3 October, 2018 at 03:43 #59780
This is the most cursed copter I’ve built…….
Everything looked great on the current limited bench power supply. First time on battery power, the left ESC and/or motor smoked. It runs, albeit very rough, and won’t go passed 50% throttle now. Either blew one fet bank or opened one motor phase, not sure which yet. I’ll diagnose tomorrow.
I should add the SBUS RX interface and SPort telem all worked correctly.3 October, 2018 at 08:30 #59786
Two words of advice… “Smoke-Stopper”.
I always power up with one attached the first time and have yet to blow components. Best yet, if i do see a fault via the smoke stopper, it cuts power. After which i can hunt down the short via a DMM. Best investment i have ever made and cheep too.3 October, 2018 at 14:09 #59796
That’s the exact function of the current limited bench supply. There were no shorts, and everything worked perfectly on the bench supply.
The blow out happened on the first full power run up under load with the battery supplying power.6 October, 2018 at 02:27 #59832
Looks simply like a component failure when I went to high throttle since everything worked fine up to about 1/2 throttle. The 3 motor phases all measure right around 0.4 ohms, no sign of a short. If you look very carefully at the ESC photo, the lower right FET is the problem, the top of the fet has blistered from the heat generated during the “poof” event.
Replacements on order!
Attachments:6 October, 2018 at 09:35 #59840
Was that a new ESC, if so you should send David a tell.6 October, 2018 at 10:14 #59841RCExplorer – DavidKeymaster
Already on it6 October, 2018 at 16:16 #59842
Good man!21 October, 2018 at 03:03 #60004
How did the ducts work out for you?
Curious to know if you had clearance over the sOat mount as that looked to be a tight fit.27 October, 2018 at 17:28 #60053
Okay, back in business. New ESC solder in, and everything seems okay on the bench supply. Some more setup to do, hopefully first flight later this afternoon or tomorrow!28 October, 2018 at 14:43 #60058
I’m “Assuming” your using dRonin? Just keep your sensitivity down after the Autotune or your likely to see oscillations. I dont know, maybe your setup will behave better with the OAT mounts.
Question: Wouldn’t larger props (disk area) allow for better pitch control? Far as i know of, larger props would have lower disk loading, higher gyroscopic mass and produce less prop-wash for the equivalent amount of lift.28 October, 2018 at 16:26 #60060
First flight done. Yes, using dRonin with the stock RCE components and KakuteF4V2. Only change made was to drop the default yaw gains to the same default values used by pitch and roll. It did not do anything odd, was able to fly forwards and backwards with no bad behavior. Roll and yaw seemed well behaved too. This was with all 3 axes in level mode, in the confines of my backyard, so nothing aggressive. Attitude mode, while behaving as I would expect, feels funny after flying everything else in rate mode.
So I went ahead and did an autotune. Again, nothing out of the ordinary, and FWIW, my motors didn’t even come down warm to the touch. I am running 3s rather than 4s, but it should still require the same watts (power) to hover, just requiring more amps to get there.
I will try the autotuned gains later, and explore horizon and rate modes too.
Larger props might be beneficial, but with the OAT mounts there would be some clearance issues.28 October, 2018 at 22:03 #60063
Okay, 2 batteries thru it, though still in the backyard. Sorry, no video at the moment. Getting 5 minutes on a 3s 1350 pack.
First attempt after autotune, pitch and yaw oscillations to the point where it wasn’t flyable. Zeroed out the D term on both pitch and yaw (same as I do on tricopters, 0 the D gain on servos), and it flew very well. Beginning to think the dRonin flight controls don’t do well with D gain on axes with servos. May have to investigate that some. It’s likely from the fact that the autotune is aimed at quads, and their typical responses. Add a servo on an axis, giving vectored thrust, and you get a much stronger control response, even though the servo is slower.
I tried Attitude, Horizon, and Rate on this battery. All modes flew well, I liked Rate mode best, although the pitch axis occasionally got confused, and it would pitch hard in the direction opposite flight. Wondering if the pitch integrator was running away.
On the second flight, I added LQG Rate, and LQG Attitude since I had an autotune the seemed to be performing nicely. They both flew well too, really couldn’t see a difference from Rate and Attitude modes.
It might be worth trying a mixed mode, where roll and yaw are Rate, and pitch is Attitude. Next time out.
The fuselage was following the commands fairly well, so I think the gyroscopic moment from the sOAT is doing it’s thing.
I need to connect my OpenLagger, so I can get some telemetry to look at when it does the hard pitch maneuver.
On one real hard pitch maneuver, I saw a slight hint of the reverse flight funky chicken dance, but it didn’t persist and flew right out of it. Hopefully I can catch this on the logger too.
This thing is going to be fun!30 October, 2018 at 02:58 #60070
No time to fly today, but played with the music feature of the blHeli32 ESCs. This is my first set of blHeli32 ESCs, in fact it’s my first set of dshot capable ESCs. So this copter is setup with dshot600 and the Star Wars Imperial March startup music. Too funny…….31 October, 2018 at 03:41 #60081
Weather has taken a turn for the worse, so did a little software work. Sort of a start on the modules necessary for when/if a custom BiCopter controller is needed, along the lines of triflight. I was real original, called this module biflight. All it currently does is calibrate the servo travels and measures the servo speed, so those values can be used in the telemetry feed.
Here’s a quick video of the process:
And here’s the settings object and status object:
Attachments:16 November, 2018 at 17:44 #60195RCExplorer – DavidKeymaster
Coooool! Good job on that!
I really want to see a video of that thing in flight. Does it look weird?
- You must be logged in to reply to this topic.