the cell count based on the input voltage, if you were to start with a 50% charged battery the safety device would fail to save your batteries anyway.

I can imagine that there is a lot of people that are not aware of this setting or have the required skills to check or change this. This should be an option that is turned off by default and can be enabled by advanced users if they want this.

As the SimonK firmware is commonly used does anybody know of the settings it uses for this?

 

[kloner]

it's all the same stuff, all works great. simonK has voltage monitoring dissabled

 

[gtranquilla]

soler,
Since ESCs aren't connected to the flight battery balance connector they have to connect to the main leads.
You are probably correct though.... a novice could connect a partially depleted flight battery and ESC would determine a wrong #S value for the battery and make several wrong decisions after that leading to many issues including crashes etc.

 

[STEVE5]

Hallo to all,
just some observation of similar (HW) setup and few data from my first flight with 14,5 Kg setup.

Thanks for any comments.

 

[fgomezd]

gtranquilla:

I finally received the diagnosis of the electronic engineer to see if the thead of the crash were the ESC´s, he founded that the ESC´s are working well, so I wrote to DJI, and after 6 months triying to receive support,they required the FLAYING DATA LOG information, after a week their diagnose was that batteries couldn´t deliver the energy and had a sudden power cut off, but the thing is that now I do not trust the answer of DJI fot the situation on my A2 so I decided to buy new ESC´s and diferent Flight Controller, what do you suggest, the most important is reliability, what do you think?????

 

[PeterLester]

DJI's assessment might be 100% correct! We have nothing to go on re: your subsequent decision and or lack of faith in DJI and their hardware.


Sent from my iPad using Tapatalk

 

[PeterLester]

I agree! FC system reliability needs to be at least, slightly better than power supply reliability! But where does that leave you? Send us more info so we can assess root cause of failure.


Sent from my iPad using Tapatalk

 

[ACP]

As mentioned ESC determine the cell count by the initial power on voltage. It needs to know what type of cells you are using to do this.
For example set to Ni, it uses a formula 1.2V = 1 Cel 2.4V = 2 cels and so on.
set to Li 3.7V = 1 cel............22.2V = 6 cells... and so on.
Ok cells charged are different voltages. A Fresh charged LiPo could be up to 4.16V so
IF volts > 3.5 and < 4.1 then cel count = 1

So a fully charger 6 cel lipo could be 25.2V
if volt >19.9 <25.1 then cel count = 6
a poor 7 cell could also be 25.2V
If volts > 25.1 and <28.7 then Cel count = 7

In this case the ESC WRONGLY presumes a 7 Cel pack. So once the motors power up the voltage drops to 21.7V and if the cut-off is set to 3.1V per cell then the ESC will shut off the motors.
But in reality the 6cel which is FULLY charged is good.
The numbers are exaggerated a little to make it easier to explain but I hope you see the point.
Powering up you system with say 50% SOC packs the ESC can get confused and have the opposite effect.

The reason we tell it we are using Ni cells is that there voltages drop more than Lipos, so the ESC used a different (IF) test on them making the cut-off voltage lower (or the voltage range grater).
This in turn keeps the ESC fooled in thinking there is still plenty of fuel left so it won't Cut-Off the motors.

The ESC would think you are using a 20 Cell Ni which has a cut-off of 18V, well below a 6cel Lipo.

Regards
Anthony

 

[fgomezd]

Dear Srs: I found what was the fail so my octocopter fall down, after almost 6 months the fail was in my XT60 connectors, so many times the sparked I didnt realice they will be spoiled and had a called GLIC, so I have brougth anti-spark conectors and now I have started to fly again with my KRAKEN with the A2 DJI FC.

I think is not a FC problem...

Thanks

Francisco

 

[Bartman]

glad to hear you got it worked out Francisco.

 

[fgomezd]

Peter:
I found the problem, it wasn't in the A2, or in the esc's it was the XT60 connectors that with a LIPO 6S sparked many times, until when the motors need energy the power was insulated from batteries and had the accident, so I brougth anti-spark conectors and back to fly my KRAKEN.

 

[EastCoastJib]

I realize this is an older thread but there is a serious amount of brain power between those that will get the notification... i hope. I finally had a fairly decent crash and managed to totally trash a gimbal, gh4, landing gear and some props. Otherwise no damage to the core at all.
SkyJib 8 with AMP 500kv motors and AMP 15x5 carbon props.
Maytech 65amp Optos
DJI WKM
OSD Mk2
Futaba 8's with 6208's core/gimbal

So I've gone through my head.. and the log and neither are making much sense about the final why..

I accidentally grabbed manual mode for a split second and it sounded like a screaming woman falling from about
60 ft agl. 2.5 G impact into the middle of the road. Came straight down.

https://files.acrobat.com/a/preview/d2b19986-f508-4238-bf5c-b9e63d233a20

37 mb file...

The log shows the throttle freaking out but I'm not quite grasping the reason why it behaved the way it did. The voltage never reached
1st level at 21.4v, weren't flying heavy. I'd love some insight if you guys would care to share it. Some of you guys have been deciphering these DJI logs a little longer than I have and may see it right away.

I'm thinking I'll go ahead and put new motors on it and ESC's since they go hand in hand. Props for sure also. Wasn't planning on changing FC as we've had great results with the WKM for all 78 flights on this rig. But now my confidence is bothered and I'm looking around to see what my options are.

Thanks in advance guys...

Bum