Flytrex Live - Voltage

[Hughie]

I tried out the Flytrex 3G yesterday it seems to work well and mission was updated to site fine, last seen position was correct too.

Nice to see two or three new metrics being logged over the Core 2, one of which is battery level.

I am a bit curious about why with a perfectly healthy battery (which gradually degraded normally from 12.5v to 11v during the flight) was logged at 0.21 v on four occaisions. The quad was not being thrown about much, which makes me doubt whether the battery pins are the cause. There were no power problems observed or heard on the quad it flew perfectly.

Is this a bug in the logging or something even more sinister ?

 

[p fandango]

Re: Flytrex Live Voltage

i've had a quick look through the first 5-6 of my missions & they've got atleast 1 spike, some alot more. But as you say i've never noticed it physically having any issues

 

[Hughie]

Re: Flytrex Live Voltage

i've had a quick look through the first 5-6 of my missions & they've got atleast 1 spike, some alot more. But as you say i've never noticed it physically having any issues

Thanks - that is weird though isn't it :?

 

[Fyod]

Re: Flytrex Live Voltage

Seems like a bad sensor reading, ie. the voltage is measured too often and the data either isn't read or gets corrupted.
They should take 3 readings per second and only send the highest to rule out errors.

 

[p fandango]

Re: Flytrex Live Voltage

Thanks - that is weird though isn't it :?

i never look at the graph & didn't even notice the spikes till you pointed them out lol

 

[Hughie]

Re: Flytrex Live Voltage

Thanks - that is weird though isn't it :?

i never look at the graph & didn't even notice the spikes till you pointed them out lol

Of course it is not just the graph, it is there to be seen in the csv. In my case each was 0.21v.

 

[p fandango]

Re: Flytrex Live Voltage

Of course it is not just the graph, it is there to be seen in the csv. In my case each was 0.21v.

i only tend to put the kml file of the flight into GoogleEarth, which doesn't show that info

 

[ianwood]

This seems odd. Are these Flytrex using CAN data? This should be reported to Flytrex.

 

[Hughie]

This seems odd. Are these Flytrex using CAN data? This should be reported to Flytrex.

Yes CAN data Ian. I have PM'd Amit at Flytrex and sent him a link to this thread. I decided not to use the Flytrex support because at this time they seemed to be snowed under. My last support request took 4 or 5 days to get answered. However, if they dont get the opportunity to pick the thread up soon I will contact them via the support channel.

I dont know enough about how the data is originally acquired by NAZA, or how the secondary acquisition by Flytex is done to be able to take a view on where in the data path the problem might be being caused.

Edit : I have now sent their support a message via their website.

 

[Sandman1962]

For Amit:
I have the same issue with the voltage graph. I'm using he new can-bus cable on
A phantom 2 vision +

Thanks

 

[p fandango]

i'm canbus on a P2 non-Vision

 

[ianwood]

I dont know enough about how the data is originally acquired by NAZA, or how the secondary acquisition by Flytex is done to be able to take a view on where in the data path the problem might be being caused.

I'd guess they've got a bug in their CAN decoder. Battery and individual cell voltages are contained in the 0910 payload which contains a lot of data and is 196 bytes long. It's often in non-consecutive packets and can be truncated from time to time.

(blue areas represent 0910 data relevant to the Phantom 2)

 

[amitregev]

Hi guys,

Thanks for the report on this! Since the Live 3G is no longer soldered directly to the main distribution board when using the CAN-BUS, the voltage reading is done slightly differently, so it is possible that our algorithms needs a bit more tweaking and optimization. We didn't see this behavior in our local testing (using a P2V+ as well), but should be fairly simple to fix.

We will discuss this more closely tomorrow morning (late night for us ATM), and we'll get this fixed quickly.

If you could post a link to the mission showing this issue on your Flytrex profile this will be great help -- thanks!

At the same time, we continue updating and refreshing our services and just pushed a new update to the Live Flight Channel on our site. Featuring new and overhauled design with much improved telemetry display and new Flight Mode indicator display - available for those flying with the CAN-BUS adapter. Flight mode will show both on the bottom telemetry panel and as colored flight path.

Many thanks for the report and help investigate this!

 

[p fandango]

everyone of my missions i've picked so far has had a voltage spike, but this one looks really scary http://www.flytrex.com/mission/quadcopt ... -AbbfFvwG/

 

[Hughie]

I dont know enough about how the data is originally acquired by NAZA, or how the secondary acquisition by Flytex is done to be able to take a view on where in the data path the problem might be being caused.

I'd guess they've got a bug in their CAN decoder. Battery and individual cell voltages are contained in the 0910 payload which contains a lot of data and is 196 bytes long. It's often in non-consecutive packets and can be truncated from time to time.

(blue areas represent 0910 data relevant to the Phantom 2)

Really interesting Ian. Thanks for posting this.

 

[Hughie]

Amit here is the link to my mission as shown earlier in the thread.
http://www.flytrex.com/mission/quadcopt ... -4ldUWCY3/

 

[amitregev]

everyone of my missions i've picked so far has had a voltage spike, but this one looks really scary http://www.flytrex.com/mission/quadcopt ... -AbbfFvwG/

Hi P fandango

Specifically with this mission, are you referring to the 3 (+1 at the end) peaks in the voltage readings that are showing in the graph?

Looking at each of these peaks, it seems that during each of the peaks your altitude have been pretty much similar to take off altitude and the aircraft barely moved. Is it possible that you landed several times during the flight and then took off each time? And that the voltage increases that are visible in the graph are actually the times you have landed on the ground for a few seconds each time?

Let me know if this might match the actual flight pattern. Thanks!

 

[amitregev]

Amit here is the link to my mission as shown earlier in the thread.
http://www.flytrex.com/mission/quadcopt ... -4ldUWCY3/

Thanks Hughie! We'll review this tomorrow and should get this fixed quickly.

The behavior in this mission though seems different from the one posted by P fandango in the post above. In his case I actually believe that the reading are correct (matching slow down in the engines) -- waiting for his feedback on this to see if makes sense.

Thanks!

 

[p fandango]

Hi P fandango

Specifically with this mission, are you referring to the 3 (+1 at the end) peaks in the voltage readings that are showing in the graph?

Looking at each of these peaks, it seems that during each of the peaks your altitude have been pretty much similar to take off altitude and the aircraft barely moved. Is it possible that you landed several times during the flight and then took off each time? And that the voltage increases that are visible in the graph are actually the times you have landed on the ground for a few seconds each time?

Let me know if this might match the actual flight pattern. Thanks!

i can't honestly recall what i did on that day, but as you say it does look alot different to the spikes that have appeared in alot of the other missions (as in Hughies). So can only assume your right on that sir

 

[Sandman1962]

Mine:

http://www.flytrex.com/mission/quadcopt ... -LJJOVRlt/

http://www.flytrex.com/mission/quadcopt ... -a4I3HDdX/