CleanFlight Ervaringen
- Topicstarter BorisB
- Startdatum
van daag weer eens getest en dat viel flink tegen kwa auto tune.
hij was bijna niet in de lucht te houden en flipte zelfs moet je echt wachten tot de piep of kan je eerder stoppen en het een paar keer doen want dit gaat niet als verwacht.
vanmiddag weer getest nu met pid5 en op de hand afgesteld wow wat een verschil zeg.
van knuppel om naar knuppel kietelen.
de p moest vlink om laag van 4 naar iets rond de 2 de i en d nog niet getest leek nu wel aardig ok toch een vraag is hier nog veel meer uit te halen of mooi stabiel laten zo.
hij was bijna niet in de lucht te houden en flipte zelfs moet je echt wachten tot de piep of kan je eerder stoppen en het een paar keer doen want dit gaat niet als verwacht.
vanmiddag weer getest nu met pid5 en op de hand afgesteld wow wat een verschil zeg.
van knuppel om naar knuppel kietelen.
de p moest vlink om laag van 4 naar iets rond de 2 de i en d nog niet getest leek nu wel aardig ok toch een vraag is hier nog veel meer uit te halen of mooi stabiel laten zo.
Laatst bewerkt:
BorisB
Forum veteraan
Geen idee. Ik heb het een keer getest en zag geen gewenste resultaat
.Zelf pids afstellen werkt toch het beste
KISS ESC
Why are you using 1.33? If you calculate dimensions for cmix it will be more 20/16=1.25.
Any good reason for that or just a hunch?
BorisB
Forum veteraan
in my case it wasn't 16. 15 and something I believe.
Cmix is not really about the dimensions in this case, but it was a workaround to get the same flip and roll speed on harakiri pid controller.
Harakiri pid controller has coupled rates to pids. That means that higher P resultd in slower rotation speed. On these H frames the pitch P is always much higher and by adding a mix you can make the flips equal to rolls on this way.
This is not needed anymore as there is a separate pitch/roll rate setting on cleanflight 1.8, but I changed to other pid controller till they decouple the rates from the pids [emoji1]
These are the Pid controllers with decoupled rates:
1, 2 and 3
Ic... Don't you think its actually good practice to use custom mix for H??? 1.3 its quite a lot of a difference from X.. Can i assume that with proper mix its wise to use same P and D on both axises? So, basically, effect of PID sum will be multiplied by cmix value to get to actual value. Its beneficial to have same pids for in flight adjustments i think.
Well I term indeed is a different but its least of a problem. The pitch and roll rate seems to be different in CF1.8, But still, what do you think?
BorisB
Forum veteraan
Only thing it will do is lower your pitch PIDs [emoji1]. In other words it will only change a number, but not the flight characteristics.
A H frame has symmetrical individial pitch, roll and yaw. By symmetrical I mean that sum of all motor vectors is always 0. All motors have same distance from the cg/cot.
Once you have a spider frame for example things can get different.
Thats exactly my point. Why to have different pids for pitch and roll it it can be the same... Makes it a bit easier, dont you think? Custom mix is one for all. So everyone can use it. If P and D at least are the same as well, will be easier to tune using inflight adjustments (Which is great btw).
BorisB
Forum veteraan
Well.....I have to say that's exactly what I thought initially. Unfortunatelly the pitch axis has more weight / inertia what makes it difficult to determine the right mix ratio. This means you have to tune the both P first and after that determine the mix. That leads into more parameters to change, which doesn't make tuning easier.
I also use inflight adjustments and love it. You don't have to worry about the same values as both values are being increased and decreased. Once you have a P ratio of pitch and roll it will always be the same [emoji1]
Understood.. Makes sense. Thanks. So, you are off harakiri for now? 1.8 RC has separate R and P rates, at least in dev version of the configurator.
BorisB
Forum veteraan
Yes I tried the latest 1.8, but my ezgui app doesn't allow me to change the rates unless I use cli. I see those separate roll pitch rates more like a workaround actually. They really need to decouple the things like on other pid controllers.
I do think harakiri still has the best potential as all features and modes just work great. But once you are used to not have to worry about the rates while tuning it is annoying in rate mode.
I am now back to PC3 and PC1.
it's roll/pitch vs yaw in this case [emoji1]. Cant make up my mind yet. Need more flights.
Ik neem aan via Bluetooth Boris? Welke module gebruik je daarvoor op de Naze32? Of klik jij er hem er gewoon via USB aan..
Hehe, ez-gui dude usualy fixes quite fast. I had my kitkat issue fixed in 2 hours
EzGui works via serial port, doesnt matter is this usb otg or bt or even telemetry radio link.
If you use it with BT, I like to know which module. You need a BT module on the Naze right?Hehe, ez-gui dude usualy fixes quite fast. I had my kitkat issue fixed in 2 hours
EzGui works via serial port, doesnt matter is this usb otg or bt or even telemetry radio link.
Just build my first FPV quad, blackout B330. I thought 90 degree headers would be handy on the Naze32.. not! Now I need to pull out the telemetry cables each time I connect the Naze through USB with my Macbook. BT could be a solution.
Laatst bewerkt door een moderator:
My experience is, that also with BT you need to disconnect in order to be able to use USB.
Plug and play BT module is also available via HK. (it has already the correct settings).
Plug and play BT module is also available via HK. (it has already the correct settings).
I think this pins (in the middle) are shared with usb serial (ON NAZE) (after usb chip there are diodes to select one or another)... What Dom says, you can put MSP on softserial or something to get it work simultaneously. Well, its not hardcoded anymore.
Pity there are no schematics of naze available.
BorisB
Forum veteraan
ik heb zowel usb als bluetooth, maar voorkeur gaat maar usb[emoji1]
Lekker makkelijk. kabeltje erin en het werkt altijd.