Replies to This Discussion

DeveloperPermalink Reply by Mark Colwell on August 10, 2011 at 5:42am

Malcolm you have earned your wings!, and save us from the same fate. Thank You for being first, it is tough, I have a pile of broken parts from AP original flight testing,
Don't let Murphy Law get you down, I admire your drive to get this software right!
Thanks Again
Cheers!

• ▶ Reply

Permalink Reply by malcolm churn on August 10, 2011 at 11:22am

Got the bit between my teeth now. Its not like its the first one i have ever put in, this by chance was a nice light crash :)

I'll get there Mark, thx's for the support :)

• ▶ Reply

Permalink Reply by malcolm churn on August 13, 2011 at 6:20pm

Just thought i would add the pre trial test without the canopy attached, sorry for me chatting away, but was really happy that it was flying :)

• ▶ Reply

Permalink Reply by 415porto on August 14, 2011 at 9:56am

Hey Malcom are you gonna update the build guide for us?

• ▶ Reply

Permalink Reply by malcolm churn on August 14, 2011 at 1:45pm

Hi porto, i'll go check the code tonight and update it tomorrow at work.
i need to have a quick chat to randy too.

• ▶ Reply

Permalink Reply by malcolm churn on August 16, 2011 at 4:25am

Have updated the guide.

• ▶ Reply

Permalink Reply by malcolm churn on August 14, 2011 at 2:18pm

Hi Randy did some more testing to day.

These are my findings.

Stabilise Mode.Tail sometimes turned without any input from me up to 150 - 180 degrees.I fixed it by putting the tail control direct to the RX

 

Loiter modeWhen switching from Loiter to Stabilise -- the helicopter would forget its level -- i would have to move the cyclic stick to try and keep the helicopter level. Then reboot the APM to get it to return to the level.

 

While in loiter mode the altitude hold was ok i think i can tune this with the PID controller.

The Loiter PID control.Couldn't hold it in loiter mode for too long, the helicopter would try and move to position but it had quite a lot of angle on the heli, and moved way too fast. I tried to turn down the PID. I think the best bet might be to have a maximum speed plus maximum angle. This would slow the heli down and also not allow too much angle on the Heli ... maybe 5 degrees.

 

What do you think ?

• ▶ Reply

Permalink Reply by Chad Colt on August 14, 2011 at 5:36pm

Malcolm and Randy,

 

With our tests on a Logo 400 this is pretty much what we saw.  We were able to get alt_hold mode to kind of work using the barometer (~3 meter accuarcy), but it was good enough.

 

We flew the helicopter about 50 yards away and switched to RTL mode.  During our first flight the heli did a very aggrestive tilt and tried to fly back to home but was dropping altitude so fast we flipped back to stabilize.  We recovered and I knocked the Max_Pitch number down.  While our aggresive behavior was inducing a large roll we thoerized that the heli code was reading just one paramater for the max movement.  It seemed to work as the next time we flipped to RTL we had a good experience.  The heli returned to home, overshot and flew back. 

 

We did two or three flights tuning the NAV PIDs using the RTL and we got it pretty locked in.  We had to stop testing though when we booted up the heli and stabalize mode had lost all swash plate control.  We've run into an issue where on random boot-ups the swash plate loses nearly all it's range.  We fix it by re-downloading the code and running through the setup again. 

 

As an aside, we've never run the tail through the ardupilot in the air because on the bench it wanders a fair amount.  We would love it though if we could get it locked in as having to constantly unplug the tail servo in and out of the ardupilot and the reciever is a pain in the tail. 

 

We would love to help fly new versions of the code with our heli.

 

Ardupilot Mega 3.6 HELI

Logo 400

Compass, UltraSonic Range Finder, all the trimmings,

• ▶ Reply

DeveloperPermalink Reply by Randy on August 15, 2011 at 4:50am

Great news.  I've never even tried RTL myself.  great to hear it kind of works.

 

so when you're saying the tail isn't run through the APM, you mean you control it directly from the radio?  Have you tried using the APM but also with an external tail gyro?

• ▶ Reply

Permalink Reply by malcolm churn on August 15, 2011 at 5:19am

yes i tried running through the APM but it kicked the tail a couple of times so i removed it and went directly to the APM, i was using the external Gyro.
So when you go to the gyro set-up screen, if you type gyro on, does this tell the APM you have a separate gyro?

• ▶ Reply

DeveloperPermalink Reply by Randy on August 15, 2011 at 5:26am

yes, in the CLI, you can type "gyro on 1200" and it will enable the external gyro and output 1200 on channel 7 (not 6 like the wiki says - i'll get that updated). 

 

The tail control in the code also works slightly differently if the have the gyro enabled - it will use a single PID loop based on heading.  if the gyro is disabled then it will use a nested control - a rate control within a heading control.

• ▶ Reply

Permalink Reply by malcolm churn on August 15, 2011 at 5:30am

Ah ok then i'll go back and try tonight with the Gyro on i had better results this way. could keep the tail locked in a whole lot better. i dont use the gain channel as my gyro's want to see a slightly different command from my TX.
I didnt think turning the "gyro on" made any difference to the PID control.
Superb that will sort that problem out.

• ▶ Reply

• ‹ Previous
• 1
• …
• 13
• 14
• 15
• 16
• 17
• …
• 161
• Next ›
• Page