Hi you all.

I'm new to arducopter drones, and drones at all..

I've built a Quad, using APM 2.8 Board (non-3DR), M8N Gps, 470kv 24p motors, 15x5,5 CF props, and tarot 650 CF frame, all running on 4s 10.000mah and turnigy 30A Plush ESC.

In the beggining all worked fine, i'm started to learn how to pilot the Quad, using stabilize and loiter. obviously loiter resulted more attractive because was hard to me maintain altitude when making horizontal displacement in stabilize..

some day the quad started to toiletbowl and i got scared and did everything wrong and crashed at high speed to a 25 meters tree, cutted off some branches, and fall down. Result: 1 broken arm, 2 wasted props, and complete landing gear destruction. apm and gps was scratched too.

recently i fixed the aircraft, and this time I DID COMPASSMOT (not done before!), to avoid toiletbowling. I mounted APM with 4 3M slow response foam earplugs in a double plate, and sticked to APM and frame with double-doublesided foam tape.

In stabilize is rock solid, but alt hold and loiter works weird.....

in loiter, when you let the quad hovering, it twitches the throttle, like an angry bee, i don't know how to explain. it's like the quad stops and re-run motors for a fraction of a second, and looses about 2 or 3 inches of altitude and go up again. its like a glitch. I tried to fly loiter anyway, was making squares about 2 meters above ground to practice, and suddenly... quad rocketed up in the sky... this time i was ready, and switched to stabilize. the quad started to fall, and with a throttle punch it leveled again, but at an altitude i never used it... so i decided to switch to LAND mode, and the quad suicided itself drifting away and descending too quick. this time: 2 more props wasted, and a crack in the lower main plate. gps stand was destroyed as usual..

i spended much money to start my aerial filming project, and i do not trust this aircraft to hang a professional camera on it. i thought it will be like a phantom, but bigger, nicer and stronger, capable of carrying a full dslr camera. But it is only a big and dangerous machine, and i'm hating it so hard.

i will leave you the last flight logs. if somebody with experience and expertise give me a hand and analyzes my logs, i will thank you very much.

i can only think i have a vibration issue, so i ordered a moongel pad to reinforce damping.

The big problem is i'm in Argentina, and any replacement part or upgrade takes a month to arrive, and a ton of paperwork to claim it in customs..

looking forward to fix this, and sorry for my neanderthal english.

Juan.

Views: 3020

Attachments:

Replies to This Discussion

@Juan,

just another drone crash.

I studied 100+ drone crashes reported on the Internet.

Vibrations, fly-away, flipping on the ground, lost control, GPS fix lost, radio signal lost,

day by day, a new problem after the another problem.

You need ESC with RPM sensor, as discussed 10 years ago, to have RPM of individual

motor/ propeller under control.

If you can control RPM, sensors, vibrations, no GPS fix cannot harm your drone.

If you can't control RPM, you are always lost with your problem due to vibrations, sensor, IMU, Kalman Filter mismatch in control.

Hypersensitivity to data received from sensors should be damped.

If all motors spin at the same RPM rate, you generally can still control your drone.

Read twice, what Stephen wrote in just another thread

"

From practical experience, I know that the "apparent" rate at which they spin is never consistent with each other when using the autopilot.  This is because even sitting on the ground getting ready to take off, the autopilot does not feed each ESC the same throttle.  "

Just email me with more details just to save us against trolls at this forum.

Open Technology Park

Open Fabrication Lab

Peer To Drone Crash Investigators

manta103g@gmail.com

there is a lot of ardupilots flying out there without rpm feedback from escs. i just want mine flying like those.

and i don't want to spend any more money on it.

thanks!

I found This.

when i switch to loiter, he ThrOut value gets totally messed up. What can be????

Don't listen to Darius, he has no idea what he is talking about.  I don't have the chops to evaluate your log I don't think, but I think someone will.

What about Bar.Alt GPS.Alt ?

No input data to draw charts ?

ThrOut at max 540 could be due airframe resonant vibrations, making barometer

sin wave modulated, making APM to read min sin wave values for baro,

making APM to think your drone is falling down.

Attach IMU1 2 AccZ charts vs.  GPS.Alt Bar.Alt

From IMU AccZ we can read if free falling was sensed.

here you got it

and zoom in the final part:

Thank you.

The last chart is really excellent.

Not sure about GPS.Alt, if calibrated to match Baro.Alt  (alt above the ground)

since out of range to be plotted on the same graph.

Baro.Alt at 0, doesn't look to be calibrated either since at Baro.Alt = 0

you still get AccZ active.

AccZ at 0 means freee fall, above 0 value means accelerated free fall,

so your airframe, controller vibrate heavily at resonant frequencies.

Do you have loose hardware parts ?

read Stephen again

"

From practical experience, I know that the "apparent" rate at which they spin is never consistent with each other when using the autopilot.  This is because even sitting on the ground getting ready to take off, the autopilot does not feed each ESC the same throttle.  "

ESC with RPM sensor, if supported, is your way to study pre-crash problems every time something unusal happens to your drone.

If you can control every propeller's RPM spin rate, you can control your drone

at 100% to fly safely, exactly as in manual mode, you don't care GPS Alt, Baro.Alt, AccZ data.

I develop next gen precision control quad Pro, based on mathematical equations to get rid problems with sensors, hypersensitivity and SelfEgo logics of the drone, Kalman Filter, I have no any control over.

Hello Argentina.

My cousins live in Buenos Aires.

I don't know why you quote that, I said it because it is correct.  APM based vehicles don't use RPM logging because it is unnecessary.  If Darius can sucessfully create an autopilot and market it, I will eat my hat.  Basically he has no real experience with ardupilot and is feeding you a line of bull.

So..... besides hating Darius.... anyone took a look at my logs?

Juan, don't be discouraged.  All home built drones are difficult to get flying at first.  

I do not consider myself a big expert, but I have built two heavy lift camera drones, the first with APM 2.5 and the current with Pixhawk FC.  Each had trouble at first.  Forgive me if I am stating the obvious. 

My suggestions:

1.  Read carefully and follow the Wiki section on first flight, particularly the compass calibration, ESC calibration.

2.  Do a test very simple test flight.  Check your logs for vibration and electromagnetic interference.  The Wiki will show you how to do this.

3.  Fly only in stabilize and alt hold modes until you have succeeded with Autotune.

4.  Read the Wiki sections on GPS glitch parameters.

Don't Stop!, you will find the reasons for your bad behavior.

Hi! Thank you for your reply.

Since i'm on APM, a can not install 3.3, so i can't auto calibrate ESC. what should i do to calibrate ESCs manually?

vibrations are on the limits of acceptable levels stated on the wiki.

Seeing other posts i realized how important autotune is. i NEVER did an autotune, so i think is a good place to start. i'm not very comfortable going 5 meters high to do it, i don't trust althold so much to do that, but if it is the only way to go.

AccZ levels of my quad are safe to use AltHold??

if not, well.... i will have to go for vibrations first. and already did a earplug mount, so... whats next? adding weight to fc? a wooden cube taped to fc will help?

Thank you all

Hi everybody..

Today i did an ESC calibration (all at once method) and balanced propellers. After that, i did a test flight, and the results are strange... the pulsating behavior when in althold mode diminished a bit, but still there. in the other hand, vibrations are worse! Z axis is totally fu**ed up!

here you got my logs, and some graphs for lazy ones who still want to help :)

cheers

Attachments:

RSS

Groups

Season Two of the Trust Time Trial (T3) Contest 
A list of all T3 contests is here. The current round, the Vertical Horizontal one, is here

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service