Copter-3.6.9 has been released as the official/stable version for Multicopters and Traditional Helicopters. The changes vs 3.6.8 are listed in the ReleaseNotes and also copied below.
Changes from 3.6.8
CX-OF flow sensor enabled on all boards
CUAVv5 Nano board support added
RangeFinders supported on all I2C ports
Compass maximum acceptable offsets increased
Septentrio GPS driver loses logging of camera feedback which could cause lag
Bug Fixes:
a) Acro mode leveling fix for when ACRO_TRAINER was 1
b) Tradheli collective position pre-arm check removed
c) Fallback to microSD for storage fixed
d) GPS blending memory access fix that could lead to unhealthy GPS
e) fixed POWR flags for 2nd power module on fmuv3 boards
bootloader binary update for many boards
This is the same as 3.6.9-rc2 which has been available for beta testing in the past few weeks except for the last two items on the list (6e and 7) which we’ve decided are low enough risk that it’s OK for us to include them even though they weren’t in the last beta.
Thanks very much for those who contributed to the beta testing and please report any issues either in this thread or perhaps better yet in a new thread.
@FireFly, this should reduce some logging failures that people see but it won’t resolve them all I’m afraid especially those that are hardware issues that we just can’t get around in software.
hello, is it possible to help me out to use this firmware for the simulation via SITL ? when I am using the simulation its always showing copter 3.7.0 in mission planner but I want to use the new firmware stable for the simulation.
Many thanks
2019-05-31 12-20-00.bin (770.4 KB)
Hello sir,I’m attached herewith log files.kindly help me.
My Hexacopter approaches landing Very smoothly,but when it Touches Ground,propellers getting more speed automatically and Flips Away left side.And Voice instruction appears that Crash Disarming.
Actually It;s unable to Detect landing.Kindly help me to solve the issue.Please click on the Below Link.
Randy, I’ve also had an undetected landing with an old machine, that I’ve upgraded from 3.5.7 to 3.6.9-rc2 right before you pushed .9 to stable. I was landing it in tuned params for a save after an autotune flight following the upgrade.
Two props hit the grass before “Crash detected, motors disarmed, parachute deployed” from my Taranis Yaapu telemetry. Luckily I didn’t arm the ballistic chute before taking off
Seems the problem is still lurking. I have my main bird running 3.6.6 still after a pair of undetected landings with 3.6.7 and I didn’t encounter one problem touching down.
Landing detection is a bit hard I’m afraid. If the vehicle moves (or the GPS signal moves) in the final moments before landing then the vehicle may touch down with a position error and then it will fight to correct that but it won’t be able to because it’s on the ground.
Some things that may help the performance of this vehicle in a general sense which may help with detecting the landing:
there is a bit of a motor imbalance and some of the motors (2 and 5 in particular) are constantly hitting the bottom of their range. These are both CCW motor (counter-clockwise) so I suspect one of the motors (perhaps not one 2 or 5, it could be another) it not mounted completely vertically. It’s a bit of a guess, I can’t be sure but re-checking the motors and props are all aligned properly would be good. Another option is to increase the battery size to add weight.
try reducing MOT_THST_MIN from 0.15 to a lower value, perhaps 0.07 but you can use the motor test page to find out where the ESC deadzone is and then set MOT_THST_MIN 2% (or so) above this value. Some details are here on the wiki.
This looks like a slightly over powered vehicle and it’s using the default tune which probably means it’s overtuned (i.e. gains are too high). The Pitch gains in particular look high (from looking at the RATE.PDes vs Rate.P logs). Maybe try reducing ATC_RAT_PIT_P/I/D by 20% each might help.
Hi,
I have upgraded to 3.6.9 and all seemed well.
I did a test flight on 6/6 and all seemed well stable as always, no issues.
However, I hovered for a couple of minutes at 6m and then started to up to about 10m, but this time I noticed a weird reading of a negative (-1) altitude values on my OSD, never happen before.
I landed immediately without incident.
Sharing my log file here, any input will be appreciated.
Gal.
A couple of years ago I remember that Curtis Youngblood implemented a strategy that improved helicopter landings.
When a heli lands, sometimes one of the skids touches the ground first. When it touches the ground the flight controller perceives it as a flight force and tries to compensate. Sometimes if your landings were not perfect the flight controller would cause a control output that could cause a tipover.
Youngblood programmed a switch on the transmitter that he could hit just before touching down. The switch would reduce the gains on the flight controller’s stabilization algorithm.
This lessened the effect of having the flight controller try to make corrections when the machine was on the ground.
Here is a perfect representation of the effect that you guys are talking about, this time with my SingleCopter experiment:
@sergbokh, sorry for the slow reply. I’m hoping we can start beta testing Copter-3.7 within 2 months or perhaps a little less. I think we need to make a few more improvements in object avoidance first. So in particular get BendyRuler and Dijkstra’s working for Copter.
@Corrado_Steri, sorry, I’m actually not sure. I must admit that there are so many moving parts and the release is still far enough way that I haven’t consolidated the full list of changes yet.
