[phpBB Debug] PHP Warning: in file [ROOT]/includes/bbcode.php on line 112: preg_replace(): The /e modifier is no longer supported, use preg_replace_callback instead
[phpBB Debug] PHP Warning: in file [ROOT]/includes/bbcode.php on line 112: preg_replace(): The /e modifier is no longer supported, use preg_replace_callback instead
[phpBB Debug] PHP Warning: in file [ROOT]/includes/bbcode.php on line 112: preg_replace(): The /e modifier is no longer supported, use preg_replace_callback instead
AutoQuad Forum • View topic - PX4FLOW & position hold
Page 2 of 2

Re: PX4FLOW & position hold

PostPosted: Mon Jan 25, 2016 12:48 am
by Max
No the PWM definitions shouldn't need changing. If there is no motor set up on a port, it won't get initialized at all. COMM gets started up early in the boot process anyway and would grab the port first.

Have you tried it w/out Bluetooth, just a UART<->PC adapter? Not sure what else to suggest at the moment.

-Max

Re: PX4FLOW & position hold

PostPosted: Mon Jan 25, 2016 1:10 am
by Matthew
Thanks for the reply. My bad. I checked Rx <-> Tx connection 3 times but I guess it wasn't enough. Switching it after I made changes to the code, which you advised, finally made it work :D Tomorrow I'll check why I experience such a significant drift with PX4FLOW.

Re: PX4FLOW & position hold

PostPosted: Mon Jan 25, 2016 3:24 am
by Max
Great, thanks for confirming it works! That extra comm port should/will be an option in future firmware. Sorry can't help with the flow sensor setup though, never played with one.

-Max

Re: PX4FLOW & position hold

PostPosted: Mon Jan 25, 2016 12:50 pm
by sandmen

Re: PX4FLOW & position hold

PostPosted: Tue Jan 26, 2016 9:02 pm
by Matthew

Re: PX4FLOW & position hold

PostPosted: Tue Jan 26, 2016 9:37 pm
by Matthew
There is something very strange going on when I'm checking Mavlink Massege Inspector. First of all when I connect PX4FLOW directly to dedicated QGC everything looks fine.

OPTICAL_FLOW.flow_x changes accordingly to the real situation:
Image

So does OPTICAL_FLOW.flow_y
Image

And OPTICAL_FLOW.quality
Image

I expected to see similar values when I connected Autouad to QGC and opened Mavlink Massege Inspector. But OPTICAL_FLOW.flow_x is always, always 0. No matter what I do.
Image

OPTICAL_FLOW.flow_y looks better but it has a tendency to not go back to 0 when quality drops down. After deeper inspection I discovered that flow_y reflects movements in x direction. So flow_y here is in fact flow_x when PX4FLOW is directly connected.
Image

Quality is always 0 but sensor_id seems to change somehow similar to the true quality. Here is how sensor_id looks like:
Image

This looks so strange I don't know what to think. Should I flash the PX4FLOW with this very old version of software which Bill Nesbitt posted a long time a go?

The last thing I checked was to flash the GPS led when flowSumX value was bigger then 10. That was supposed to show whether or not flow_x really is 0 as diagram shows. And diode blinked. This probably means that everything is ok and just Mavlink Massege Inspector shows incorrect values?

Re: PX4FLOW & position hold

PostPosted: Wed Jan 27, 2016 8:08 am
by sandmen
Matthew,
what I posted was not a patch.
It should show you roughly, what you need to do....

I didn't know a other way, to get the Px4Flow working !

https://www.youtube.com/watch?v=vULx47b06Ho

Re: PX4FLOW & position hold

PostPosted: Sat Jan 30, 2016 10:26 pm
by Matthew