Definitely you will like Loiter for that type of work. It was taylor made for it. It will basically fly the helicopter, short of whatever angles and altitude you want to make it do, which you can do with the sticks. So you can concentrate on running your payload.
So you can make a pass over a site, for instance, where you want to get an aerial view. Set up your payload for the right angle and whatnot. Now put a little cyclic into the heli and start it moving over the site and make your shot, and itâs really smooth and fluid. Even letting go of the cyclic and letting the heli come to a stop is really smooth. Thatâs with the default settings. Like Bill said, you can turn 'em up to make it more âsnappyâ but then it kind of ruins what the mode is for, because even with a gimbal youâll see the jerky movements especially in video.
Super smooth is the name of the game in aerial videography and cinematographyâ.
Chris,
Sounds good. If the wind ever dies down a bit i will hopefully get a few more tuning flights in tonight. After watching Robâs videos of his helicopter just sitting there in loiter under quite windy conditions, ive been quite keen on trying it out.
My aerial platform was admittedly flying quite well already, but the pilot load gets to you after 5 or 6 packs under less than ideal conditions. Im truly hoping this makes my helicopter more of a tool to use vs something i have to fly constantly and pay less attention to the job at hand. Thats the only reason i like my multis, i find i can concentrate on positioning for shots with ease vs having to constantly battle the wind correcting etc. Im hoping this does the same for my helicopter let alone the possibility of guided auto missions for site mapping. Very excited about the prospect of that.
Thanks,
Tim
Oh, it will. And youâll find the helicopter much more capable and efficient than big multiâs.
You did something I have not seen done yet - you made a helicopter fly with a Pixhawk 2.1 running the latest beta. I think itâs pretty much just a matter of tuning it up now. Too bad you had some hardware problems there with the pin on the wiring. But I guess that happens. Still never got your video of the flight but when I look at it on my Android tablet maybe that will get it.
Which, BTW - for flying in the field a 11-12" Android tablet with Tower works pretty awesome. Itâs all I use in the field. My tablet has a full-size USB port on it and the MavLink radio plugs into it and âjust worksâ. If your tablet only has a microUSB, you can get a OTG cable to make it work. And the mission files are fully compatible between the PC ground stations and Tower on the tablet. I find the tablet and Tower much more intuitive to use - to create a mission you can just draw on the screen with your finger where you want the heli to fly and it creates the path and puts in all the waypoints for you
Chris,
Im going to upload the videos to youtube and link from there this evening. I used a DSLR so the file sizes are quite large and maybe that is causing the issues?.
As to the pin issue, i guess that stuff just happens. I wont sweat it too much, im just glad is turned out to be such a simple fix.
I use a Microsoft Surface Pro tablet, so its running the newest incarnation of windows 10⌠When windows isnt beind windows i love it. Its what i picked up this christmas just for field work. I also intend on seeing if i can use my Galaxy Note 5 phone with Tower and an OTG cable for quicker missions/flights. Im hoping my luck with the Pixhawk 2.1 continues and i get a good flying heli out of it.
Tim
Well, i got the P fain up there. Roughly .15-.17, ill have to check in MP to get an exact figure of where i left it. I bumped it up and snapped it back and forth on the roll axis to induce ossicilation and backed it off a tad. Tried to add in .15 on the I gain and it had a little slow ossicilation in hover so i backed it off a bit to .12 and it seemed good.
Since it was late in the day and i was a bit weary from the day, i left the dowels on âjust in caseâ. Im guessing they could contribute quite a bit to ossicilations in of themselves. Tomorrow is supposed to good weather here and i have the day off so i should get some flights in to dial it in more without the dowels. Im sure ot will act a bit different. I will say as it sits, even with those silly 6â 3/4" dowels through the landing gear, it was hard to get a negative tendency out of it when jambing the sticks hard left then right in an attempt to get it to blow out. I went back and forth in a rythmic fashion trying to tease something out, but dident get anything.
Cant wait to keep dialing it in, man it feels a fair bit better with some I gain in there.
Tim
Tim,
Great video! Looks really good. What made you stop increasing the P gain? Iâm assuming you saw some type of oscillation developing? I would like to see the log file. If you did see oscillations was it as high of a frequency as mine?
Itâs interesting that you noticed an oscillation with the increased I gain. Was this more when you were trying to control it or did it occur with no stick inputs?
Bill,
That video was from yesterday. I have another from today I will upload that shows the slow ossicilation from setting the I gain to .15 initially. Dropped down a few points and it went away. It would ossicilate on its own sitting in a hover without any cyclic input.
As to the P gain, i could have gone higher i think had i not had the dowels attached but what was happening was when i stabbed the cyclic hard it would ossicilate so i backed off until it stopped when jabbing the cyclic. Just sitting there in a hover i probably could have gone past .2. Ill revisit that tomorrow without the extra landing gear on.
Ill upload logs from todays flights with the I gain added in.
Tim
Tim,
The aircraft is starting to see an oscillation with the higher P gain at approximately 5 hz in the roll axis. This may go away widen you remove the wood dowels but I suspect it will return at some higher P gain. I didnât have a chance to look at the oscillation due to the I gain. Did you put it on the tuning knob or just change it in the param file.
Bill,
With the I gain i just plugged in .15 and test flew. Then backed if down to .10 and it seemed much better. Im likely going to remove the dowels and start P gain tuning again and then add in some I gain now that im more comfortable with the unit and know what to expect. I have a sneaking suspicion that extra gear is fouling things up.
After im dead set on P gain ill assign I gain to channel 6 and see where i can get that.
Tim
Bill,
Before i forget again, how are you determining there is an ossicilation at 5hz? Is this something that is easily ascertained by looking in the MP log browser? Or is more of a you know what to look for by experience thing? Is there a set procedure to determine instability frequencies?
Thanks,
Tim
Click on review log in the dataflash logs tab. Load your log. so there are a few things you can look at. along the right side youâll see all of the messages with data. I either look at the desired and actual attitudes in the ATT message. If you can see them there then you should have saw it during the flight. The other place to look is the IMU message and look at the gyrX for roll and gyrY for pitch. The other message that is pretty conclusive is the RATE message. Look at R and Rdes for roll and P and Pdes for pitch. The first one is the actual rate and the second is the desired rate. To determine the frequency you will have to click on the time button right below the plot and then right click in the plot and click on âset scale to defaultâ. Then you can click and drag to zoom in. I normally zoom until I see the x axis show 1 second for each tick mark or 3 or 4 seconds for each tick mark labled. Then it is just a matter of counting peaks per second which translates to frequency in hz.
I looked at the logs again the I gain did not seem to have the effect you though it had. The contribution of the I gain to the command signal was 2 orders of magnitude less than the P gain.
Bill,
Thanks for the detailed explanation, its appreciated. Ill have a go at working through the process tomorrow.
As to the I gain, on my second to last flight there definatly was a continuous ossicilation at .15. A slow wobble on axis, kind of like a top thats about to tip over at the end of its spin if that makes sence.
What ill do tomorrow is re-create the scenario and attach a video of that specific condition along with the appropriate log before i take the wood dowels off and retune the P gain etc.
I admittedly have to get better at matching logs with videos etc for the sake of useful records. The whole logging while disarmed thing created a ridiculous amount of logs while setting this up initially and now its time to wipe it all and get to a more orderly file system that will be helpful for tuning.
Thanks again,
Tim
Tim, Iâm curious about your tail setup. Did my params from the 700 work with your 800 rotor pretty much, except for reducing the P a bit for the longer throw on your servo arm?
Great videos BTW - I really doubt the sticks are affecting it much. If was a 450 or 500, yeah. But not that much with a 700 or 800. Might want to consider just leaving them on there until you get done tuning. I remember back when I was figuring out how to tune my first flybar I put a bunch of P gain in it and got it to head shake 25 degrees in roll both directions at about 8-9 times per second. It wasnât even controllable and I wished I had some sticks like that on there because I ended up with grass stains on the blade tips.
When windows isnt beind windows i love it. Its what i picked up this christmas just for field work. I also intend on seeing if i can use my Galaxy Note 5 phone with Tower and an OTG cable for quicker missions/flights. Im hoping my luck with the Pixhawk 2.1 continues and i get a good flying heli out of it.