I built this quad initially with KK2.1 and then upgraded to APM2.8
Basic specs are, 450 frame with 2212 920kV motors and 1045 props.
First, I was using FS TH9X tx with R9B rx. Was looking to improve range and in last flight, battery accidentally came off and by the time I landed the quad, it took fast spins in air.
I tested immediately after that and found Quad to spin in air.
Now I got FrSky DJT module with D4R-II receiver and switched to it.
My first test flight and same problem of spinning.
I checked on Mission Planner Raw sensor and found Roll & Pitch to be at 0 but Yaw was at 310.
I calibrated Accel, Radio and also compass.
Still the problem remains same.
Is APM sensors are damaged?
If I reset APM, will it help with clearing off-set and will be ok?

How is the APM mounted on the frame ?

On shock absorber board. X position.
I did factory rest but still position is same.

I just got to understand from one of the post that 'The yaw value is 0 = north, 90 = east, 180, south, 270 west.'
So, raw sensors results are fine.
But am not able to figure out, why this spin (yaw error).

Connect the FC to mission planner. Move the quad in roll ,pitch, and yaw axis.Does the HUD move accordingly?
Point the quad to north, east, west and south .Does the HUD show the correct orientation ?.
Have you disabled the internal compass ?.
Have you calibrated the ESC's after the radio calibration ?
Can you post some pictures of the copter ?
PS. APM is a retired board. Last firmware that can be loaded is 3.2.1
Anyway hope for the best. We shall make it fly !

Hope you have read the wiki.
http://ardupilot.org/copter/docs/connect-escs-and-motors.html
Are there any trims in your Tx?
Enable logging (RCin and RC out) and post a  bin.log here
Or http://ardupilot.org/

I tried almost all
Connect the FC to mission planner. Move the quad in roll ,pitch, and yaw axis.Does the HUD move accordingly? - Yes, it moves accordingly

Point the quad to north, east, west and south .Does the HUD show the correct orientation ?. - Yes, it shows correct direction

Have you disabled the internal compass ?. - APM has in-build compass but I have installed GPS+Compass (UBLOX). I am not sure, how to disable and will check how to do that. I remember reading somewhere about a jumper pin and noticed it is in place. Will try to explore if I need to remove that or not. But, with same settings, quad was flying good and last crash caused this issue. So, am not sure what is the fault.

Have you calibrated the ESC's after the radio calibration ? - Yes, I did couple of times.

Can you post some pictures of the copter ? - I completed today's flying time and am back to work. Will post today evening.

PS. APM is a retired board. Last firmware that can be loaded is 3.2.1 - I updated latest firmware. Do you think, Beta firmware is good to try?

Anyway hope for the best. We shall make it fly ! Bow - Thanks, am sure will do it with your help. Monsoon about to hit this part of the country soon. Would like to fix it before that.

Compass-there is a jumper near the GPS port. Remove it and select external compass in MP.
Twisted arms, 'not so up right' motors can induce yaw . but fc can adjust. Check all the prop tips align each other.
Check motor temp, hot motors (1&2) or (3&4)  are fighting yaw.
APM (8bit) ,MP will only load 3.2.1 !!!
If you have time , load rover or plane firmware and then load arducopter. It will ensure the board is reset fully.
BUT you will have to start from scratch !!!!.
PS. I think it is a compass calibration issue. From memory, external compass orientation for APM is roll 180
Good luck.

Clear. I infact did load another configuration and then back to X...to reset fully.
I calibrated inbuilt compass couple of times earlier also and was working fine.
Now what I noticed at the time of calibration is, while am rotating quad, points on screen are a bit off the axis and then after few data points it adjusts to centre. This was not earlier. May be, this is what offset is.
In stabilize mode, does bad compass make the drone go spinning?

I was going thru one of the similar thread where you were suggesting on external compass
https://discuss.ardupilot.org/t/compass-unhealthy-help-needed-for-a-rookie/21424/8
Was the issue resolved?

REMOVE props, go to motor test page in MP and check all the motors start (almost) at the same %.
Calibrate esc once again(all at once) ,check motor ,prop direction . do motor test again
Is there any chance for an ESC damage due to the crash?

I think ,to keep yaw orientation, it uses compass and accelerometer in all flight modes.
The two compasses might be conflicting. In flight the internal compass is affected by magnetic interferences from pdb and power wires.so disable internal compass ,do a compass calibration (an older version of MP will be better)

ok, thanks

On ESC damage, I gained lot of experience (4 in 2 months).
I don't think, ESC is damaged. All props start almost at same time.

Will check motor test also.

Hi are your props in the right order ? I mean if they are placed in the right order that is cw with cw motor and ccw with ccw motor.
Regards

Quote from: knvshrm on June 09, 2018, 11:26:19 AM
Hi are your props in the right order ? I mean if they are placed in the right order that is cw with cw motor and ccw with ccw motor.
Regards

I did this mistake earlier and learnt hard way. Result was flipping.
This time I checked that part and seems correct as quad was flying till the crash.
Even now, it takes-off but spins.
I was of the opinion that if propos are not in order, drone will flip.
I will try with working on external compass but somehow, I feel that, there is something else to do for this yawing.
Hopefully, FC is not damaged!!!

I removed the jumper from APM2.8 so that, I can use external compass (UBLOX NEO M8N).
I did factory reset: From Config->Full parameter list-> Rest to default
Loaded Firmware and then did Accel calibration.
In the meanwhile, HUD shows Bad compass health.
Still tried Compass calibration by selecting 'APM and External Compass'. Ticked Use this compass and also ticked external mounted.
Progress shows, Compass 1 Error: 99
I found in one of the thread 'if you have a 2.8 with external compass do not use the 12c, plug it into the plug under the GPS plug and you have to remove the jumper to the right of the GPS'.
Did that but no use.

Does the GPS unit get power?( LED lighting).
I haven't used a m8n, so can't  help you there....
But swapping the compass wires ( tx- rx)  might do the trick.

I just got from another thread....I was plugging I2C cable from GPS to port next to power.
In case of external, I was supposed to plug it below GPS.
Now calibration done.
Will try to fly tomorrow morning and update on progress.
Even if it works, one point is still left.....why internal compass not working!!!
Thanks for your support till now.

 
Glad to hear that.
The jumper is removed to disable the internal compass.

Now it is established, compass is innocent. Culprit is something else. Same spinning issue persists. Today, winds were a bit strong on top if it and craaaash.....frame arm where motor is mounted, broke.

Need to get what exactly is wrong.

I disassembled it completely. Changing the frame to a new one.
I should receive a new one (now it would be F450 instead of S500) in a day.
I think, one of the motor is rotating slower / faster.
In next couple of days, will try to fix this.

Hi,
     In ur first post ,you said it is a 450 frame. Original DJI F450 frame is a very good one. All other clones are crap.
Arms flex and twist. S500 ( with carbon fiber rods in arms) is a better frame. Anyway choice is yours !
        Have you turned logging on.?  Without a log , it will be difficult to trouble shoot. We will have to go by assumptions !. You said ,it happened after the crash. Check the motors for bend shaft, debris, or any abnormal sounds .
  Hope to see you in the air soon......

PS. ArduPilot has a steep learning curve !.....

My mistake because the description of earlier one said 'PCB Version S500 SK500 Four Axis Quadcopter Frame w/ High Landing Gear For F450' and posted 450.
I ordered F450 now, from Amazon....am sure it is a clone. Let's see, how it works.

I turned logging and tried to read files in MP. I downloaded all of them. Loads of data and very less I could understand and rest trying to. I forgot to post it as suggested earlier and I will do today evening.
After Y'day crash, I noticed one of the motor have debri in it and donno if it was there earlier also. This motor is having trouble in free rotation. I will open it and clean or try with a spare one.

https://diydrones.com/forum/topics/quad-is-spinning-on-take-off?id=705844%3ATopic%3A1580726&page=2#comments
This thread talks about similar issue and solution in IMAX settings / PID / changing FC.
Settings are the last one to touch at this point as I usually go with default except minor changes like props spin at arming etc.
Will study about them and auto-tuning while I go with re-assembling.

I am learning....

In MP, go to dataflash logs- auto analysis, then select the log you want .
Look for motor balance.
Or review log tab, and graph rcout1( motor 1) to rcout 4.( tick what you want in the right side drop down list)
From the graph you can see the commanded ( requested)pwm to each motor. All should somewhat overlap.

Till now, I used to filter on only error or things like altitude, SAT count etc. Also path on map.
Graphs I view for Yaw error and it is very erratic throughout the flight - while it is spinning in air.
rcout....I will check.