Solaris8x86 views on the Parrot Anafi crash's(5/Nov)

[Gipsz Jakab]

1. For used Anafi battery. There are two methods. First, is to drain it once just likes the video. If it falls. You know what to do. But you must charge it back to nominal voltage (1-2 leds) immediatley after the test or it will degrade significantly within days if remains totally empty. The second method is you get it fully charged at 100% before you go sleep. On next morning. Check again its % percentage. If it has already dropped to 95% or lower. It can be considered to be retired. For a new lthium battery, in the RC airport. professional RC players usually would strictly follow the 3.7v per cell rule (11.1v for 3S) on each flight to maintain the lithium health cycle. They would also use a small paper note to note down each discharge cycle. When the no. of discharge has reached to an expected number (such as 100 times). The battery is retired. They retired the battery first before to let the crash happens. I also do this way on Bebop 2. That's why my Bebop 2 could last for 2000 times flight spreading across 4 generations of battery replacement cycle within 4 years. Due to the discharge is always being controlled to not less than 3.7v per cell on each flight. So the estimation of day of retirement can be quite precisely.

2. Only choice is offered. Nothing more. Which choice to take is not a concern. As battery is not a very profitable goods. That's why preferring customers to buy the Parrot stock battery first. (for battery only, You go bother Parrot, don't bother me). In case people also need the Mod. My super bundled offer can help. The plan for the returning customer has already been described.

In whichever case, for the first time of purchase. The amount you spend for 3 batteries is very same. Because you buy it or not will not gain too much profit so no one really cares. My practical offer is I let people to know the truth and reality. Which may change the destiny of your drone. This's a major contribution. Thus, I have already offered my part for all people. Don't request further unreasonably.

In short, need battery only. Just go Parrot to help yourself.

You did not answer the question, if Anafi does not calculate the remaining charge based on the cell voltage, then on what basis? Because if cell voltage were the basis for the calculation, an aging battery would reach 2% charge sooner where it turns off.
Maybe you measure the elapsed time?
This problem could be aided by a firmware that monitors the cell voltage. So this is a firmware error because it does not take into account the specifics of the battery.
The 80% crash is not a hardware failure in my opinion ...
In any case, the useful advice you gave is not to submerge below 43%, but the solution is a firmware that informs you about the charge based on the battery voltage and brings the drone home safely when the battery voltage requires it.

 

[Eagle]

Parrot is pathetic. First of all not having a repair facility is unacceptable for a such a big company that wants to be taken serious. Also customer support is tragic
If parrot 's batteries have an issue then i think is their responsibility to look into it carefully and not ignoring all those anafi crashes. My drone switched off mid air and dropped and they refused to replace it or even look at the cause.
My advice STAY away from parrot there are a lot better companies selling drones

 

[Skydreamer]

I feel like going to Henri Seydoux house and yelling :”Henriii,Henriiiii,WTF,....Henriiiiiii...!!! “
?

 

[Fmaxii]

I feel like going to Henri Seydoux house and yelling :”Henriii,Henriiiii,WTF,....Henriiiiiii...!!! “
?

He heard you !

3 questions à... Henri Seydoux (PARROT)

Nous avons rencontré Henri Seydoux, co-fondateur de Parrot, créateur de l'A.R.drone (hélicoptère radiocommandé depuis son iPhone) et du Zik, un casque haut de gamme conçu avec Philippe Starck.

france3-regions.francetvinfo.fr

 

[Gipsz Jakab]

No one is asking your question.

The voltage vs percentage is calcuated by a voltage table predefined in the drone. It exists on all generations of Parrot drones. Including on Disco, Bebop series and Anafi. Parrot's work. Not me.
Skyshit FPV Suite is created by Parrot SDK software. I'm not measuring anything. It does. Parrot's work. Not me.
Firmware is alway monitoring the total voltage and cell voltage even their diff voltage in between cells. It just doesn't help on anything when cell has physically reached to total empty level within a few seconds. No time to do anything before reaching to its cut-off voltage. Operation terminate -> crash.

The root cause of all these, is the damage to the lithium caused by an intended misleading of the percentage left on Freeflight. If you always land at 40% - 50% at 3.7v-3.8v. It will last 100 cycles or more. When you mistrust the gauge and always land on 0-30 % at 3.0v - 3.6v. The total lifetime of battery shorten to 20-30 cycles. When applying this bad practise on any Lithium Polymer batteries. The result is the similar. Not Anafi proprietary.

Assume Parrot is willing to correct this intended "off scale" percentrage issue and have it corrected to align the voltage per cell level of Bebop 2 and Disco to bring ltihium back to a normal health cycle. The Anafi would be a "10 minutes drone" at max range 2km roundtrp when restricting the landing percentage at 43% (I tried once days ago). Afterall, using 2 cells 8.7v is a flaw design and the flaw was intentionally done in the design because of to trim weight to make a lighter drone.

I can't reproduce a 80% crash scientifically based on fact data. May be the next Einstern to tell us more about it some day.

This is not possible. If you were to display the actual charge percentage based on the voltage, as you wrote, we would not see 23% at 3.00V, but 0-2%.
So the Parrot drone ignores the battery voltage and does not show the remaining charge based on it. It is precisely because of this "lie" that it shuts off and falls out of the sky at 30%.
My question remains: so what does the Anafi firmware count on?

 

[jaja6009]

Thank you for the interesting video. I have had 2 sudden drops on two different Anafis. The sound in your video of the cutoff of power is exactly what I heard in my living room. I was testing a battery which the day before had a forced low battery landing. I was testing it at about 8 feet and at around 20-30% battery the power cut off, the engines stopped and she fell.

But I also had an Anafi Thermal fall after 5 minutes of hovering. I can only estimate that the battery was 70-80%.

At one point I was pretty close to a Parrot fanboy, but after my nightmare getting a warranty replacement and how Parrot has treated several members of this forum, I have pretty much written off ever purchasing another Parrot Drone any time in the next few years. They will need a huge turnaround for me to come back.


I'm with Augustine and the others that have said that a good solution would be to let us know the voltage on Free Flight. This simple thing could help to save a lot of drones in the future.

Please correct me if any of tis is wrong.

I should never let a 100% battery discharge on its own (10 day self discharge and sleep)
I should use the battery after full discharge within 2 days.
But what about this 40% landing? Even with the 8 batteries I have I would never be able to map the large areas that I currently do. I minus well give up on it for doing anything other than training or toying around with it.

 

[Fmaxii]

Here is an other example of what explain Solaris.
Anafi cut off at 50%.
then we can see a red light on the battery, just before 1 green led.
In fact, it’s not totally a battery power off, because it powers on after the anifi cut off, but it shows the battery power loss, at 50%..

Stupid battery level reporting !

 

[Gipsz Jakab]

I say again. It's all Parrot's work. Not me.
if you have watched thru the entire video. The 3.0v is at 0-2% finally. But on the new cells. The damaged cells were at 3.0 at 23%.

I'm not the one who should be in position to explain how the firmware made of or how it is calcuated or how it is offset the measurement scale when the cells were damaged. It's all Parrot's work, they wrote it. So, you've asked to a wrong person. We can just evident the data output and outcome only. The outcome is when cell is damaged. Their calculation doesn't work.

Why apologize? I'm not blaming you! I think that's Parrot's fault, too.
I ask because you understand him and you were the one taking the measurements.
You stated that Anafi shows the remaining charge% based on the values in the battery voltage table.
I'm saying that's not the case. If this were the case, the new battery would reach 100% -2% in about 20 minutes, and the used battery would reach 100% -2% in 10-12 minutes. Since this is not the reality, I am sure it does not take into account the tension !!! Then there could be no 23% display at 3.00V!
It’s not your fault, but your theory that pointing by measuring voltage is flawed. It would be desirable to show (and trigger RTH) exactly the remaining charge based on the voltage. Although he would come home sooner, he would not fall! And from the short flight time, you know you have to buy (renew) a battery. This is firmware (deliberate?) bug!

 

[Gipsz Jakab]

Thank you for the interesting video. I have had 2 sudden drops on two different Anafis. The sound in your video of the cutoff of power is exactly what I heard in my living room. I was testing a battery which the day before had a forced low battery landing. I was testing it at about 8 feet and at around 20-30% battery the power cut off, the engines stopped and she fell.

But I also had an Anafi Thermal fall after 5 minutes of hovering. I can only estimate that the battery was 70-80%.

At one point I was pretty close to a Parrot fanboy, but after my nightmare getting a warranty replacement and how Parrot has treated several members of this forum, I have pretty much written off ever purchasing another Parrot Drone any time in the next few years. They will need a huge turnaround for me to come back.


I'm with Augustine and the others that have said that a good solution would be to let us know the voltage on Free Flight. This simple thing could help to save a lot of drones in the future.

Please correct me if any of tis is wrong.

I should never let a 100% battery discharge on its own (10 day self discharge and sleep)
I should use the battery after full discharge within 2 days.
But what about this 40% landing? Even with the 8 batteries I have I would never be able to map the large areas that I currently do. I minus well give up on it for doing anything other than training or toying around with it.

"I'm with Augustine and the others that have said that a good solution would be to let us know the voltage on Free Flight. This simple thing could help to save a lot of drones in the future."
This would also be a solution, but for this we need to understand the specifics and operation of batteries.
It would be much better for the firmware to calculate the remaining charge based on the actual voltage and, if necessary, bring it back to the home point before reaching the switch-off voltage.
This is not a devil thing, this is how all other drones work except Anafi.
Of course, it would be best if it showed the voltage of the cells in addition to the properly functioning firmware ...

 

[Jagerbomb52]

On my old X-Star this is how it works for low battery. Only problem with this action is with a damaged battery. (bad cell) when the cell drops lower then the others the drone lands on the spot which is not a good thing if over water.

Failsafe
The Failsafe function is designed to help your X-Star/X-Star Premium automatically return home or land onsite when necessary.

• Low Aircraft Battery
Failsafe will also be activated if one of the following low battery conditions are met.
A.Your aircraft’s battery constantly calculates the required battery level for the aircraft to return to the home point from its current location. When the battery level reaches the minimum level required for the aircraft to return to the home point, Failsafe will be activated and your aircraft will automatically Go Home. You can regain control of the aircraft during the Go Home process.
B.Anytime your aircraft’s remaining battery level reaches 25% (Low Battery Warning), Failsafe will be activated and your aircraft will automatically Go Home. If you choose to regain control during this process, the aircraft will automatically land onsite when the remaining battery reaches 15% (Critical Low Battery Warning). In this case, you may push the Ascent stick to stop the aircraft landing and command it to fly away from the current location for a safe landing in case of an emergency.

 

[Gipsz Jakab]

On my old X-Star this is how it works for low battery. Only problem with this action is with a damaged battery. (bad cell) when the cell drops lower then the others the drone lands on the spot which is not a good thing if over water.

Failsafe
The Failsafe function is designed to help your X-Star/X-Star Premium automatically return home or land onsite when necessary.

• Low Aircraft Battery
Failsafe will also be activated if one of the following low battery conditions are met.
A.Your aircraft’s battery constantly calculates the required battery level for the aircraft to return to the home point from its current location. When the battery level reaches the minimum level required for the aircraft to return to the home point, Failsafe will be activated and your aircraft will automatically Go Home. You can regain control of the aircraft during the Go Home process.
B.Anytime your aircraft’s remaining battery level reaches 25% (Low Battery Warning), Failsafe will be activated and your aircraft will automatically Go Home. If you choose to regain control during this process, the aircraft will automatically land onsite when the remaining battery reaches 15% (Critical Low Battery Warning). In this case, you may push the Ascent stick to stop the aircraft landing and command it to fly away from the current location for a safe landing in case of an emergency.

That's all true, that's right! But as Solaris8x86 says, there is a table that records the% charge for battery voltages. Then why doesn't Anafi use this spreadsheet?
In the firmware if it detects that the 3.00V (6.00V) voltage, then why it writes to 23% charge?
I think this is a firmware error!

 

[Gipsz Jakab]

Ja látom, Gipsz éhes. Ezért teljesen félreértette a videó feliratát. Lehet, hogy legközelebb a google fordítót kell használnom egy magyar felirat beágyazásához. A videót kezdettől fogva nem érti helyesen.

​

I understand exactly. For technical reasons I don't understand see my previous answer to agustine.

 

[zandoli]

Parrot should have had 0% to be ~3.2V, whichever cell is lower.
Why have the battery go down to 3.00V to the point where the motherboard does not have enough power?
Parrot should not cut it this close.
Percentages and failsafe actions should also be based on whichever cell is lower.
It appears that they are using total voltage to do calculations.
Their programming is definitely flawed.

 

[Ana789]

Parrot should have had 0% to be ~3.2V, whichever cell is lower.
Why have the battery go down to 3.00V to the point where the motherboard does not have enough power?
Parrot should not cut it this close.
Percentages and failsafe actions should also be based on whichever cell is lower.
It appears that they are using total voltage to do calculations.
Their programming is definitely flawed.

The "big problem" for Parrot was: They wanted a ~300g drone, which needed a 2S battery. But also they wanted 50% more flightime as the Spark and still a longer one, in marketing, as the Mavic Air had.
So my "educated" guess is, that it needed to be this close to the threshold!
Which would have been ok, if that would have been corrected in pendency with the degradation of the battery abilities.
Which, it seems, other drone manufacturer are be able to implement into the intelligence of the fw of the batts.

 

[Gipsz Jakab]

Zandoli and Ana789 understood me!
The problem is monitoring the firmware voltage.
The issue is whether the battery firmware update is still this error or only since (1.6.5)

 

[Fmaxii]

I agree with you. The problem is the battery monitoring which should monitor the voltage, but not this % which is a % of we don’t know what.

 

[AnafiPilot1]

That's all true, that's right! But as Solaris8x86 says, there is a table that records the% charge for battery voltages. Then why doesn't Anafi use this spreadsheet?
In the firmware if it detects that the 3.00V (6.00V) voltage, then why it writes to 23% charge?
I think this is a firmware error!

Lipo have a very flat discharge curve, so actual measured voltage is a poor indicator of charge level especially under load. Under load and depending on current drawn vs. maximum discharge C of the battery, the voltage tends to be quite flat (can be as low as 0.15v) between 85% to 20% of battery level.

This is why the Anafi 'smart' battery uses like most batteries of this type use the concept of counting charge in and charge out, this means it measures the charge flowing in to the battery, and then counts the charge flowing out of it - this is what is used to display a battery level.

The Anafi itself clearly handles low voltage or critical voltage related problems by initiating auto landing itself.

Which ever way you look at it this is a pretty big fail by Parrot - it is actually not very hard to use current load, voltage and stated charge level from the battery to calculate a truer more realistic actual useable charge remaining.

There is no excuse for this - it is a closed system the battery and drone are designed by the same company - they know the parameters of the drone, and the parameters of the battery - create a operating map of load + voltage + battery level = accurate flight time remaining. Simple!

 

[jaja6009]

"Obivously wrong tool and wrong expectation I think. You should use fixed wing. Thing likes this small is designed for portability. Not functionality. Not intended for industrial application. I think the teams who run their sister company are smarter than their parent company. Take a look."


While I really do want a fixed wing solution for my mapping projects, they will not work. I cannot fly over people here in the US. Quad copters enable me to stop a mission and either wait for people to clear or avoid them, then continue the mission. I will make some final decisions after Remote ID rolls out in the US as I would like to ensure I don't spend money on something that may not work out as planned.

Thanks for your video and insight into the "Anafi problem"!

 

[Gipsz Jakab]

The discussion of changing the calculation of landing doesn't change anything.

The cell would still die in 20-30 cycles that it shouldn't be. It's not changing the fundamental problem. Would you accept to spend $99.99 for 20 cycles or something evenif it now can land safety at 3.0v per cell? You guys just got ahead of yourself. It has nothing to do with the app, ever. What you see in the app is just an outcome. Not the root cause.

When the physical resource does never exist. App and programming stuff are unimportant. Nothing can be done from nothing. So updating the app or firmware won't change a thing to your health of your battery. As long as you continue to drain power lower than 43%. Your battery dies sonner and game over sooner.


What you need to care is just two things:

1. Go get new batteries immediately and try to always land at 43% if possible in the future. As long as you are with Anafi. Your primary objective is to maintain its battery health cycle. Parrot cannot change this condition physically thru the future firmware updates as long as cells being used on Anafi remains just two cells. It is impossible to change this physical condition. So firmware updates won't change the damage to the lithium polymer if you drain lower than 43%. Evenif they can change the offset scale issue and improve thd autolanding sooner to occur. Lithium Polymer remains being damaged constantly. So firmware update is not that important to the root cause and it won't change the fact and the outcome.
2. In order to maintain a good health cycle to the battery. The storage practice is also extremely very important. Stated on my other posts already in this thread. Not repeating.
3. Software has never played a fundamental role in human aviation history. The physic does. Since 1903 - before Vietname war. In that half century. There was no computer onboard any plane and those planes still did amazing jobs including could do dogfights. They did 100% rely on mechanical physic is the nature of aircraft. So never expect software or programming can fundamental change what physic does. When physic says No. Nothing can be done. When the fuel tank is empty. It falls is just normal. Thus, so by spending effort to irmprove the firmware still can't improve the dealth rate of your battery until you change the way you handle your battery. Keep that in mind. And just don't get ahead of yourself by thinking too much on the firmware. It won't help you out in either way when physic says a No.

"When the physical resource does never exist. The app and programming stuff are unimportant. Nothing can be done from nothing. So updating the app or firmware will not change a thing to your health of your battery. As long as you continue to drain power lower than 43%. Your battery dies sonner and game over sooner. "

Unfortunately, you are not right! If Parrot writes the firmware well, we would not notice the decrease in battery health not by the bird falling out of the sky, but by the failsafe RTH starting much sooner and coming home before something goes wrong ...
It is true that there is nothing we can do, but Parrot can.
Or we can do that we fly little (43%) and buy a new battery from you (or Parrot).
I agree that it is poorly designed, but could still be improved with the right firmware: It would display the real cell voltage and come home even sooner, thus saving battery life. If you had a choice: fewer flights or a sudden stop in the air, which one would you choose? I also The first ...

 

[jspg]

Hi Solaris, I don't mind to buy new battery from Parrot, but your video is great to inform how to better use the battery, I have a question: if a battery has physical failure (bad quality) or becomes worn out, does the "RTH" no more work?