Interesting conclusion on disconnects from app and rc

[mnoutdoors]

ok, after 2 days of pulling my hair out this is what I’ve found.
I have been experiencing disconnects from the RC, while flying, or even just powered up sitting and what I have found is, it is very repeatable almost set your watch to, as a matter of fact I used a stop watch at 4 minutes 13 seconds the RC will disconnect for 1 second then reconnect but this is enough to initiate a RTH
Because that what I have it set at. It does not matter on the distance it can be 5 feet or 1500 feet away
It indicates (weak signal readust antenna)reboots and is good for 4 minutes 13 seconds then repeats I had tried different versions of the app (IOS) using on iPad, and no difference I have reinstalled software on the aircraft and no change. I’m still thinking it’s a software issue,but will not rule out a hardware issue yet either.

Has anybody ever had this? I have refresh so maybe it’s time. I have about 30 hours of flying on the bird
And over 100 flight miles, never crashed

 

[Anthony Viscomi]

Very interesting that its always exactly at 4 mins 13 seconds. Waiting to hear the outcome of this.

 

[Robert Mitchell]

Following.

 

[Keule]

@mnoutdoors Interesting find.
Being in IT for a long time, i'm always looking for bugs, flaws or other glitches in software code.

So, t seems there could be a buffer overrun somewhere in the code (either on the RC or AC).
Why? Covert your finding into seconds will give you 254 seconds. [(4 min * 60) + 14 seconds)]
Thinking in Bytes, then there is 1 Byte = 8 bit = 256 decimal. (2^8)

Given the fact that your measurements are precise, whenever an internal counter (for whatever reason) or buffer, an array reaches 254 or even 255 (more likely), the firmware initiates a 'mini' reboot of a module, leading to the disconnects.
Counters, variables are reset, and 'redo from start' for another period of 254 seconds.

Note: the above is only an assumption! So don't take it as granted! But i am on it.

What firmware did you tried?

 

[Anthony Viscomi]

@mnoutdoors Interesting find.
Being in IT for a long time, i'm always looking for bugs, flaws or other glitches in software code.

So, t seems there could be a buffer overrun somewhere in the code (either on the RC or AC).
Why? Covert your finding into seconds will give you 254 seconds. [(4 min * 60) + 14 seconds)]
Thinking in Bytes, then there is 1 Byte = 8 bit = 256 decimal. (2^8)

Given the fact that your measurements are precise, whenever an internal counter (for whatever reason) or buffer, an array reaches 254 or even 255 (more likely), the firmware initiates a 'mini' reboot of a module, leading to the disconnects.
Counters, variables are reset, and 'redo from start' for another period of 254 seconds.

Note: the above is only an assumption! So don't take it as granted! But i am on it.

What firmware did you tried?

Well freaking done!

 

[mnoutdoors]

@mnoutdoors Interesting find.
Being in IT for a long time, i'm always looking for bugs, flaws or other glitches in software code.

So, t seems there could be a buffer overrun somewhere in the code (either on the RC or AC).
Why? Covert your finding into seconds will give you 254 seconds. [(4 min * 60) + 14 seconds)]
Thinking in Bytes, then there is 1 Byte = 8 bit = 256 decimal. (2^8)

Given the fact that your measurements are precise, whenever an internal counter (for whatever reason) or buffer, an array reaches 254 or even 255 (more likely), the firmware initiates a 'mini' reboot of a module, leading to the disconnects.
Counters, variables are reset, and 'redo from start' for another period of 254 seconds.

Note: the above is only an assumption! So don't take it as granted! But i am on it.

What firmware did you tried?

Thanks for your input, I’m a engineer but not electrical, mechanical

Here is where I’m at,I went and bought a new mavic today and it operates fine on my app 4.1.20
So I paired the new controller to the AC,and same thing, 4minutes 13 seconds it has a 1 second hiccup
RC disconnects,just brief enough to initiate my RTH which I override.

I have installed new Firmware on the AC .01.04.0100 ( wish I wouldn’t have ) the RC upgraded also to
.01.04.0100 neither helped anything but the old RC will run the new Mavic I got today just fine so that leads me
To believe it’s either the firmware in the AC or something in the hardware in the AC

The new Mavic I bought today has .01.03.0800 FW on AC, .01.03.0550 FW on the RC I will not upgrade if I can help it.

I do have refresh so I guess that’s about where I’m at. And that is why I purchased the new Mavic Today
Can’t be without, I’m retired, and fly almost everyday.

When I get the damaged one replaced, I will dedicate it to a battery mod/ antenna flier also have thought about using it to place my game cameras set up on three carbon fiber arrow blanks tripod and hook about 1.5 pounds total weight

If anyone has any other ideas or clues they are very welcomed, Happy New Year!

 

[Dan]

Remove the SIM card and try your test again

 

[HighlandsVideo]

Remove the SIM card and try your test again

Where is there a SIM card on the AC and how would one remove it?

 

[mnoutdoors]

Took it out and that did not cure it either, reset all the cache also video and logs also.

But Thanks for the idea !

 

[mnoutdoors]

Where is there a SIM card on the AC and how would one remove it?

It on the side of the AC under the larger of the two flap doors it is where the RC/WIFI switch and pairing button are. I use the tip of a pen to push it slightly and then it unlatches and pops out ( and can fly and hide quite nicely)

 

[Simmo]

It on the side of the AC under the larger of the two flap doors it is where the RC/WIFI switch and pairing button are. I use the tip of a pen to push it slightly and then it unlatches and pops out ( and can fly and hide quite nicely)

Thats an SD card...?

 

[mnoutdoors]

Yes sir, a micro Sd,the factory one is only 16GB,but I run 64GB now

 

[Simmo]

Ahh ok, the above were calling it a SIM card.

Dan said: ↑
Remove the SIM card and try your test again

Where is there a SIM card on the AC and how would one remove it?

#8 HighlandsVideo, 31 minutes ago

 

[Qoncussion]

Where is there a SIM card on the AC and how would one remove it?

There are no SIM cards on a Mavic. Therefore, one would not remove it. I have a feeling that Dan meant microSD card in his earlier post.

 

[Logger]

That’s odd. A 1 sec signal loss should not trigger RTH. Should require 3 sec or more.
Have you tried flying from a different location. It occurred to me that you could be getting signal interfence at a regular interval from another source. You could rule that out by moving the RC and Mavic elsewhere and see if it still occurs.

 

[mnoutdoors]

I thought about that, and I never had a problem in the past, nor do I have that same problem with the new MP I purchased today. So is guess I discounted that. I live right next to a 500 KVA power line and I have flown right over the top of that before without any problem.

I did wonder if someone created a spark gap generator,close by,it would interfere with RF,but once again does not effect the other MP at all

 

[Dan]

Ahh ok, the above were calling it a SIM card.

Dan said: ↑
Remove the SIM card and try your test again

Where is there a SIM card on the AC and how would one remove it?

#8 HighlandsVideo, 31 minutes ago

Correct meant the SD Card, I'm wondering if the timing your mentioning is going through the process and checking the SD Card and these if where the failure is happening...

 

[HighlandsVideo]

There are no SIM cards on a Mavic.

Quite right and I just wanted to point out that the use of correct terminology is ultra critical in these communications.

 

[Keule]

I did run a test with one of my Mavics, i could not reproduce the dropout as the OP described.

The craft was on 01.04.0100, the RC had the same version.
I used GO 4 app 4.1.22 on a Fire HD 7 tablet.
I also tried Litchi Version 1.18.0 as control app.

I ran the test twice and for each app (GO4 / Litchi) 10 minutes long.
Craft was only powered ON, props removed, sitting on a table and an external ventilator cooling the Mavic from the front.

No dropouts, signal losses or similar. I watched the onscreen GO4 status as well the RC's display.

I also had wireless networks, a mobile phone and another tablet in the test area.

As @Logger already mentioned, could that issue eventually triggered by external means?

 

[mnoutdoors]

But only on one of my MP, I have two as of yesterday,not the other,it’s going in to repair the RC seems fine but they want it sent in