Old gas fixed wing pilot just getting into quads…
I have bench run my new Iris+ tied down (luckily!) and the throttle is a mess. It behaves almost normally below 50% in that I can increase and decrease power with the throttle stick. But when I go above 50% and increase to full power the throttle just quits working. The motors stay at full power no matter where I move the throttle stick down to 50%. They just run on as if I am doing nothing. Then if I chop the throttle the motors will slow down but in 2 “steps”, not smooth and proportional.
I found and have done the ESC calibration procedure twice. After doing the cal while still in cal mode the throttle works perfectly and is smooth and proportional thoughout the entire range. Exit cal and restart Iris+ and the throttle is back to it bad behavior. And not just bad, this is dangerous as there is essentially no proportional throttle control and the beast would surely get away from me if not tied down.
Anyone have any ideas how to correct this? As it is I am afraid to fly this thing, even in the automated modes.
Here are a couple of video clips.
- This one shows the issue with bad throttle behavior with Iris+ armed for normal operation.
- This one shows the normal correct throttle behavior in calibrate mode after calibrating the ESC.
Providing tlogs and/or data flash logs will greatly assist in troubleshooting your issue.
Calibrate the radio too. Maybe the throttle channel is acting up. Wouldn’t hurt, but if it’s okay in calibration maybe not. Curious…why the bench test. Did you try flying it and it acted oddly? Look at the radio’s screen with the radio on. Look at the read out of the throttle channel as you move the stick and see if the bars are linear to movement or jumpy. If jumpy that could be the issue.
Bench tested because I am a newbie to quads and wanted to check it out… plus it is really cold out. I did also try the radio throttle calibration. Made no difference. Throttle isn’t jumpy. It behaves exactly the same every time.
Looks like the youtubes didn’t embed so I’ll post the links later.
I’m not sure that a quad-copter that relies on g-sensors and GPS (in certain modes) will behave correctly tied to a bench indoors. You’re telling it to “go up”, so it applies power to initiate the initial "takeoff maneuver"and is waiting for the g-sensors to report back that the request was performed (quad lifted off the ground). As Far as the “computer” is concerned, it appears the quad must be “very heavy”.
I may be way off, though. I’m just operating on hunches and guess at this point.
Interesting thoughts, Texan. But I am not sure it is that smart. Hopefully 3DR will weigh in here at some point… soon before I send it back.
Direct links to youtube vids in case embed didn’t work above:
- Showing the problem
- Showing normal throttle in cal mode.
You know, the more I think about this the more I think Stexan is onto something. Standard altitude hold mode must allow the baro and accel sensors to influence the throttle. So maybe being tied down is the issue. Probably same for Loiter. Is there any way to put Iris+ in a purely manual flight mode? Not that I’d fly it that way but would like to do a ground test in that config. I think I’ll get into MPlanner and see if I can temporarily define Standard Mode as manual.
I am just not used to having a throttle that doesn’t react instantly and accurately to my inputs. Hopefully this is normal on Iris+ and this is just a case of me slowly getting into the brave new world of automated aircraft.
I was thinking the same. It’s all fly by wire. Maybe the Pixhawk is confused by not seeing any vertical movement. I guess you could pick it up props off, move it vertically and see if it changes.
OK, I found this on DIY Drones while researching something else: “AltHold and Loiter are Fly-By-Wire modes, no direct link to throttle. Above 50% on the stick is the rate of height gain, below 50% gain is the rate of descent.”
So I’ll mark this thread solved and re-post later only if necessary after a real flight… but since it is 5 degrees F with 40 mph winds outside safe to say that won’t be today!
Anxiously awaiting your flight test. I too agree with some from above, that the Pixhawk “IS” smarter than you think. My only experience so far is with Naza Lite, Naze, KK2 and some Multiwii. My Iris+ w/Pixhawk4 is totally different, so to speak. On my first test flight, where I went up a bit higher than comfortable, when I pulled the throttle back to descend, it hesitated for just long enough to get my stomach in my throat, then slowly descended, I wound up at just a tad below half throttle, for a nice touch down. Also the “LAND” switch works quite well too. Same thing when throttling up, not at all like giving the gas to a little 250 quad. Nice and easy. In my mind, the Iris+ acts like a good camera ship should act. I suppose you could tweak things, and get different results, but so far, I like mine, just the way it is.
i have read thru this post and i am having the same exact problem.
i have not seen if this was a normal thing or not for the IRIS+
i am afraid to try to fly mine before i know…
Its not a problem. In the default mode which is Alt Hold, it’s fly by wire, so below 50% throttle you are controlling the decent rate, and above you controlling the ascent rate. If it’s not ascending it gives more throttle, no props, not movement, max throttle (well near max).
Again, you are controlling the rate of ascent/descent, not the actual throttle
If you want to see linear throttle response you will need to put it in stabilize mode (which requires reassigning the modes using Mission Planner to APM Planner 2.0)
You will then see throttle response linear to the stick input, and more like the KK boards mentioned
ok, i kinda thought that…
also in some of the forums i saw that it says the iris should be configured in the cross “V” and not in the "X"
configuration…if this is so,where in MP do i find that ?
If it’s a new Iris and you have not updated the FW it will be configured as V type. But either X, V works,the same. The V just adds slightly different thrust curves to the motors mix due to its less uniform motor layout.
I see a lots of posts of these bench tests, pretty much most them are incorrect interpretations of what should happen, if it’s RTF, just better off and go fly it. It is after all RTF!
For maiden flights, if you fly over grass, in a wide open space. If something is wrong, a quick landing or flip won’t do any damage. Test fly it over concrete in a confined space, and the outcomes would be much worse.
ok, thanks… i’ll reply with my results later, thanks again