Firmware update Ver 2.25 After you update post your outcome here.

[vr-pilot]

Did the update on about a dozen of our customer's drones and all well here.

The FW 2.25 on my YTH PRO RS is the first update that changed the overall behaviour apparently to be unpredictable:
Flying in Angle mode on long distances is not a problem. The bird in my case starts behaving erratically when using Smart Mode(s) close to the ground. Also landings in tight spots that require some precision in Angle Mode became really twitchy. I was wondering where this sudden "strange close to ground behaviour" could come from.
I analyzed the FlightLog data files and compared problematic Follow/Watch Me Mode hovers and Angle Mode landings and I was surprised:
After the FW 2.25 update low altitude hovers and landings had high negative altitude readouts while still in the air!
Before this update negative altitude readouts were much more seldom (and with lower values).
Below: Follow/Watch Me Mode (green) hovers at low altitudes (and switching to Angle Mode (purple)) with negative readouts of up to 5 m:

The result: It was not possible to land the bird as throttling down did not land the aircraft. As a "toilet bowl" effect began to develop additionally the only way out in between some high trees was UP!
(Image taken from "Flight Data Evaluation" programme v2.6 by Helmut Elsner. Flight on Sat, OCT 1st, 10:10 UTC)

Below: Landing with high negative altitude readout of 3 m while still in the air:

The result: it was difficult to land the bird as it behaved like skating on a wet piece of soap!
(Image taken from "Flight Data Evaluation" programme v2.6 by Helmut Elsner. Flight on Tue, OCT 4th, 17:20 UTC)

I was always able to safely land the bird, but without some practice, especially during the first event, it could have been the end of the story...
So is it possible that there came a general baro sensor read out problem with the new FW?

 

[ArnhemAnt]

I was always able to safely land the bird, but without some practice, especially during the first event, it could have been the end of the story...
So is it possible that there came a general baro sensor read out problem with the new FW?

Silly question, but did you do all the calibrations, and got confirmation, after the new FW was installed?

 

[vr-pilot]

Silly question, but did you do all the calibrations, and got confirmation, after the new FW was installed?

There are no silly questions... I did all the calibrations as required and recommended. Until FW 2.24 I did not have this kind of effects on exactly the same spots doing exactly the same things under comparable conditions. Even the K-Index was <4 during those flights. Radio Blackout and Geomagnetic Storm impacts were also "green"...
Something went astray with this FW update (even for me this time).
Changing software IMO is often like trying to make a table stop wiggling by starting to shorten one of its legs...

 

[THoff]

What kind of surface were you over when descending? The IPS has a ultrasonic sensor and a high-definition camera to recognize ground features and try to maintain the drone's position, but some surfaces (water, featureless concrete, wind-swept vegetation) can present problems.

 

[ArnhemAnt]

vr-pilot - maybe try to install the FW again and then re-do calibrations? Kinda like a Windoze restart when things go bad.

 

[Rayray]

The FW 2.25 on my YTH PRO RS is the first update that changed the overall behaviour apparently to be unpredictable:

Confusing. My Pro with the 225 update is as steady as before. At times I intend to switch from RTH to Angle and discover (with no crash) that I pushed it to Smart mode. Also I fly mostly in Angle for everything, and forget when I switch to Smart to first move away from the H if it's in the air close to me. I've learned not to pay much attention to the height shown on the ST-16. It usually zeros +/- 5' at startup. I doubt the update is at fault.

 

[Aguster]

Completed the FW update last night and did 1.1 hours of flight today without issue. Stays on station, maintains altitude...rock solid. I've only had the drone for a few days and I am really impressed with how it fly's. One thing to note, as vr-pilot indicates, the altitude in the ST-16 is notoriously erratic while on the ground. Not sure if this is normal or if it can be calibrated for 0. Overall, very impressed!

 

[Tuna]

As I understand it, altitude is largely measured by the barometer and so is quite sensitive to changes in conditions during the flight, as well as drifting somewhat. Now that we're heading into autumn I have no idea if that's going to add to the problem.

As someone noted, drift in the Typhoon in general can come and go with the software taking readings from the combined GPS, compass and barometer. Flying a 10 meter square pattern has been reported to help the system adjust. Let us know if the issue gets better or worse with successive flights.

The altitude is automatically reset to zero at the point of take off, so drift whilst on the ground is 'forgotten'.

 

[QuadBart]

Thank you, Thank you!!!!! The behavior you describe is almost exactly how mine behaves after the last update. Landing as I described in my other post is almost a battle with it constantly moving around in all directions. You can't simply push the left stick down gently, it takes some throttle up as well to keep it descending smoothly and constantly making corrections to get it to land where I want....

More data points to show/give Yuneec when I send my H in for diagnostics and hopefully repair.

The FW 2.25 on my YTH PRO RS is the first update that changed the overall behaviour apparently to be unpredictable:
Flying in Angle mode on long distances is not a problem. The bird in my case starts behaving erratically when using Smart Mode(s) close to the ground. Also landings in tight spots that require some precision in Angle Mode became really twitchy. I was wondering where this sudden "strange close to ground behaviour" could come from.
I analyzed the FlightLog data files and compared problematic Follow/Watch Me Mode hovers and Angle Mode landings and I was surprised:
After the FW 2.25 update low altitude hovers and landings had high negative altitude readouts while still in the air!
Before this update negative altitude readouts were much more seldom (and with lower values).
Below: Follow/Watch Me Mode (green) hovers at low altitudes (and switching to Angle Mode (purple)) with negative readouts of up to 5 m:
View attachment 3353
The result: It was not possible to land the bird as throttling down did not land the aircraft. As a "toilet bowl" effect began to develop additionally the only way out in between some high trees was UP!
(Image taken from "Flight Data Evaluation" programme v2.6 by Helmut Elsner. Flight on Sat, OCT 1st, 10:10 UTC)

Below: Landing with high negative altitude readout of 3 m while still in the air:
View attachment 3354
The result: it was difficult to land the bird as it behaved like skating on a wet piece of soap!
(Image taken from "Flight Data Evaluation" programme v2.6 by Helmut Elsner. Flight on Tue, OCT 4th, 17:20 UTC)

I was always able to safely land the bird, but without some practice, especially during the first event, it could have been the end of the story...
So is it possible that there came a general baro sensor read out problem with the new FW?

 

[RiverRunner]

There will usually be some "ground effect" with any multirotor as you get very close to the ground. The air being pushed down by the rotors meets the ground and pushes back up. This makes the craft a bit wobbly. It could also affect the barometer. This is normal behavior. All of my multirotors do this. My H does not behave any different in this regard after the upgrade.
If your problems are beyond what is expected from ground effect, then try recalibrating everything or contact customer service.

 

[gwhuntoon]

That skating around effect when trying to land makes me think it's due to a negative altitude reading. The skating is when it thinks it's on the ground and you fight to get it lower. Due to incorrect data the drone is fighting being pushed into the ground, normally a good thing.

I would also like to better understand how the barometer actually recalibrates and when. We have a lot of high and low frontal boundaries pass through and I'm beginning to think if I was flying before a front passed through and continued after I should see an altitude change on the ST16 even if the H is holding a steady altitude. If this is the case then flying when a front passes through is probably not advised.

Greg

 

[RiverRunner]

That skating around effect when trying to land makes me think it's due to a negative altitude reading. The skating is when it thinks it's on the ground and you fight to get it lower. Due to incorrect data the drone is fighting being pushed into the ground, normally a good thing.

I would also like to better understand how the barometer actually recalibrates and when. We have a lot of high and low frontal boundaries pass through and I'm beginning to think if I was flying before a front passed through and continued after I should see an altitude change on the ST16 even if the H is holding a steady altitude. If this is the case then flying when a front passes through is probably not advised.

Greg

The H is perfectly capable of landing at a lower altitude than it took off from. I think It only knows it is on the ground when you give it down throttle and it stops descending. Make sure you have obstacle avoidance turned off when you land.

 

[vr-pilot]

That skating around effect when trying to land makes me think it's due to a negative altitude reading. The skating is when it thinks it's on the ground and you fight to get it lower. Due to incorrect data the drone is fighting being pushed into the ground, normally a good thing.
Greg

That is exactly what I think and how it feels. Just 1 or 2 meters above the ground my YTH PRO RS (since FW 2.25 update) shows signs of avoiding to descend to the ground due to already negative altitude readouts.
IIRC in the beginning of August this year (FW 2.10), the software made the YTH stop/slow down descends below 1 or 2 meters when the gear was still up. Maybe here something has changed or causes confusion?
I will only have a chance to make some more tests from next Saturday on (during daytime).

 

[vr-pilot]

What kind of surface were you over when descending? The IPS has a ultrasonic sensor and a high-definition camera to recognize ground features and try to maintain the drone's position, but some surfaces (water, featureless concrete, wind-swept vegetation) can present problems.

My flying/landing area is over green grass, mower cut high. Rayray was told by Yuneec that the IPS only kicks in indoors when GPS is lost or unstable. According to this IPS may run outdoors in standby only. But maybe this is only valid for the camera and not for the ultra sonic sensors?

 

[vr-pilot]

There will usually be some "ground effect" with any multirotor as you get very close to the ground. The air being pushed down by the rotors meets the ground and pushes back up. This makes the craft a bit wobbly. It could also affect the barometer. This is normal behavior. All of my multirotors do this. My H does not behave any different in this regard after the upgrade.
If your problems are beyond what is expected from ground effect, then try recalibrating everything or contact customer service.

I know what you mean, but I had no such issues with FWs 2.10, 2.15, 2.21 and 2.24 doing all the calibrations in a standardized manner under the same conditions in always the same area...
Something went astray with this FW for my YTH PRO RS this time, so I won't take too much of a risk until the next FW is out...

 

[vr-pilot]

The H is perfectly capable of landing at a lower altitude than it took off from. I think It only knows it is on the ground when you give it down throttle and it stops descending. Make sure you have obstacle avoidance turned off when you land.

Yep, OBS is off. The unusual behaviour seems to be worser the more sophisticated the selected flight mode is: WatchMeMode very bad, FollowMeMode bad, AngleMode somewhat bad.
I also have the impression that using other than rabbit speed modes are worsen the problem additionally because the bird really gets unresponsive.
I used to take off and land in a middle speed position (where the beep comes) for 200 flights: no problems...

 

[vr-pilot]

Thank you, Thank you!!!!! The behavior you describe is almost exactly how mine behaves after the last update. Landing as I described in my other post is almost a battle with it constantly moving around in all directions. You can't simply push the left stick down gently, it takes some throttle up as well to keep it descending smoothly and constantly making corrections to get it to land where I want....

More data points to show/give Yuneec when I send my H in for diagnostics and hopefully repair.

Is yours a YTH PRO with RS? Maybe the FW fixed some problems for early released Standard/Advanced (or Pro w/o RS) models and broke something for PROs with RS (in combination with IPS' ultra sonic sensor?) Just a guess...

 

[QuadBart]

No, mine is the standard. I just got off the phone with Yuneec. Sending it in for a health check...

Is yours a YTH PRO with RS? Maybe the FW fixed some problems for early released Standard/Advanced (or Pro w/o RS) models and broke something for PROs with RS (in combination with IPS' ultra sonic sensor?) Just a guess...

 

[dxer]

I have Typhoon H Pro version (EU version) and several days ago I upgraded firmware to the latest. Before it was flying without any problems. Finally yesterday was better weather with some wind so I decided to give it a try. After i upgraded I did gilbal and accelerometer calibration and yesterday before flight I performed compass calibration. First time id didn't go well because after some dance with H, it just switched OFF while I was doing it. Ok, I tried again and it went well this time. But after I was in the air I realized that something is different. First of all couple of times it didn't react on command at all. I mean, left/right commands. So I tried some faster flying backwards and after I stop to hover it was not stopped at this location but rather slides to the right (looking from behind) in a curved line and then hovered. There was a windy weather , I don't know if this is because of that, but never had such behaviour. The most scary thing is that it didn't react on command couple of times.
And one more thing , when switching RealSense ON, it is now flying more slowly than before.
I must say that after I did this upgrade , I didn't re-bind as it was all working fine. Now I re-bind just in case so I will today try to calibrate compass again and give it another try hopefully without wind problems.

Here are two last telemetry files if someone more proficient can say if there is something that looks wrong in there. Thanks

 

[Drone On]

I was just on the Yuneec, America, support/download page. There is no 2.25 drone firmware update. The only 2 drone updates on the page are are for the H Standard (USA) model:
Firmware Typhoon H STD Gimbal:v1.22, Autopilot:v1.28, Camera:v3.2.16
and For the H EU model:
Firmware Typhoon H EU Gimbal:v1.22, Autopilot:v1.28, Camera:v3.2.16 which states in the description Warning: THIS VERSION IS FOR EUROPE.
There is no other firmware for the H pro realsense for the USA models on the Yuneec USA website.
Can it be that people are installing the EU drone firmware in the U.S. H pro models by mistake that is causing the problem?