Welp, fell out of the sky this morning.

[Ty Pilot]

This brings up the next question for those that lost a Plus - how hot was it when it happened? This might help rule - in or out - heat.

 

[Eagle's Eye Video]

At this point @CraigCam is in one of the hottest locations over the last couple of months in Tucson AZ... but I know he flies early/late for that very reason in the summer.

 

[Ty Pilot]

Yeah I have been encouraged by seeing his progress. Until the last couple of days I thought all was looking pretty good.

 

[Eagle's Eye Video]

mebbe it's because they're not using them new-fangled $500 batteries...

 

[Wingnutz]

This is from the 520 forum. Thought I’d check there as I’ve not read anything in awhile. Interestingly enough, this is from post #9. It indicates there could be an issue with the IMU. Not sure if this is relevant to this thread, but IMO it’s something else to consider. Not that heat couldn’t be the issue, because it certainly could.

H520 fell from skies above!!

“The report said: H520 telemetry data has been checked and we can identify there was an issue with the IMU board which has caused the mid flight failure.”

 

[PatR]

Most of our hobby stuff comes with temp limit ranging from +5 or 10C to around +40C. I can tell ya that military stuff specs a minimum of -10C and often -20C to +60C because of electrical component degradation and failures. A lot of those failures occur with the aircraft sitting on the ground. I’ve personally measured heat being reflected off the ground, asphalt, and steel decks that exceeded 140*F.

 

[glider]

4.68v at the flight controller is death for a PX4/Pixhawk based unit.

Being on my phone it's difficult to read the log, but seeing a low V warning for avionics is bad news.

I wonder if the power module failed. Even intermittent voltage drops will at minimum cause brown outs. There doesn't need to be a complete power loss.

 

[h-elsner]

Sorry, I slept all the time (night time in Europe) and missed a lot of discussings. The problem is that H Plus with current SW will not fill the Error Flags. But, there are a lot of useless and wrong datasets inside the telemetry files.
I tried to supress most of them to get a proper analysis and useful histograms. In Settings > Data analysis > TH+ clean option shall be checked.

The list of error flags above is not real. It contains also Yuneecs magic number 85 (which is 0101 0101 binary ).

br HE

 

[mbernholdt]

4.68v at the flight controller is death for a PX4/Pixhawk based unit.

Being on my phone it's difficult to read the log, but seeing a low V warning for avionics is bad news.

I wonder if the power module failed. Even intermittent voltage drops will at minimum cause brown outs. There doesn't need to be a complete power loss.

I have noticed this on motor start, but not in flight. But googling the error tells me it's an common error with the PX4 on all platforms.

 

[oliver]

I have been hot swapping, and while I have not experienced the same as @Wingnutz I have see some behavior that I believe to be from heat. I have checked body and motor temperatures during operation and in flight, using an infrared laser thermometer and have found that upon landing, as the hot components sit; the aircraft body and surrounding components heat up an additional 20 degrees (F) or so. Upon lift off I am seeing the aircraft in a slight climb but this goes away within a minute of operation and seems to correlate with the temps going back down. I have thought this was due to heat as I have read that the barometers can be effected by heat. Just a thought.




Here is a screen shot of the error flags from @toddinroch fatal flight. I see there are temperature warnings in there as well.

View attachment 11026

I also have this type of error message on flights with more or less of line, I do not know now if it is necessary to worry about it

 

[biltno]

Me too, after reading through these threads I have decided to ground my Plus pending an understanding of what is going on and what we are dealing with. I have looked back through my history notes and am seeing some of the same messages; Caution: avionics power low, Manual Control the Interruption(?) and others that don't look right.

Sounds like an excellent idea to me. These cases are very similar

 

[oliver]

I correct my previous message.
If I open fligh2tlog with version 3.5 of Q500log2kml, I have plenty of errorflag

If I open the same flight with version 3.8, I have no more errors

What do you think? did I miss something

 

[Ty Pilot]

Maybe, maybe not. Are you not seeing the messages in "History" on the ST-16s @oliver ?

 

[oliver]

Maybe, maybe not. Are you not seeing the messages in "History" on the ST-16s @oliver ?

the messages I have on the ST16 are the compass.
I just watch the other flights is the same. With version 3.5 I got scared. I even sent the full file to Yuneec Europe to get their opinion.
there are even errors when the engines are shut down.

 

[Ty Pilot]

Okay so, just to be sure of things, I wanted to see these 'history notes' and how they correlate to a flight so I cleared all the old ones out. I charged a battery and went into the backyard and performed a 12 minute flight with a landing and engine stop in between, so essentially it was two flights. From that came these notes.





So the 'Avionics Power Low' warning is happening right after engine start but before takeoff. It is interesting that the 1st take-off occurs in the first photo at the top (11:02:24 AM) followed by only one message - 'NOT DISARMING: Not in Manual Mode or Landed Yet' then the Landing nine minutes later at 11:11:38 at the bottom of photo 2. Then for the second part of the flight the notes seem to repeat.


Is it safe to assume this is normal? I will look into the telemetry and see what it looks like.

 

[PatR]

Sounds like a voltage sag generated by application of ESC/motor loading. As Glider mentioned it's tatamount to a "brown out" as it's what happened with high performance aerobatic aircraft using high draw digital servos with batteries that were too small in capactiy and C rating during high deflection maneuvers. This is exactly what happens when low C rated batteries are used.

 

[Ty Pilot]

Sorry, I slept all the time (night time in Europe) and missed a lot of discussings. The problem is that H Plus with current SW will not fill the Error Flags. But, there are a lot of useless and wrong datasets inside the telemetry files.
I tried to supress most of them to get a proper analysis and useful histograms. In Settings > Data analysis > TH+ clean option shall be checked.

The list of error flags above is not real. It contains also Yuneecs magic number 85 (which is 0101 0101 binary ).

br HE

So you are saying that screen shot I posted on the other page with all the error flags is not real because the software for the Plus is different and causes fake warnings - In the way the telemetry is read in your program but the warnings are not actually there and being seen by the Plus software? Well that makes things a little better because I just looked at the telemetry from the flight this afternoon and it has the same sort of warnings, but here again - in the notes, all looks normal. This is encouraging.

 

[Ty Pilot]

Sounds like a voltage sag generated by application of ESC/motor loading. As Glider mentioned it's tatamount to a "brown out" as it's what happened with high performance aerobatic aircraft using high draw digital servos with batteries that were too small in capactiy and C rating during high deflection maneuvers. This is exactly what happens when low C rated batteries are used.

Yep that is what I am thinking too, start six big motors at once and there should be a quick zap to the battery.

 

[h-elsner]

What do you think? did I miss something

V3.5 was before H Plus. H plus support was introduced with V3.7. With V3.8 the support of Sensor files was improved and AppLog was introduced to record the text messages from Sensor files (the same as History logs on ST16 I think).

Background: In the telemetry data from ST16S there are a lot of useless messages with chaotic values in. Starting with V3.7 I tried to suppress those values in order to allow a useful analysis of the rest of the data. The supression can switched on and off by check box "TH+ clean" in Settings > Data analysis. Default is 'on'.

With FW 1.0.08 the following columns are not supported and empty:
- roll
- yaw (always 90°)
- pitch
- gpsPosUsed
- errorFlags1
- gpsAccH

Current do not contain the current as expected but remaining LiPo capacity in %. Misleading header, but a good idea to report this. Why it is so hard to rename this, I do not unstand. On the other hand Yuneec removed '_' in the headlines.

Vehicle type is 5 as previous used for Typhoon H. Not a good idea.

It's no more possible to detect real flights by fMode. There is no indication for initializing as it was at Typhoon H (f_mode=16). I'm not sure, but it looks like also the task modes (FollowMe, CCC, POI and so on) are no more indicated by fMode.

Two new columns added: vSpeed and hSpeed. Very good.
cgpsStatus was added again (was already there at Q500). Useless, because it represents only the two highest bits from IMU status.

The numbering of sensor files is different from the other FlightLog file numbers. It can only be matched by time stamp of the files. RecordID from Typhoon H was $BC, now at Typhoon H Plus it is $FD.

Remark: It seems, the roll/yaw/pitch values will come with FW 1.0.10 but this one was withdrawn. Let's see what will be changed/improved next time.

br HE

 

[PatR]

Yep that is what I am thinking too, start six big motors at once and there should be a quick zap to the battery.

FYI, they have this problem with the 920 batteries too. They worked fine for the original 920 but their conversion to the 920 "+" increased current draw, and flight times fell off quite a lot for the Plus version using the same batteries. Changing to batteries with higher C ratings ended up providing a lot more flight time using batteries having 4000mA less total capacity. Low C rated batteries are less expensive than higher C rated batteries. Supply chain economics?