RE: [Paparazzi-devel] reply to: Re: a way to measure time? (antoinedrouin)

[gisela.noci]

As indicated before, the problem is not the computation of the climb/descend
rate. The problems is getting a sound (!) , repeatable sensor reading. If
every 10Hz sample were good, the job is 90% complete, but the reflections
are weak, lost during small A/C pitch and roll maneuvers, etc. If you get 5
or 6 sample out of 10, and your approach speed is , say 20m/s (ground speed)
then things happens fast, and the aircraft does not respond in tenths of a
second to rapid pitch commands, and so the whole process becomes touch and
go. Also, I use a different vertical guidance PID control loop, with very
different gains, for the landing, to get the aircraft as 'tight' as
possible, ie, generally not much integral gain, but more proportional and
quite high differential. We are not interested in long term gentle, stable
flight here, but very tight control down an approach flight line.

Have fun
Joe


-----Original Message-----
From: address@hidden
[mailto:address@hidden
] On Behalf Of Chris
Sent: Friday, November 13, 2009 12:24 AM
To: address@hidden
Subject: [Paparazzi-devel] reply to: Re: a way to measure time?
(antoinedrouin)

Hi Antoine.
I am trying to measure the rate of climb/descent by using the ultrasonic
range finder.
The analog value is coming from the sonar.
Since it is updated at 10Hz i thought that i could use the indicated height
which is (hopefully) more accurate than every other method for landing,
to compute the rate of climb or descent.
Chris




_______________________________________________
Paparazzi-devel mailing list
address@hidden
http://lists.nongnu.org/mailman/listinfo/paparazzi-devel