Open AndreaPro opened 1 year ago
Duplicate of #482
In my opinion it's not a duplicate, it is a different issue. I have compiled the firmware with VTX and OSD already. If it can be of any help the VTX is a BetaFPV M03 with SmartAudio
Does this happen when making the same changes in the betaflight configurator?
No, it happens only when making changes to VTX parameters using the lua script from the radio. I've tried changing accelerometer trims and PID profiles without issues
can you try downloading the vtx tables again? https://github.com/betaflight/betaflight-tx-lua-scripts#setting-up-vtx-tables
Yes, I've tried multiple times both downloading them from the script on the radio and even generating the lua script from the configurator and copied it manually on the radio but the outcome is the same, as soon as I do one change the VTX table is altered and becomes unusable
That's strange. I'll have to see if I can reproduce
If there's something that I can do to troubleshoot the issue just let me know
Is it the setting of a specific parameter that causes the issue? Could you try 4.4.1 and maybe 4.4.0?
It happened when changing VTX power. I can try to go back to these versions and let you know
I have downgraded to BF 4.4.0 that was the previous release available for my board, and it happens the same. I've noticed one thing, that if I copy the VTX table lua script generated by Betaflight inside the script it works until I change a parameter, but if I download the VTX table from the quad it downloads a different VTX table that is the wiped one that I found inside the configurator. Attached there is the Lua script generated by Betaflight configurator and the lua script that is created inside the radio if I download the VTX table. And there is the VTX table json saved before and after the issue.
Seems like the lua scripts end up with the "empty" vtx table after downloading? We'll have to figure out why that happens then. I just tested this with a betafpv meteor65 board with 4.4.2, and I'm not able to reproduce. Vtx table is correctly downloaded, and I can change all parameters.
@AndreaPro Can you share a diff please?
@AndreaPro Can you share a diff please?
For sure! I have tried both with the stock CLI provided by BetaFPV and with my custom settings but I have the same issue.
Do you have stock settings for your VTX table? Because the CLI on the BetaFPV website for the Meteor 65 appears to have the same VTX table as mine
Got the same issue if I download the vtx table or copy the lua script from betaflight to my controller, it overwrites my vtx table.
Got the same issue if I download the vtx table or copy the lua script from betaflight to my controller, it overwrites my vtx table.
Could you please share more infos about radio firmware version, Betaflight version, script version, drone model, etc...?
Oh ye sure sry, Betaflight version: 4.3 Edge TX version: 2.8.0 (Radiomaster Boxer) Betaflight lua script version: 1.7 Flight controller: Speedybee F405 VTX: TBS Unify Pro32 (https://www.team-blacksheep.com/vtx-table-for-betaflight.txt) (unlocked)
After I try to switch a setting, my VTX table on the flight controller gets overwritten to the same table u shared
The only thing I can think of is that the payload somehow contains more bytes than it should, and this happens: https://github.com/betaflight/betaflight/blob/99be07729963e4b5e5cb540555450bda28bd2781/src/main/msp/msp.c#L3380-L3407
It's strange because in the BF lua scripts we're only sending four bytes back to the FC
@tomekste5 Which RX system are you using?
I am using Express LRS version 3.3 2.4GHz and as a Receiver I use the Radiomaster RP1
I am using Express LRS version 3.3 2.4GHz and as a Receiver I use the Radiomaster RP1
So we have in common ELRS 3.30 2.4GHz
maybe try disabling vtx administrator in the elrs lua script.
maybe try disabling vtx administrator in the elrs lua script.
I have tried VTX administrator in elrs lua script because the Betaflight script was destroying the VTX table, but I have never used it before discovering this issue
try disabling it by setting "band" to "off". did this happen with earlier versions of elrs? the payload sent back from the betaflight lua scripts is only 4 bytes, but if more bytes are sent the vtx tables could be reset. not sure where those bytes are coming from. needs more investigating.
try disabling it by setting "band" to "off". did this happen with earlier versions of elrs? the payload sent back from the betaflight lua scripts is only 4 bytes, but if more bytes are sent the vtx tables could be reset. not sure where those bytes are coming from. needs more investigating.
I have tried the Betaflight lua directly starting with ELRS 3.3.0. I will try disabling it but I am sure that the first time that I have noticed this issue I haven't enabled VTX admin yet
I have never enabled VTX administrator
maybe try disabling vtx administrator in the elrs lua script.
If it's not too much hassle, could you guys please try ELRS 3.0? I'm not saying elrs is the issue, but we need to rule things out. As I said, the payload sent from the bf lua scripts is 4 bytes, but it could be that zero padding or something is added somewhere along the way to the FC.
There is no target for the internal module of the Boxer radio in ExpressLRS v3.0
If it's not too much hassle, could you guys please try ELRS 3.0? I'm not saying elrs is the issue, but we need to rule things out. As I said, the payload sent from the bf lua scripts is 4 bytes, but it could be that zero padding or something is added somewhere along the way to the FC.
I can try it and let you know
Have the same issue with
Transmitter
Configuration through betaflight works fine (OSD or Configurator), just activating the LUA script even with vtx tables from betflight configurator switches the vtx channel to another band. After that nothing works and a reboot is required to reset to the old settings.
[EDIT] Tested VTX Admin in the ELRS lua script: that works fine and is able to reset the vtx config to a normal state. Never tried that before, but after this is set in ELRS the betaflight lua vtx script no longer changes the band.
Hi Any update on this one?
Same issue here: SpeedyBee F405 stack BF 4.4.0 BF Lua 1.7.0 ELRS 2.4GHz v 3.3.0 Radiomaster TX12 Mk2 EdgeTX 2.9.1
Works on my Mobula8 without any issues
If it's not too much hassle, could you guys please try ELRS 3.0? I'm not saying elrs is the issue, but we need to rule things out. As I said, the payload sent from the bf lua scripts is 4 bytes, but it could be that zero padding or something is added somewhere along the way to the FC.
I can try it and let you know
Hello, finally I have been able to do some tests. I couldn't downgrade to ELRS 3.0 because there was no Target available for BETA FPV AIO RX. I've tried the following:
TX3.1.1 and RX3.1.1 = OK TX3.2.1 and RX3.2.1 = OK TX3.3.0 and RX 3.3.0 = BAD TX 3.3.0 and RX 3.2.1 = OK
So the issue appears to happen with ELRS 3.3.0 on the RX side
@AndreaPro Thanks for testing! Maybe open an issue on the ELRS github, or ask on their discord. I'm sure they have an explanation for it.
@AndreaPro Thanks for testing! Maybe open an issue on the ELRS github, or ask on their discord. I'm sure they have an explanation for it.
Done! On ELRS Discord they suggested me to try the nightly of the luascripts, I've tried v20231020.10 but still the same issue with ELRS 3.3
Same here. Latest stable v1.7.0 or nightly v20231020.10 wipes VTX tables on save. Even reading the VTX tables goes wrong. Reading gives: Band 4 Channel 4 Power LV2 Pit Mode OFF Protocol SA Frequency 5800
While in fact my diff contains this:
vtxtable bands 6 vtxtable channels 8 vtxtable band 1 BOSCAM_A A FACTORY 5865 5845 5825 5805 5785 5765 5745 5725 vtxtable band 2 BOSCAM_B B FACTORY 5733 5752 5771 5790 5809 5828 5847 5866 vtxtable band 3 BOSCAM_E E FACTORY 5705 5685 5665 5645 5885 5905 5925 5945 vtxtable band 4 FATSHARK F FACTORY 5740 5760 5780 5800 5820 5840 5860 5880 vtxtable band 5 RACEBAND R FACTORY 5658 5695 5732 5769 5806 5843 5880 5917 vtxtable band 6 LOWBAND L CUSTOM 5362 5399 5436 5473 5510 5547 5584 5621 vtxtable powerlevels 3 vtxtable powervalues 14 20 26 vtxtable powerlabels 25 100 400
set vtx_band = 5 set vtx_channel = 3 set vtx_power = 3 set vtx_low_power_disarm = UNTIL_FIRST_ARM set vtx_freq = 5732
The mcu_id.lua on my SD card \SCRIPTS\BF\VTX_TABLES contains:
return { frequencyTable = { { 0, 0, 0, 0, 0, 0, 0, 0, }, { 0, 0, 0, 0, 0, 0, 0, 0, }, { 0, 0, 0, 0, 0, 0, 0, 0, }, { 0, 0, 0, 0, 0, 0, 0, 0, }, { 0, 0, 0, 0, 0, 0, 0, 0, }, { 0, 0, 0, 0, 0, 0, 0, 0, }, }, frequenciesPerBand = 8, bandTable = { [0]="U", "1", "2", "3", "4", "5", "6", }, powerTable = { "LV0", "LV1", "LV2", }, }
Same here. Latest stable v1.7.0 or nightly v20231020.10 wipes VTX tables on save. Even reading the VTX tables goes wrong. Reading gives: Band 4 Channel 4 Power LV2 Pit Mode OFF Protocol SA Frequency 5800
While in fact my diff contains this:
vtxtable
vtxtable bands 6 vtxtable channels 8 vtxtable band 1 BOSCAM_A A FACTORY 5865 5845 5825 5805 5785 5765 5745 5725 vtxtable band 2 BOSCAM_B B FACTORY 5733 5752 5771 5790 5809 5828 5847 5866 vtxtable band 3 BOSCAM_E E FACTORY 5705 5685 5665 5645 5885 5905 5925 5945 vtxtable band 4 FATSHARK F FACTORY 5740 5760 5780 5800 5820 5840 5860 5880 vtxtable band 5 RACEBAND R FACTORY 5658 5695 5732 5769 5806 5843 5880 5917 vtxtable band 6 LOWBAND L CUSTOM 5362 5399 5436 5473 5510 5547 5584 5621 vtxtable powerlevels 3 vtxtable powervalues 14 20 26 vtxtable powerlabels 25 100 400
set vtx_band = 5 set vtx_channel = 3 set vtx_power = 3 set vtx_low_power_disarm = UNTIL_FIRST_ARM set vtx_freq = 5732
Which drone and which ELRS version on TX and RX?
@AndreaPro home made quad. ELRS 3.3.0 on both TX and RX.
@AndreaPro home made quad. ELRS 3.3.0 on both TX and RX.
Which FC?
board_name OMNIBUSF4SD manufacturer_id AIRB
version Betaflight / STM32F405 (S405) 4.4.2 Jun 1 2023 / 02:07:34 (23d066d08) MSP API: 1.45
it probably has something to do with the new MSP VTX code in ELRS 3.3.0
updated to 3.3.0 now. All I have to do is start the bf lua and let it download vtx tables, and then they disappear from the FC
updated to 3.3.0 now. All I have to do is start the bf lua and let it download vtx tables, and then they disappear from the FC
That is not what happens here. Only when I hit the "save page" button from the lua script my vtxtables is missing from the diff all. Reading alone does not ruin my vtxtables.
iirc just saving a PIDs page with the lua script goes fine. (But i should re-test that when back at home.)
I just built elrs 3.3.0 without msp vtx support, and now it works fine, so there's definitely something going on in the RX code. downloading vtx tables works, and saving the vtx page works. vtx tables still there.
this problem is still there on elrs 3.3.1...will there be a fix or should we build our own fw without msp support? and does it only ereases the vtx table? i dont trust the lua script when changing on field pids or rates...
using edgetx 2.9.1 / tx16s mark ii iflight blitz f7 stack & blitz 1.6w vtx (nazgul v3)
btw i am new to this fpv hobby :) started 2 weeks ago
ahh...and i also tried to export the vtx table directly to the bf lua script vtx table folder but without any luck... the bf lua script just ereases the saved lua vtx table and sets everything to zero.
@bloudman648 It only affects vtx/vtx tables. Everything else should work.
There's not much we can do about this in the betaflight lua scripts. The script is working as it should and only sending the necessary bytes back when setting vtx parameters. The data is being tampered with somewhere else.
You could report this issue here too https://github.com/ExpressLRS/ExpressLRS/issues/2452#issue-1955368086
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs within a week.
I have exactly same issue with this. I'm using Radiomaster boxer using elrs 3.3.0 dan rx radiomaster rp1 same elrs versionon FC Speedybee F405 v3 Everytime I download vtx table, all reset just like this case. I did edit vtx-table.lua and copy the correct table and power but problem still remain even I have restart both tx and rx
https://github.com/betaflight/betaflight-tx-lua-scripts#setting-up-vtx-tables
Any luck with RX 3.3.1?
If it's not too much hassle, could you guys please try ELRS 3.0? I'm not saying elrs is the issue, but we need to rule things out. As I said, the payload sent from the bf lua scripts is 4 bytes, but it could be that zero padding or something is added somewhere along the way to the FC.
I can try it and let you know
Hello, finally I have been able to do some tests. I couldn't downgrade to ELRS 3.0 because there was no Target available for BETA FPV AIO RX. I've tried the following:
TX3.1.1 and RX3.1.1 = OK TX3.2.1 and RX3.2.1 = OK TX3.3.0 and RX 3.3.0 = BAD TX 3.3.0 and RX 3.2.1 = OK
So the issue appears to happen with ELRS 3.3.0 on the RX side
Any luck with RX 3.3.1?
If it's not too much hassle, could you guys please try ELRS 3.0? I'm not saying elrs is the issue, but we need to rule things out. As I said, the payload sent from the bf lua scripts is 4 bytes, but it could be that zero padding or something is added somewhere along the way to the FC.
I can try it and let you know
Hello, finally I have been able to do some tests. I couldn't downgrade to ELRS 3.0 because there was no Target available for BETA FPV AIO RX. I've tried the following: TX3.1.1 and RX3.1.1 = OK TX3.2.1 and RX3.2.1 = OK TX3.3.0 and RX 3.3.0 = BAD TX 3.3.0 and RX 3.2.1 = OK So the issue appears to happen with ELRS 3.3.0 on the RX side
I haven't tried it yet but if no one fixed it in the Betaflight or ELRS code repos I doubt that it will be fixed by itself
I am using Betaflight Lua Scripts 1.7.0 (I've even tried the nightly), Betaflight Configurator 10.9.0, Radiomaster TX16S with EdgeTX 2.8.5 and a BetaFPV Meteor 75 Pro updated to BF 4.4.2 and I am using ELRS 3.3.0 on BetaFPV Micro 1W TX and on the quad. If I open the lua script and I adjust some settings in the VTX section and save them and I power off the quad and connect it to Betaflight configurator the VTX table is wiped completely and the settings are not stored. I've tried generating the lua script VTX table from the configurator and retrieving them from the quad but the result is the same. Pratically the VTX section of the script is unusable because after restart the VTX table is completely wiped.