You can ignore these warnings with an H743.
2 Likes
Iām attaching pictures of 2 crossfire receivers.
The 50 version is the suspect - loosing bing, the 45 is a working receiver.
As you can see, there is a lot of differences between the boards.
Another detail,
the version 45 is using a PIC microcontroller, while the 50 is using an ESP32 microcontroller.
Trappy think that if it works with encryption on then someone near you is using a ādebinderā which are apparently a thing.
@juvinski
Thats interesting about the 45 version. I wish I had one to try. Perhaps the code that was written for the pic has some hidden issue that does not translate well to the ESP32.
I think its reasonable to say the problem shows up with Ardupilot and ESP32 versions. I have been using one of my nanos on a 5" Betaflight quad for a few weeks without any problems.
The only problem with that theory is info from Alex Ivanhoff and his YT channel (see previous posts) is that he has reports of this loss of bind with nano 45/50 pic and ESP32, diversity RX and with Ardupilot, Inav and Betaflight. Im not convinced its the same problem.
I think I am at the point of giving up with CRSF. Who would fly any RC that was 50% likely to disconnect on the next flight. Everything I have is now using CRSF and all the Ardupilot craft are disconnecting.
ELRS is looking very tempting with its rapid development and hardware choices.
And, the last stable firmware update was 2022. Unbelievable.
The reports of this seem a bit too widespread to support the theory of anti-drone activists. Unless they are getting really aggressive all over the world.
Other than the initial configuration pain associated with any new system, Iāve been very happy with ExpressLRS. Highly recommend it if youāre looking to make a change.
2 Likes
And Im getting disconnect with the TX - RX 1mtr apart indoors in a rural part of the UK so no, I think I can rule it out in my case.
I did start to think about the rise of Meshtastic LORA systems in the UK but checking on a āheatmapā of owners I think that is also not the cause.
1 Like
Hi @Yuri_Rage
I donāt think an anti-drone activists or something.
I flew in the order - Swordfish (13 km o max distance and 40 minutes flight), Zohd Altus - (40 minutes),
ZoHD 3 minutes flight ( and the crash of crossfire), and the same Swordfish more 40 minutes and a Dolphin more 20 minutes.
The only difference is the receiver of the Zohd - version 50, while all other is using the 45 and Iām flying flawless for more than 2 years.
If you succeed in getting past the range where the link rate drops to 50Hz you can go for many miles. For me that drop occurs at about 200mts. I estimate there is a 5% chance that the link drop will cause the problem, so if you are flying around within that area, the link will be jumping 150/50 and in my case, it will defiantly disconnect sooner or later.
Hi @Vince_Hogg
the first flight I made with the suspect receiver I had several link rate changes.
While the second and fatal flight
1 Like
Like I said, I think there is roughly a 5% chance of failure at every link rate drop. I think it may be less in Stab mode (or manual in plane) believe it or not.
I have 100s of logs exactly like that. I dare not fly my plane anymore, twin Snowgoose. I only fly the quad and have set rally points. Every flight since February is to try a new theory about the disconnect. Designing, building, flying for fun are all on hold waiting for this problem to go away.
@Vince_Hogg so happy to see your post!
Normally I ran my Nano and Nano Pro at 150Hz as MAVLink 2 telemetry with rc override. Range with fsk was not great as expected, especially for the Nanos. However, I never had the loss of bindā¦until nowā¦
I converted 2 drones to SerialN option 23 in order to try out Yaapu and 50Hz mode. I also turned on multibind so I could try out different RCs.
Everything seemed to go ok but on usb the gcs would report a lot of messages about telemetry rate. I took the drones out to run an autotune and the nano rx lost bind a few minutes into the flight. I never flew beyond 200m, but yaapu was reporting the same crsf telemetry messages.
I took the drone inside and had to rebind. Chalked it up to an anomaly. But today the exact same thing happens. Drone is 100m away and my tx is blasting 500mW and I lose bind.
Fw
-Ardupilot 4.5.1
-TBS 6.33 beta
-edge tx or freedom tx (depending on rc)
Hw
-matek H743
-nano rx
-nano rx pro
-tangon and rm boxer + fullsize crossfire module
could you please share the version (what is the text on the board) of your Rx Nano ? 43, 45, 50 ?
I have reports and a log from a client with the same things occurring. This caused an RC loss and crash into the ground.
The following messages occurred a bunch right before the end. approx. 1min 30s worth.
Andy if you have settings (firmware/ CRSF) you would like me to try let me know? I do have a few CRSF and Tracer modules.
All of my current builds are ELRS.
2 Likes
Good news! We have a working group between AP and TBS engineers setup to investigate this. What I would like to do first is establish the exact setups that have run into this problem. So if you have seen this please can you post the following:
ArduPilot Version
TBS RX and TX Hardware and versions (You should be able to get these from Agent-M)
TBS RX and TX Firmware versions
Flight Controller
Flight Controller UART used for RX
Also if you have been flying for a while without issue and then an upgrade caused this please mention this and what you upgraded.
2 Likes
@andyp1per
Its very good news its getting looked at.
Ive had lost bind with all the following setups
Radio Frsky X20s running Ethos
Arducopter 4.4.4 and 4.5.1
Arduplane 4.4.4
TBS TX Full size HW 1.03
bootloader 2.05
FW - many from 6.16 to 6.33 (6.33 was terrible. Very frequent lost bind)
Also another Nano bay TX which bricked after FW update.
TBS RX HW nano 1.01 marked XF nano 50 (3 the same)
FW as above.
TBS RX nano pro HW 1.32 (2 the same)
FW as above.
Matek H743 v3 (2 off)
Matek H743 v1
Many uarts tried - no difference.
All flying before this was with Frsky RX. Problems began as soon as I started using CRSF.
It may be worth looking at the difference between old and new nano RX. Older ones used a pic processor and I think there are no confirmed reports with these. I think the last version with pic was marked 45. Newer ones use ESP32.
But ALL my failures have occurred as the RC link rate drops from 150 to 50. I believe it drops into LORA modulation at that point.
Great news @andyp1per
ArduPilot Version: 4.5.1
TBS RX and TX Hardware and versions Rx v3 hardware 1.01, bootloader 2.04
Flight Controller SpeedyBee F405
Flight Controller UART used for RX - Uart1
I have 4 wings with the version RX 45 and working like a charm, there version 50 I had the first flight and in the second one after 2 minutes the receiver lost the bind with the transmitter. No modifications between the flights and in the OSD I can see when the message of link rate droping from 150 to 50 I lost the bind. In the first flight I have a lot of this messages and I didnāt lost the bind.
When were those newer RXās purchased, this year? Might it be safe to say that any RX bought last year is probably still V45?
Iāve put this yearās start of the (real) flying on hold because of these reports, but most of my receivers are a few years old, some are also Nano Divās. So maybe I shouldnāt worry?
EDIT: Of course I could check every PCB for version number, but I was hoping to avoid that since some of my receivers are not that easy to access.
Most helpful to just post what you have and if there are issues.
Hi @UnknownPilot ,
the receiver Iām having trouble was purchased 1 month ago(version 50).
All others (version 45) were purchased since 2022.