- Posts: 14
Mid flight TX reboot
- syphear
-
Topic Author
- Offline
Less
More
15 May 2013 10:19 #10014
by syphear
Mid flight TX reboot was created by syphear
My Devo8 running Deviation 3 rebooted itself halfway through a FPV flight today resulting in my Bixler2 landing on its wing (no damage luckily)
The receiver I was using is a Walkera Devention RX701
My TX was set to 100MW power output and the FPV TX was a Fatshark 5.8GHz 250MW.
The TX just out of the blue decided to reboot and the receiver could not rebind. The TX looked and worked completely normally after the reboot aswell.
Here is my errors.txt file contents hopefully someone can make sense of it.
The receiver I was using is a Walkera Devention RX701
My TX was set to 100MW power output and the FPV TX was a Fatshark 5.8GHz 250MW.
The TX just out of the blue decided to reboot and the receiver could not rebind. The TX looked and worked completely normally after the reboot aswell.
Here is my errors.txt file contents hopefully someone can make sense of it.
[Hard fault handler00000000
R0 = 40003800
R1 = 00000000
R2 = 200030e1
R3 = 00000003
R12 = 00000050
LR [R14] (subroutine call return address) = 080093cd
PC [R15] (program counter) = 08020696
PSR = 6100002e
BFAR = e000ed38
CFSR = 00000000
HFSR = 00000000
DFSR = 00000000
AFSR = 00000000
SCB_SHCSR = 00000000
impure_ptr = 08021874
00000026
20000082
0000001aPlease Log in or Create an account to join the conversation.
- cmpang
-
- Offline
Less
More
- Posts: 296
15 May 2013 10:56 #10016
by cmpang
Replied by cmpang on topic Mid flight TX reboot
The safest way is to bind the RX701 with fixed ID so that after power recycling or reboot the RX can be rebinded.. Since you are flying fixed wing so there is at least some chance that can save the plane..
Please Log in or Create an account to join the conversation.
- syphear
-
- Offline
Less
More
- Posts: 14
15 May 2013 11:47 #10017
by syphear
Replied by syphear on topic Mid flight TX reboot
Good call, I will definitely start using fixed ID for future flights. Cheers
Please Log in or Create an account to join the conversation.
- PhracturedBlue
-
- Offline
Less
More
- Posts: 4402
15 May 2013 13:37 #10023
by PhracturedBlue
Replied by PhracturedBlue on topic Mid flight TX reboot
Note that deviation will rarely be able to rebind after a reboot. The problem is that the channel selection happens uniquely after each reboot and if any conditions change, a different set of channels will be chosen. The solutions that would prevent this all have significant drawbacks.
Please Log in or Create an account to join the conversation.
- syphear
-
- Offline
Less
More
- Posts: 14
15 May 2013 13:57 #10024
by syphear
Replied by syphear on topic Mid flight TX reboot
So after a TX reboot even using fixed ID the chances of a successful re-bind
are slim to none?
are slim to none?
Please Log in or Create an account to join the conversation.
- vlad_vy
-
- Offline
Less
More
- Posts: 3333
15 May 2013 14:00 #10025
by vlad_vy
Replied by vlad_vy on topic Mid flight TX reboot
Ladybird with Fixed ID after switch off and switch on Tx rebind 10 times from 10.
Please Log in or Create an account to join the conversation.
- PhracturedBlue
-
- Offline
Less
More
- Posts: 4402
15 May 2013 14:04 #10026
by PhracturedBlue
Replied by PhracturedBlue on topic Mid flight TX reboot
It depends on the environment.
If there aren't any other 2.4GHz radios in the area, it will reliably rebind. In a noisy environment, it won't work very reliably. the longer it has been since the last bind, the less likely it is.
If there aren't any other 2.4GHz radios in the area, it will reliably rebind. In a noisy environment, it won't work very reliably. the longer it has been since the last bind, the less likely it is.
Please Log in or Create an account to join the conversation.
- syphear
-
- Offline
Less
More
- Posts: 14
15 May 2013 14:23 #10027
by syphear
Replied by syphear on topic Mid flight TX reboot
That is good to know, I fly in a fairly remote area so not much chance of stray 2.4GHz RF.
Does the Deviation bind procedure always start by looking at the same frequency, to see if its clear to begin with and move up or down from there? Or is the first part of the spectrum it looks in completely random every time you turn on the radio?
Does the Deviation bind procedure always start by looking at the same frequency, to see if its clear to begin with and move up or down from there? Or is the first part of the spectrum it looks in completely random every time you turn on the radio?
Please Log in or Create an account to join the conversation.
- PhracturedBlue
-
- Offline
Less
More
- Posts: 4402
15 May 2013 14:28 #10028
by PhracturedBlue
Replied by PhracturedBlue on topic Mid flight TX reboot
The code works by finding the 3 least noisy channels. It always searches in the same order. Since in a low-noise env, all channels can be virtually the same, any spurious noise can cause it to choose a different solution. We should, perhaps find a more robust solution that will always start with the same channels (maybe based on the fixed-id) and choose different ones only if those are too noisy. I'm not sure how to identify 'too noisy' though, so am not sure how to implement something like this.
Please Log in or Create an account to join the conversation.
- syphear
-
- Offline
Less
More
- Posts: 14
15 May 2013 14:50 #10031
by syphear
Replied by syphear on topic Mid flight TX reboot
Anything that could help an emergency re-bind would be effort well spent. The idea of starting with the same channel based on a fixed ID sounds like a good idea. Even if it is skipped over due to being too noisy at least it has a chance of working in a low noise environment.
Please Log in or Create an account to join the conversation.
- RRacer
-
- Offline
Less
More
- Posts: 43
04 Jun 2013 08:46 #10548
by RRacer
Replied by RRacer on topic Mid flight TX reboot
Yesterday I too had an in flight tx reboot.
I'm flying a Master CP with a Deviation 3.0.0 Devo 10.
Motor rpm quickly went down, but not as if you just cut power.
Luckily I was just a few feet off the ground so no damage.
When the tx had booted up again I had no control over the helicopter.
Did a rebind and everything was well again.
I really like Deviation, but this worries me a bit.
I use the Devo 10 for helicopters, planes and quads so a solution for a (very) quick rebind in the event of a reboot would make me feel a lot better.
I'm flying a Master CP with a Deviation 3.0.0 Devo 10.
Motor rpm quickly went down, but not as if you just cut power.
Luckily I was just a few feet off the ground so no damage.
When the tx had booted up again I had no control over the helicopter.
Did a rebind and everything was well again.
I really like Deviation, but this worries me a bit.
I use the Devo 10 for helicopters, planes and quads so a solution for a (very) quick rebind in the event of a reboot would make me feel a lot better.
Please Log in or Create an account to join the conversation.
- vlad_vy
-
- Offline
Less
More
- Posts: 3333
04 Jun 2013 08:56 #10549
by vlad_vy
Replied by vlad_vy on topic Mid flight TX reboot
With Fixed ID Tx rebind in most cases.
Please Log in or Create an account to join the conversation.
- FDR
-
- Offline
04 Jun 2013 09:30 #10550
by FDR
Replied by FDR on topic Mid flight TX reboot
It would be useful if you could post the errors.txt file from the root of your tx...
Please Log in or Create an account to join the conversation.
- RRacer
-
- Offline
Less
More
- Posts: 43
04 Jun 2013 14:01 #10558
by RRacer
Replied by RRacer on topic Mid flight TX reboot
Will post the error file as soon as I get back home.
Cheers!
Cheers!
Please Log in or Create an account to join the conversation.
- RRacer
-
- Offline
Less
More
- Posts: 43
04 Jun 2013 19:49 #10561
by RRacer
Replied by RRacer on topic Mid flight TX reboot
Here's the contents of errors.txt.
I'm running the tx from a 3s Lipo, and voltage was about 11,6 when it happened.
Today I flew six packs without any hickups apart from my own sketchy stick management
I'm running the tx from a 3s Lipo, and voltage was about 11,6 when it happened.
Today I flew six packs without any hickups apart from my own sketchy stick management
[Hard fault handler00000000
R0 = 40003800
R1 = 00000000
R2 = 20003729
R3 = 00000082
R12 = 00000020
LR [R14] (subroutine call return address) = 0800941d
PC [R15] (program counter) = 080201fe
PSR = 6100002e
BFAR = e000ed38
CFSR = 00000000
HFSR = 00000000
DFSR = 00000000
AFSR = 00000000
SCB_SHCSR = 00000000
impure_ptr = 080213e0
00000026
20000082
0000001aPlease Log in or Create an account to join the conversation.
- PhracturedBlue
-
- Offline
Less
More
- Posts: 4402
04 Jun 2013 21:35 #10568
by PhracturedBlue
Replied by PhracturedBlue on topic Mid flight TX reboot
This looks the same as the other hangs I've seen on 3.0.0
Basically the code is getting stuck waiting for the Module to complete data transfer. I have an idea how to prevent the transmitter from hanging in this case, but I have no idea whether it will actually resolve the issue or not.
Basically the code is getting stuck waiting for the Module to complete data transfer. I have an idea how to prevent the transmitter from hanging in this case, but I have no idea whether it will actually resolve the issue or not.
Please Log in or Create an account to join the conversation.
- RRacer
-
- Offline
Less
More
- Posts: 43
04 Jun 2013 22:50 #10569
by RRacer
Replied by RRacer on topic Mid flight TX reboot
Dear PB,
I have all the confidence in your work, I'm positive you will crack this one.
Only wish I could contribute in some way.
I have all the confidence in your work, I'm positive you will crack this one.
Only wish I could contribute in some way.
Please Log in or Create an account to join the conversation.
- PhracturedBlue
-
- Offline
Less
More
- Posts: 4402
05 Jun 2013 04:17 #10584
by PhracturedBlue
Replied by PhracturedBlue on topic Mid flight TX reboot
I just checked in code that should prevent the reboot you (and others) have seen on the Devo10. However, it is still possible that you will completely lose control of your model should the condition that was causing a reboot happen again. Since I can't reproduce it, there is no way for me to know.
I am currently also trying to get more log info than I currently have from the error.log, but it won't help any in this specific situation since the issue is a hardware communication error, and the real question is what needs to be done to recover from it.
I am currently also trying to get more log info than I currently have from the error.log, but it won't help any in this specific situation since the issue is a hardware communication error, and the real question is what needs to be done to recover from it.
Please Log in or Create an account to join the conversation.
- cmpang
-
- Offline
Less
More
- Posts: 296
06 Jun 2013 00:09 #10619
by cmpang
Replied by cmpang on topic Mid flight TX reboot
the usual way to trouble shoot such cases is to simulate the problematic hardware environment to test out if the corrective coding works...
In this case, may be you can, say, putting in a resistor in one or some of the communication connections to the RF module to weaken the link, try it in the original code to see if the problem can be reproduced.. if it could, then try it with the corrective code..
Another way is to simulate the communication error by a piece of code, may be a prolonged loop in between the RF module's acknowlegement..
In this case, may be you can, say, putting in a resistor in one or some of the communication connections to the RF module to weaken the link, try it in the original code to see if the problem can be reproduced.. if it could, then try it with the corrective code..
Another way is to simulate the communication error by a piece of code, may be a prolonged loop in between the RF module's acknowlegement..
Please Log in or Create an account to join the conversation.
- PhracturedBlue
-
- Offline
Less
More
- Posts: 4402
06 Jun 2013 00:51 #10620
by PhracturedBlue
Replied by PhracturedBlue on topic Mid flight TX reboot
It isn't a question of how the software works. The question is what state is the CYRF module in when it doesn't return the expected state. I can't easily guess at this, nor can I do much to recover from it without knowing the answer. So simulating it isn't valuable because I'm not trying to diagnose symptoms, I now need to know the underlying cause and actual response. Without the ability to reproduce the problem, that is not possible to determine.
Please Log in or Create an account to join the conversation.
Time to create page: 0.210 seconds
-
Home
-
Forum
-
General
-
General Discussions
- Mid flight TX reboot