Anafi in the Ohio river

[AnafiPilot1]

I should mention my Drone isn't on the latest firmware yet(i haven't had time to update). So i think its not an issue with the new firmware.

"Auto landing"

The drone is back home Apart from the broken arm and a a few scuffed props everything seems to be working. Connection, the motors "click" on start up and the gimbal calibrates but i've not tried firing up the motors yet!

I might have a look at gluing the arm but a replacement looks more sensible(i'll post photos of the break tomorrow).

So what point was the autoland - after the forward and down flight near to the tree? It is not clear exactly when the autoland starts.

 

[Jagerbomb52]

I might have a look at gluing the arm but a replacement looks more sensible(i'll post photos of the break tomorrow).

Someone had recommended this stuff for a strong bond. Never heard of it till then.

InstaMorph - Reusable Moldable Plastic

InstaMorph is a reusable lightweight polyester thermoplastic that acts like clay when warm, but when it cools, becomes a strong plastic. Our moldable plastic can be sculpted with your hands, tools, or pressed around an object to make a mold. It can be painted, machined, carved or attached to...

www.instamorph.com

 

[ROBANAFI]

So what point was the autoland - after the forward and down flight near to the tree? It is not clear exactly when the autoland starts.

I'd say around 35 seconds. i had control but couldn't cancel auto land and the rate of descent meant it i couldn't get back.

 

[BRC]

I'd say around 35 seconds. i had control but couldn't cancel auto land and the rate of descent meant it i couldn't get back.

it almost looks like it initiated a RTH around 35 seconds - the way it rotated heading back to your position, then right into auto landing?
Maybe I am miss reading it.....
Anyway I am very pleased that you were able to recover it

 

[ROBANAFI]

it almost looks like it initiated a RTH around 35 seconds - the way it rotated heading back to your position, then right into auto landing?
Maybe I am miss reading it.....
Anyway I am very pleased that you were able to recover it

i rotated the drone but couldn't make it back. Thanks for you're kind comment though "i'm made up to get it back" - translation = Made Up | Scouse Dictionary - Scouse Sayings and Words from Liverpool

 

[DiscoBeb2anafiuser]

So it is not the battery firmware but the drone firmware that is causing the error? Then why did Parrot have to upgrade the battery firmware and why did weird warnings come down with landings and drops with 20-80% battery?

 

[DiscoBeb2anafiuser]

I tried to charge the batteries i left in the car last night. They wouldn't charge. Too cold. Hmmmm. Not just too cold for usbc. Too cold for direct terminals. Hmmm Hmm. I couldn't charge with my Anbee fast charger direct either. The battery voltage at the terminals of both was... 8.6v the charge levels of the batteries were 1 led and the other 2 leds. Hmmmm. Now that is my data with no interpretation. Except i said too cold. That was an assumption. But yes, 10 minutes in the warmth and they started charging. But what is important to me is that Rarely, does anyone, put something in the path of a critical circuit. I say that because i was aerospace, and i am compelled to warn the hell out ot the astronauts, but i won't take the decision out of their hands, or control. Cars, and outboard motors are the same way. They will overtemp and run without oil but they wont normally cut off the "pilot in control". The pilot makes the decision whether to save the engine, or get 20 to 50 more seconds of power before it locks up. So, i dont have the "smart battery" circuit diagram. But, I would not expect to find any component that could fail or degrade to reduce the full voltage and current the two cells contained within .. from the outputs. Which really means: Direct Connection. Why. These toys are not. These toys can do damage as small as they are. Damage to humans. And damage to a human must be performed (decided) by a human. Robotics law. God's zero commandment. Anyway, I did take a smart battery apart once... its still here somewhere, it sat in a ditch of 6 inch water for 6 days before i found it...and the anafi it was attached to. Lol. Hit a powerline and landed in a ditch only 2 feet wide. What are the odds? Well the ditch was bigger but the water was only 2 feet wide in it. And 6 inches deep. MY POINT. Which I almost forgot by now, IS. IF the battery "smarts" and circuitry can control the output... then it must be considered. I realize that the screen still reports good battery... but in the instance of the decision of anafi to autoland.... did it sample it? Is it a frozen value? I know nothing. I reported my data in the beginning. That is what I had to add. And of course. Robots may not defend themselves as humans are allowed to... they can run hide or just take it. They can identify targets and present them to a human. Even Dozens at a time. But someone has to be responsible and accountable for the kill authorization. Each one. When it comes to robots... we are GOD. Im not making a robot god. Lol

 

[ROBANAFI]

Someone had recommended this stuff for a strong bond. Never heard of it till then.

InstaMorph - Reusable Moldable Plastic

InstaMorph is a reusable lightweight polyester thermoplastic that acts like clay when warm, but when it cools, becomes a strong plastic. Our moldable plastic can be sculpted with your hands, tools, or pressed around an object to make a mold. It can be painted, machined, carved or attached to...

www.instamorph.com

Ok here is the break in the arm.





It looks pretty clean so i'm thinking a bit of glue then apply that instamorph for a bit of added strength. The last photo shows a tab which i assume is to give the arms a bit more stability when folded out, lucky that this is still intact!

Quite a cool video i found showing how that instamorph can be used with drones.

 

[Gipsz Jakab]

Just came in from the back yard. First battery that came with the Anafi.
44°F outside right now. Not to bad of a day. This battery is the one that spent 3 months outside in a tree in the winter in Northern Ontario. The total life percentage of the battery as of today is 87%. It has been charged 68 times. From start to landing it flew 20 minutes and 19 seconds. It used 99% of the battery during this period. It gave a Auto landing message at 12% saying it will auto land in 3 minutes. At 1% the Anafi auto landed.
So this about the 3rd time I have conducted this test over the last 6 months. Never has it tried to auto land at a higher percentage and it has never just dropped and shut off. So either I have one of the better Anafis or there is another reason unknown to us. Some how I don't think we will ever know.

View attachment 5023

View attachment 5024


As you can see the battery has seen better days
Duct tape to the rescue
View attachment 5025

It can also be a firmware error because under certain conditions (temperature, charge, instantaneous load, etc.), it does not handle (misinterpret) the case properly and this leads to a false automatic landing.
Anyway, these batteries are made in completely different versions (smartbattery.version, smartbattery.serial, smartbattery.hw_version) this may also be the reason for the different behavior.
quote: "Parrot today solved this problem!
This version enhances the battery level management and fixes the data synchronization issue caused by the previous version. "
Or not?

 

[AnafiPilot1]

I'd say around 35 seconds. i had control but couldn't cancel auto land and the rate of descent meant it i couldn't get back.

Oh Ok - I always thought with critical landing you had no control at all - but looks like you had horizontal control just not vertical.

 

[AnafiPilot1]

Ok here is the break in the arm.

View attachment 5033
View attachment 5035
View attachment 5036

It looks pretty clean so i'm thinking a bit of glue then apply that instamorph for a bit of added strength. The last photo shows a tab which i assume is to give the arms a bit more stability when folded out, lucky that this is still intact!

Quite a cool video i found showing how that instamorph can be used with drones.

My fix would be to use some CA glue to get leg in held together. Then with an soldering iron melt in some paperclips in to plastic (on the 3 sides), crossing the break. Finally sand down the area and apply some epoxy.

The leg will need to be removed however to do it properly!

 

[Liger 1956]

Ok here is the break in the arm.

View attachment 5033
View attachment 5035
View attachment 5036

It looks pretty clean so i'm thinking a bit of glue then apply that instamorph for a bit of added strength. The last photo shows a tab which i assume is to give the arms a bit more stability when folded out, lucky that this is still intact!

Quite a cool video i found showing how that instamorph can be used with drones.

I have used Polymorph, similar to Instamorph, in the past and it is not good for making fine or thin pieces. Think of trying to mould boiling hot clay that sets very quickly!

I would go with either Cyano, if it will stick the plastic, or Epoxy. You could then reinforce the break by heating some staples and pushing them into the arm to span the repair. Just make sure that you cut the legs of the staples short enough that they do not penetrate the arm and damage the wires inside.

 

[WWTV]

I haven't read all of the comments, but It appears many have had issues with battery voltage and drones falling out of the sky. I had that happen a couple months ago on a replacement ANAFI. It had the battery it came with and it fell just at around 29%. Luckily it was within a 30 day period and I was able to return it for a refund explaining what happened to the seller.

Got another replacement shortly afterwards. I haven't flown it too much recently, but I only used my original ANAFI's batteries with no issue (I legitimately crashed my original/1st ANAFI).

Anyway, I subscribe to a YouTube channel. The guy does a lot of range testing somewhere in SE Asia. He also promotes/sells a controller that extends the range of the ANAFI among other drone products. He just put out a video claiming that Parrot has a design flaw in how the battery voltage vs. % of battery left reported.

First, the battery has two cells, so all voltages are per cell. The upshot is that the battery goes below 3.7v (per cell) on a fresh battery at around 43%. According to him, that's the critical voltage as far as battery health. When the battery gets to 3.0v, the drone shuts down. Not lands, shuts down, as there's not enough power to operate the drone. The motherboard needs 6.1v total.

So if we are flying our drones like a normal person where we land somewhere between 10-20% of battery left according to the app, then we are killing the battery. So at some point we will experience the ANAFI falling out of the sky when the app says we have plenty of battery.

He said that Parrot intentionally removed a cell vs. the Bebop series to save weight. But didn't factor that into the Freeflight app as far as battery health since a fresh battery will operate just fine. Somewhere between 20-30 cycles of normal flying will cause these crashes when you have 20-30% battery left.

Here is the link to his video. It's a bit dramatic (that's his style), but he demonstrates the voltage draining in real time in a before and after type of video. The after from replacing the LiPo cells in the battery with new ones.

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

A little theatrical performance but he gets his point across. Something we have suspected for some time now.

parrotpilots.com

Bottom line is that the drone has become untrustworthy after a small number of flights on a battery unless you strictly adhere to landing it with 43% or more battery left. Or just risk it and then keep buying new batteries.

I'm no expert in LiPo or drones in general. I've only had mine for a year. But this all seems legit. What do you all think?

 

[ROBANAFI]

I have used Polymorph, similar to Instamorph, in the past and it is not good for making fine or thin pieces. Think of trying to mould boiling hot clay that sets very quickly!

I would go with either Cyano, if it will stick the plastic, or Epoxy. You could then reinforce the break by heating some staples and pushing them into the arm to span the repair. Just make sure that you cut the legs of the staples short enough that they do not penetrate the arm and damage the wires inside.

Ok i've took the risk and and i'm trying a bit of yours and @AnafiPilot1 advice. Cyano+Epoxy+Tape(with long pins instead of "melted" staples). I know i should have removed the arm but i'm going to give it a go attached to the drone. ?

 

[Ana789]

[...] Somewhere between 20-30 cycles of normal flying will cause these crashes when you have 20-30% battery left.

[...]

I can assure you, my batteries all had more as 20-30 cycles. And I used them almost every time below 20% in FF6. So it is not that easy. The fall with 20% juice in the battery, eh, in FF6, was with about 90ish cycles on that battery.


Because if a manufacturer says "26 minutes flight time", I am flying minimum up to 18/20 minutes, even try some flights to get close to "26 minutes".

And if flying the Anafi longer as ~14minutes is damaging the batteries, I'd say there is something wrong with the advertised flight time! Which is a "guaranteed feature", a buying argument.

But as I thought after my "event", the intelligent batteries arent (by fw flaw) not intelligent enough, to tell the bird and the FF6, that the live cycles are coming to an end.
Which ist on Parrot!

The mentioned video is really interesting, since it is giving a good advise, how to change the batteries!
Maybe I should go and buy another Anafi, they are really cheap, now! And change the cells every 50 flights? Have to check if these cells are available, here!