Hi DIY community,
Flying this arducopter is turning out to be the biggest nightmare of my life. I have been using a jDrones frame with the standard 850kV motors and 20 amp ESCs. I assembled the frame fairly easily, configured everything and started testing it. After great effort I managed to correct the drift it had been suffering. So I thought of testing some Loiter and thought the GPS would help a little in keeping it in place and not drift off. Unfortunately, that was the worst decision, since Loiter actually did not work at all. It ended up crashing into a building nearby. I was flying on an open ground, with atleast 15m on each side. I was the logs I collected. Could some tell me whats going on? I have sort of convinced myself that the arducopter is not as friendly as it is being advertised. Is it just me or this thing does not work for any noob?
Another possibility is that the ppm (pwm) signal is too close to a mode change. Or your RC radio connection looses signals at a certain times.
Did you confirm that you had GPS lock before going into Loiter mode?
George, you need to get the level readings into the AP. It needs to be level, and you need to tell the AP to store these accelerometer offsets into eeprom. Sounds complicated, but the MP does it for you. You just got to do it. Flying isn't user friendly or easy unless you by a plane ticket.
Listen to me. The accelerometer chips and gyro chips aren't perfect due to manurfacturing defects in the micrometer scale. These nanometrical imperfections must be compensated for on a case by case basis. That is why mems chips and pretty much everything else int the high tech world must be calibrated.
Ya, I guess it's not user friendly for the average guy.
One other problem I realized that could affect it is the radio. I am using a cheap FlySky (32 bucks one). So even when I calibrate it nicely and many times, in the setup menu of CLI, I still get the error that 'your radio is not calibrated'. However it performs farily well in tests>radio. So I ignored this part. However sometimes to arm my motors, I need to use little yaw trims since pushing the stick to the right is unable to arm the motors.
How much would that make a difference?? Do I need a better radio??
My first copter it's an hex a from jdrones. I had no problem at all, except some oscillations on high winds....but no problem at all. Just take off, flight and land again. After that, I've installed the sonar (already with shielded cable and LPF) and guess: AL working perfectly!
Last test: loiter mode. My loiter is working fine.....without wind. I've made my second test last weekend and need running for loiter....when wind start, my hex a goes for about 4~5 meters then go back an be stable. Without wind, they still stable at same place from start.
I also tested for first time RTL and it worked perfectly too. The copter goes high to default altitude (which i guess it's 8 meters) then go to home, stay there at 8 meters until i get control back.
For me everything worked from the first time....just need some running now.
BTW: i don't have xBee, so i didn't tested auto mode yet.
I think of it as very friendly. What normally would take years of research and development is functional enough that any n00b can build and crash it. :)
There are a lot of little verification and tuning steps, like checking COG, using Auto Trim (you mentioned correcting drift? did you use TX trims or auto trim?) and setting up certain parameters ... it is so user friendly that you might even skip some of these steps and still get it working well enough that it crashes well when you try to do something your quad is not yet ready, by virtue of skipping a step, to do.
There have been some changes which I am not tracking on completely, but generally speaking, I think the order of operations is to get stabilize working well, but some people recommend acro instead, and then working on AH, which can be tricky, then going to loiter, then working on Auto. Each step has its own tuning and some special knowledge about how to switch between modes and what to expect (such as how the throttle operation varies between stabilize and AH.)
I know this does not address your question, I hope it helps to address your expectations.
The brilliance of ArduCopter is not in how "plug and play" it is, as you might expect with some commercial APs using in pre-designed commercial frames, with no variation in the build. It is rather in its adaptability to handling different frames, to being expanded to suit your needs. That adaptability, the potential to get it working well with all different frame types, styles, different prop and motor combinations, and being able to expand on the base system.... that comes at the cost of needing to test several different modes methodically. Unless, that is, you buy it preassembled and pretested from jDrones, where Jani's team do it for you.
I know how you feel. I was starting out a month ago, and thought that if I got it together all right frame-wise, solder-wise, and TX- config wise, all would be cool, and I could just start flying and playing. Well it was not! I scratched my head over several problemes, and with not much success in reading the manual either, info scattered all over the place. All sorts of stuff gave me some hard days. Almost gave up in the beginning, thinking this is just not working. BUT :
I kept in there and after a while I got a better understanding of things, got help from the community (very responsive), and kept on, and today I am more than happy. I have flown my first Auto, it was amazing watching the self governed vehicle fly the way it had been told by clicking a map! WOW. I have still some problems with adjusting PIDs, and you know what, but when you at last get to a stable loiter, and work your way from there, the rewards are great, I think.
But again, there is a mismatch between the advertising in the wiki, and real life, it is not THAT easy!
I had the same problem in the beginning, LOITER made the copter fly away. I think for me it helped a lot to get the copter stabilized in flight with "auto-leveling" after that it was all another game.