Pixhawk 2.4.8 ( bangood version) crash analysis

Hi all, i am new in the drone building world ( aabout 1 year) and finaly i found this community where people share experience but most of all are helping others.

So, first of all excuse my English.

I build this drone :

Tarot 680 pro

Pixhawk 2.4.8 bought from bangood

ZTW spider pro esc 30A opto tiny version ( i've changed some parameters as in the attachment)

Sunnysky v3508 580 kv with 13*55 prop

Multistar 2x 4s 5200mah 10c total of 10000mah

Current sensor and voltage but never made it to work...

Holybro telemetry, gps, gimbal, 2xstepdown for 12 and 5v

Well, as you can see from the crash video, he didn't lose power, the motors rotate during the crash like it was just armed. . At middle of descending you can hear beeps becouse i believed that i touched the kill switch by mistake, and did quickly on off to be sure it isn't switched. Beside, before touching ground, i give some throtle and i could hear the props trying to spin.

I must say another thing : this pixhawk was already in a quad before ( with an totaly another configuration ) and did a similar crash, don't know if i can find the log. But after that crash i put itin a plane and did some wonderful flights, it flow for a few hours with no issue, and even did a few autonomous flights of 10 15 kilometers with absolutely no issue. Even did autonomous landing.

In this one i flow less than 1 hour. I did autotune of roll and pitch only.

The ztw esc i bought them by mistake, didn't see they are small, but then i tried them with one motor at home and on my stand with motor and prop, at full throttle they drink about 15 amps with 1200g of weight.

After crash i controlled all electronics and was all perfect. Battery was at 3,85 aprox all cell of both batteries.

Only Gimbal broken. :( and a two arms and prop

I already repair it and is functioning well without props, becouse i was without spare, now i'm waiting them.

Size (kb) 11300.671875

No of lines 134260

Duration 0:07:22

Vehicletype ArduCopter

Firmware Version V3.4.4

Firmware Hash 8a196281

Hardware Type

Free Mem 0

Skipped Lines 0

Test: Autotune = NA -

Test: Brownout = GOOD -

Test: Compass = GOOD - mag_field interference within limits (20.55%)

Max mag field length (641.26) > recommended (550.00)

Test: Dupe Log Data = GOOD -

Test: Empty = GOOD -

Test: Event/Failsafe = FAIL - ERR found: CRASH

Test: GPS = GOOD -

Test: IMU Mismatch = WARN - Check vibration or accelerometer calibration. (Mismatch: 0.99, WARN: 0.75, FAIL: 1.50)

Test: Motor Balance = WARN - Motor channel averages = [1649, 1560, 1604, 1608, 1549, 1656]

Average motor output = 1604

Difference between min and max motor averages = 107

Test: Parameters = GOOD -

Test: PM = NA -

Test: Pitch/Roll = NA -

Test: Thrust = NA -

Test: VCC = UNKNOWN - No CURR log data

these are links for the other documentation

Log file : https://www.dropbox.com/s/y94nvghm1vxlrya/CRASH%20tarot%20680%20PRO.log?dl=0

Short video of crash : https://www.dropbox.com/s/7et62ag03dd0bcb/crash.mp4?dl=0

ZTW esc setup : https://www.dropbox.com/s/pmkglglpqokqs3p/ztw%20changes%20for%20sunnysky%20v3508%20580%20%20%20kv.png?dl=0

Thank you in advance for patient.

If voltage dropped enough, given wind, instability , a flip is not unthinkable.

I would attach a power module that works.

would test the ESC's for low-voltage behavior, (using a lab PSU) , but you can do that with batteries.

depending of how far you'll go:

you could just fly low/hover till battery gets critical (cells don't really get damaged until below 3.0v/cell , several manufacturers say 2.8v/cell, - but given poor calibration and no momentary landing ,people tend to operate at higher minimum voltages)

I would not recommend to fly at anything close to 3v/cell , but as a controlled test it can be done whithout damaging cells - that would verify motors/PIDS at such voltage , given proper safety /(net or soft area) that's an essential test, if you can do, you'll get to know how it behaves. (it's most likely not flyable already at 3,2v, as the internal resistance will be too big.)

Replies

Mihai Nicolae Sichim March 4, 2017 at 12:36pm

No, I haven't made the tests how it should.

I will first make some low alt flights with various flight modes and than after a 1 hour of flght i will sacrifice 1 battery.

Hope that my esc setup is right for my low kv motors.

Thank you for your help.

If i find a problem i will post it in this discussion so that other people learn without destroy anything.
Andre K. March 4, 2017 at 11:52am

If voltage dropped enough, given wind, instability , a flip is not unthinkable.

I would attach a power module that works.

would test the ESC's for low-voltage behavior, (using a lab PSU) , but you can do that with batteries.

depending of how far you'll go:

you could just fly low/hover till battery gets critical (cells don't really get damaged until below 3.0v/cell , several manufacturers say 2.8v/cell, - but given poor calibration and no momentary landing ,people tend to operate at higher minimum voltages)

I would not recommend to fly at anything close to 3v/cell , but as a controlled test it can be done whithout damaging cells - that would verify motors/PIDS at such voltage , given proper safety /(net or soft area) that's an essential test, if you can do, you'll get to know how it behaves. (it's most likely not flyable already at 3,2v, as the internal resistance will be too big.)
Mihai Nicolae Sichim March 4, 2017 at 11:06am

In this case i must mention that the battery as it is only 10c is very weak, and i know that after the crash was 3,85 per cell, but in the air with all six motors maybe failed under 3,5 per cell and after crash restabilished.

But i remember that when voltage drops dramaticaly ( it happend more than 1 time with other quads ) the drone it is loosing weght easy before loosing all, not fliping.

Can it be anyway the battery but if i didn't setup failsafe on battery he did this flip? And the sensor i bought it from bangood too, and cannot calibrate it by folowing the guides, so maybe the sensor isn't trustable and could have to do with the crash??

If it was you what kind of test would you run before going up so high again?
Andre K. March 4, 2017 at 10:36am

yes , a single motor failure would look different, I could have been PID's that were too slow, if some esc's reduced performance due to low voltage or heat.

that could explain instability and flip,

I forgot to mention; in the video you hear motors spin down, then flip.

in the log, at 527sec , before flip, you can see all motors getting high demand - so as it failed, and clearly reduced thrust, the pixhawk commanded very high throttle.

So - voltage, heat, .. anything.. but it's clear that the motors (not only one) failed to perform.
Mihai Nicolae Sichim March 4, 2017 at 9:51am

Yes, you maybe right , but i forget to say that here was a strong wind, and maybe did that to correct it's self becouse was in loiter mode?

How did you get that kind of grafic?- yours is more clear and easy to read. I only get the black one in mission planner with log analyse with different axis view.

Anyway, should't an six axis drone remain in the air even with one motor-esc failing?

Thank you for the analyse.
Andre K. March 4, 2017 at 1:00am

the last 20 seconds before the flip, you can see it oscillate in pitch axis pretty badly, +-15 degrees.

Can it be some ESC's hitting voltage cutoff, low voltage, or overtemperature protection ?

You have no battery voltage/current logging, that would have helped..

Screenshot from 2017-03-04 09-58-45.png

This reply was deleted.

Discussions

Pixhawk 2.4.8 ( bangood version) crash analysis

Replies

Activity