Hello,
I recently rebuilt my quad after looking at a few pictures on the vibration damping page. Unfortunately after all that effort my AltHold seem totally messed up.
Im not sure if the picture shows it so ill explain what I did:
1. The Yellow casing is a plastic box lid.
2. Level 0: On the plastics box lid, below the Purple acrylic plate is the RX. (Hidden from View)
3. Level 1: The APM 2.6 is mounted of 4 foam stilts on the Purple Acrylic Plate.
4. Level 1: The small chip on the side is the Darlington Array ULN2803A to control the LED's (works Amazing) - I have made sure nothing touchs the APM.
5. Leve 2: GPS & Compas.
After doing all this the AltHold seems to have got totally messed up.
Checked Vibration Again. Seems ok. See below:
Image below is BarAlt, WPAlt, RelAlt on Left Axis & CURR: ThrIn on Right Axis.
The BarometerAlt seems all over the place... and the Quad moves up and down nearly 3 - 5 Feet! One thing I noticed that is also reflected in the graph is that when I engage LOITER it work very well holding altitude for a little time, then starts to drop and I need to compensate with throttle.
I was wondering whether the way I have powered the APM could be causing a problem. I do not have a power module. I power the the APM with ONE ESC Power Wire. The other 3 have the power wire clipped. I have the JP1 Jumper connected.
CTUN: ThrOn & CTUN: ThrOut on Left & CURR: VCC on Right
Note: I tried with the Darlington Array driving the LED's and without the LED's - Same behavior on Alt Hold.
Not Sure if this is giving any valuable info but the graph below is CTUN: ThrOut (GREEN) vs CURR: ThrOut.
I cannot plot any TLogs as I don't have the Wireless module. Whats strange is that when I did a very rudimentary temporary install the Alt Hold worked much much better. Still had the same issue but on a much smaller scale. The ALTHold is so bad now that I cant LOITER becomes dangerous after a few seconds.
Thanks to Peddles2Paddles I managed to get the GPS and Loiter working now the Alt Hold is messed up. One step forward and 2 steps backward :-(
Any help will be much appreciated.
Thanks Arshish
Replies
Thank you So much for replying. I really appreciate the help. In an attempt to improve the VCC I have now switched over to a Castle Creations 10Amp BEC at 5V connected to the input rail. The VCC is a lot more consistent as a result of this however as I am not using the Power Module any more I have lost Volt & Current sensing until I get the AttoPilot.
I have attached the Logs & some graphs I created. Again .. Thank you for your help...
Vibrations
GPS
ALT
ALT + Throttle
VCC
Logs Also Attached below... I get a strong feeling that it is the wind gusts because on a calm day I get pretty good ALTHOLD performance. Im am positive that the wind does not reach the barometer as its covered with Foam and the whole APM unit is in a box. I think its just that the APM cannot keep woth the wind.... Perhaps some ALTHOLD PID tuning will help?
2014-06-14 16-30-44.log
A Video showing Barometer variations when APM is stationary on ground
https://www.youtube.com/watch?v=G7k1boiNa4w&feature=youtu.be
Barometer issue unfortunately is still not solved.....
Flew it today in slightly windy conditions and AltHold was all over the place again.
Came indoors and ran the Baro test with the quad placed on the table and the value was fluctuating between -0.3 to 1.2M. That's sitting stationary on the table.
Somone please please help me out with this. Ive painted the case black and put a block of foam abd the whole APM is enclosed in a box! Vibrations are OK. I just dont know what else to do :-(
That doesn't seem that bad. Do you have a log of the recent flight?
Another 2 min flight in Windy conditions in ALTHold. Throttle was constant.
Ok.. made the following changes:
1. Got the power module and fitted it. VCC is pretty locked in at 5v compared to 4.5v without the module.
2. Put a small block of white foam to cover the baromtere.
The Alt hold improved considerably for hover with these changes but was checked at night. Need to check in bright sunlight now.
I have 3 questions...
With these 2 changes the AltHold is much much better for hover. However for manouvers it still cant hold the altitude. For example when I engager autotune with alt hold enabled it kept dropping and I had to keep giving throttle to maintain the height. Any suggestions on how to improve this??
Also regarding vibrations, mine are at +-1 on all axis in a hover. But on checking logs for the time the autotune was working they are much heigher. Closer to +-3 with spikes even bigger. What is the recommended state to check vibrations?? Hover or aggressive flying??
Lastly after connecting the PM I now see the Volts Amps and bat% on the mission planner screen. Volts show correctly, amps are always 0.3, 0.4 but thats expected because im plugged into the usb and motors off.But the % battery is always 99%. Even on a drained battery? How do I fix this?
No replies...
Ok, Ive managed to sort out the AltHold Problem. The foam piece seems to have fixed it.
However, can someone please advise one how i should fix the battery status: The battery Status % is always 99%. even on a completely drained battery....
I had something similar: http://diydrones.com/forum/topics/quad-descends-in-loiter-but-not-i...
My Z-vibration was pretty bad using the blocks of foam that came with the Pixhawk, but X & Y were fine.
Eventually I just used small squares of velcro and the problem went away.
Thanks for the reply Graham.
What were your vibrations like on the z axis. If you look at the pic above my z vibrations are same as x and y.. and it appears pretty low.
-
1
-
2
of 2 Next