Acro mode! And more...

[Patscherpofel]

so I installed 1923, put to default pid and reloaded the previous setting...and it also flies all fine, and there is no yaw problem

but I can isolate and reproduce the problem with yaw now...
if I change the yaw rate tau to 0,05 the following happens:
if I let it hover straight for a few seconds and do a yaw input it is fine
but if I do a stronger yaw input ,then let it hover and then do another smooth yaw input straight afterwards it strangely counterattacks and also jumps...so absolutely not according to my stick input which is very slow

so far this is only a problem with different tau settings (yaw and tilt deadband is at 4)
I hope this may help you to find the problem Max

[Patscherpofel]

another problem I am having now is with the minimosd over mavlink I was using before
not sure if itis the new fw now, but I dont get the rssi info and I cannot switch the osd pages through channel 5 anymore
but everything else works fine ...gps, voltage,horizon,flight modes
could this be a change in the fw or should I search the problem somewhere else?

[Max]

Hi Pat,

Yeah, problem with the minim has been reported before... I "broke" it a while ago when I disabled sending radio channel Mavlink telemetry data by default. This data has to be specifically requested now, but minim can't do that. There's a thread about it somewhere. I guess I need to add an option of enabling that telemetry data on AQ side.

So far unable to recreate the yaw jumping problem, but I'll keep trying. Are you pretty sure this depends on the firmware version? (vs. eg. mechanical/frame issue.)

Also I'm still interested about the control/stability problems/differences you mentioned earlier on. Did you have to re-tune PIDs (Quatos?) for v1923 vs. 1852? If so, what did you need to change? What kind of differences were you seeing?

Thanks,
-Max

[Patscherpofel]

Yeah, problem with the minim has been reported before... I "broke" it a while ago when I disabled sending radio channel Mavlink telemetry data by default. This data has to be specifically requested now, but minim can't do that. There's a thread about it somewhere. I guess I need to add an option of enabling that telemetry data on AQ side.

that would be nice if I could have that option back
does this explain also I dont see rssi anymore?

So far unable to recreate the yaw jumping problem, but I'll keep trying. Are you pretty sure this depends on the firmware version? (vs. eg. mechanical/frame issue.)

I am pretty much sure I only had this issue with the newer fw (1919+1923)
I dont hink it has anything to do with the frame, since it flies really good now
and I also noticed it on a small 25cm frame I used 1923 first

right now I can easily reproduce it by chanfing smoothing filter of rate yaw...but I dont know if this is the cause because I never touched theses settings before the last weeks, and I noticed this strange yaw thing before that allready
maybe it has someting to do with the deadband I allways have changed, but never that much
the thing is I turned the deadband much further down then before since it now also effects the yaw, and that is not working for me at all to have a 40 deadband on yaw

Also I'm still interested about the control/stability problems/differences you mentioned earlier on. Did you have to re-tune PIDs (Quatos?) for v1923 vs. 1852? If so, what did you need to change? What kind of differences were you seeing?

I think this was due to a faulty escv3 I had...with this esc it was reacting very strange and I tuned it to pids I never had before, then I changed the esc and was surprised I could use my old settings again...so I think all is fine on this side

[Patscherpofel]

Max, any news on the the lua script for the taranis?
I have build the cable now and get some of the parameters to show up
really looking forward to test this more now

[Patscherpofel]

doing some testing today with camera and a 50mm lens
and I can still see the problem...
left yaw starts pretty directly
right is like this...first id doenst really happen anything sometimes it even goes to the left...then it sudddendly jumps and then goes slower again

not sure if you can see it is this log screenshot...mot_yaw against yaw input channel
it was a bit windy...if I would upload a video of the camera you would clearly see it though

I can see that yaw is not balanced well, (even though I aligned everything with extra care) but this should be a problem right?
it wasnt doing this on the older fw,I will try to set the yaw smooting filter off now and see what will happen

[afernan]

I have also a yaw "rebound" also in all my last build (r1923-Quatos-ESC32V3). It happens in any flight mode. It´s something we haven´t before.

I´ve tryed to tune without success changing yaw speed, yaw filter, etc. It can be reduced if attitude quatos params are reduced, but this will affect to the rest of Pitch/Roll control.

I´ll print some quatos curves to show that effect.

Angel

[Simson]

viewtopic.php?p=37319#p37319

Hi Max,

I would appreciate that , if you could add an option to enable RSSI over MavLink.
You are right, we discussed that a while ago in the MinimOSD Thread together with Adam.

Cheers, Harry

[Max]

Finally published my Lua scripts for custom S-Port telemetry: https://github.com/mpaperno/AQ-OTx-Lua

Note I have not tested this with OpenTx 2.1 (only 2.0). I added a layer which will help ease the transition, but it will need some work to get properly set up. If you use 2.1+, please help me document this transition.

The main difference in 2.1 is that all the telemetry field names have changed, and in fact can now be edited by the user. In some cases values that were available previously now have to be "calculated" in OTx telemetry setup.

Check the file in SCRIPTS/MIXES folder called TFlds21x.lua. This holds the mapping of telemetry fields used in the AQ scripts to the names used in OpenTx setup. Many of these are set to OTx 2.1 default values, but some (probably) aren't and some need to be set up specifically ("calculated" fields). Check the comments in the file, it should be fairly obvious what's what.

Also, for Angel... Obviously none of this is tested on Horus. I'm not sure what the display scripts will do... maybe they'll just work, maybe not. The code is fairly modular so it should be relatively easy to get to the telemetry data and "just" create custom display scripts for the Horus. AQTelem.lua is where the data decoding magic happens.

If any questions/difficulties, just ask.

Cheers,
-Max

[Max]

I've also published a new firmware update, build 1925.

##### 7.1.1925 - Nov. 13, 2016 - `feature_integration` branch

* Fix new YAW deadband parameter not implemented correctly.
~ Adjust yaw controls for smoother manual operation around stick deadband area.
+ Add new option to send RC_CHANNELS Mavlink telemetry message by default (eg. for minimOSD).

Pat, I'm curious if this helps the yaw control in your situation.

For the last item, to enable sending the RC channel message you'll need to manually change the CONFIG_FLAGS parameter (until I update QGC). Just take the current value and add 64 to it. So:

Code: Select all

CONFIG_FLAGS = CONFIG_FLAGS + 64

Example using default value:

Code: Select all

CONFIG_FLAGS = 16 + 64 = 80

(You can also OR the values together if you really want to get technical about it )

Builds are up now at ftp://ftp.autoquad.org/3/334693_22529/a ... tegration/

Cheers,
-Max