Thread Tools
Jul 01, 2011, 07:17 AM
Gaftopher
Gary Mortimer's Avatar
Because its not written, if you would like to you can. As Jeremy says we are not at a solid release version yet.
Sign up now
to remove ads between posts
Jul 01, 2011, 07:37 AM
Registered User
Quote:
Originally Posted by Joecnc2006
The OpenPilot TV Episode 2 - PID Tuning, was just posted, this is very informative for people who want to dial in their MultiCopters.

http://www.openpilot.org/pid-tuning/


http://vimeo.com/25839978
I really like the last section, for me it really gives a good visualisation of the effect of the parameters. The blue spectrogram thing may give colour blind people a few issues though
Jul 01, 2011, 10:22 AM
Registered User
Quote:
Originally Posted by Joshbb
I really like the last section, for me it really gives a good visualisation of the effect of the parameters. The blue spectrogram thing may give colour blind people a few issues though
Yeah at the bottom of this page http://www.openpilot.org/pid-tuning/ I plot them a different way. You're right, it could even be gray scale. I just can't bring myself to redo the presentation to fix it though.
Jul 01, 2011, 01:52 PM
Suspended Account
Quote:
Originally Posted by peabody124
Yeah at the bottom of this page http://www.openpilot.org/pid-tuning/ I plot them a different way. You're right, it could even be gray scale. I just can't bring myself to redo the presentation to fix it though.
i thought the video was great but i think you are confusing oscillating with motor hunting. when your rate is to high the motors make the revving sound like there fighting each other thats not oscillating. oscillating is like hanging the quad from a string like a tire on a rope it kinda does some circles without yaw spinning. just thought i mention that, great vid though!
Jul 01, 2011, 02:36 PM
Registered User
Quote:
Originally Posted by Quad_Addict
i thought the video was great but i think you are confusing oscillating with motor hunting. when your rate is to high the motors make the revving sound like there fighting each other thats not oscillating. oscillating is like hanging the quad from a string like a tire on a rope it kinda does some circles without yaw spinning. just thought i mention that, great vid though!
I would consider that a semantic rather than real distinction. If I put it in manual mode and run the motors at that rate you don't hear that sound so it's not the ESC itself hunting. It is the outcome of my control loops, and specifically the interaction of them with the feedback from the gyros combined with some delays.

This is not hunting in the sense of the ESC trying to maintain an RPM if you were using a governor mode. You might get that though if the ESC is running a feedback loop though. In fact you can easily model this (without putting any control loops on the ESC for what I would call actual motor hunting).



So the top plot shows the spectrum (i.e. amount of energy at each frequency) for a quad that is undertuned. You can see it decays fairly smoothly when exposed to noise. The next plot simulates what happens
if the quad is overtuned and you can see there is a shoulder out to 2 Hz which is pretty close to the measured frequencies. The bottom plot shows the response to the overtuned quad to a little bump and you can see this rings too.

Anyway, I'm certainly not excluding motor hunting (i.e. ESC trying to maintain RPM) as a relevant factor, but I don't believe that's what happens when you increase the gains too high as you can easily explain it without that.

If you meant something else by motor hunting I apologize for the long winded reply that is off topic
Jul 01, 2011, 02:45 PM
RC Addict
anatoly's Avatar
Nice video, very informative. I guess it helps to have a bluetooth adapter hooked up, so you don't have to keep reconnecting the usb cable each time you want to change a parameter.

So if I understand correctly, axis lock is like heading hold when applied to yaw, but you can also apply it to pitch and roll so it will try to maintain that angle?
Jul 01, 2011, 04:40 PM
Suspended Account
Quote:
Originally Posted by peabody124
I would consider that a semantic rather than real distinction. If I put it in manual mode and run the motors at that rate you don't hear that sound so it's not the ESC itself hunting. It is the outcome of my control loops, and specifically the interaction of them with the feedback from the gyros combined with some delays.

This is not hunting in the sense of the ESC trying to maintain an RPM if you were using a governor mode. You might get that though if the ESC is running a feedback loop though. In fact you can easily model this (without putting any control loops on the ESC for what I would call actual motor hunting).



So the top plot shows the spectrum (i.e. amount of energy at each frequency) for a quad that is undertuned. You can see it decays fairly smoothly when exposed to noise. The next plot simulates what happens
if the quad is overtuned and you can see there is a shoulder out to 2 Hz which is pretty close to the measured frequencies. The bottom plot shows the response to the overtuned quad to a little bump and you can see this rings too.

Anyway, I'm certainly not excluding motor hunting (i.e. ESC trying to maintain RPM) as a relevant factor, but I don't believe that's what happens when you increase the gains too high as you can easily explain it without that.

If you meant something else by motor hunting I apologize for the long winded reply that is off topic
u had me at hello lol...

well when your hearing the motors in Layman terms its motors fighting each other to maintain level. oscillation usually is like a lazy or delay rate. but i do understand "partially" what you mean lol. very cool video! and it makes me want the bluetooth yesterday. i just bought one for my MK
Jul 01, 2011, 07:09 PM
Registered User
Quote:
Originally Posted by Quad_Addict
u had me at hello lol...

well when your hearing the motors in Layman terms its motors fighting each other to maintain level. oscillation usually is like a lazy or delay rate. but i do understand "partially" what you mean lol. very cool video! and it makes me want the bluetooth yesterday. i just bought one for my MK
Heh. I thought it was probably just using words differently. You'd said something similar on the OP forums that puzzled me at the time but now makes sense. I'm relatively new to quads so just use oscillations in the sense of st that goes back and forth
Jul 01, 2011, 07:19 PM
Suspended Account
Quote:
Originally Posted by peabody124
Heh. I thought it was probably just using words differently. You'd said something similar on the OP forums that puzzled me at the time but now makes sense. I'm relatively new to quads so just use oscillations in the sense of st that goes back and forth
yes im ok with that terminology lol
Jul 01, 2011, 07:39 PM
Suspended Account
i guess we really dont need vibration dampeners with cc eh?
just did a video because i ran out of rubber standoffs from building some other junky quads lol so i decided to try just regular nylon standoffs.
-
http://www.youtube.com/watch?v=ZWYzcd4CPwY
Jul 01, 2011, 09:18 PM
'FPV'er...not a "LOS'er
Vantasstic's Avatar
Woo-Hoo. I got in my replacement ESCs for my T-copter and CC board. I changed over from 1000kv to 930kv motors with the 10x4.5 props. After getting calibrating the ESCs and checking the wiring I took the little T-copter out in the back yard. Wow...nothing could have been more easy. The thing was nice and smooth...with no vibration isolators on the CC board using aluminum standoffs. The stock settings are pretty slow reacting and tame. I guess I'll be cutting my teeth on PID and Angle Limits. I'd my yaw is a little slow and mushy...I'll try tweaking that up more first. In all though, I'm quite happy. Oh, I only hovered in auto-level mode. I'll try rate(?) mode where I have a little more room to fly other than my back yard.
Jul 01, 2011, 09:40 PM
MultiRotor Jedi
simonjardine's Avatar
You will need rubber stand offs in outdoor windy conditions.

Have been playing PID Algorithms too, my settings are totally different yet the Quad on the video looks to be the same size as mine...

Also where can i get a plug/cable so i can use Xbee?

Simon
Jul 01, 2011, 09:44 PM
www.joescnc.com
Joecnc2006's Avatar
Quote:
Originally Posted by simonjardine
You will need rubber stand offs in outdoor windy conditions.

Have been playing PID Algorithms too, my settings are totally different yet the Quad on the video looks to be the same size as mine...

Also where can i get a plug/cable so i can use Xbee?

Simon
the cable is in the OP stores, you cut it in half and solder to your Xbee. as per pinouts in the wiki.

http://wiki.openpilot.org/display/Do...lemetry+Modems

Connecting your XBee Modem to your Openpilot or Coptercontrol board

The simplest way to connect your XBee modem to your OP or CC board is via a carrier board that provides a socket for the XBee and a voltage regulator and IO breakouts.

You'll need to make your own connector to match whichever board you're planning to connect this to. This will require some (very simple) soldering to the carrier board's vias.
The below is an example pinout from the CC to the Sparkfun Xbee explorer regulated

CC ----------- Sparkfun
Gnd ----------- Gnd
5v ----------- 5v
Tx ---------- Din
Rx ----------- Dout
Jul 01, 2011, 09:49 PM
Suspended Account
Quote:
Originally Posted by simonjardine
You will need rubber stand offs in outdoor windy conditions.

Have been playing PID Algorithms too, my settings are totally different yet the Quad on the video looks to be the same size as mine...

Also where can i get a plug/cable so i can use Xbee?

Simon
tried it in breezy winds still good. like i said i only used it without because i used my other rubber standoffs, but i have 2 sets coming, i would still rather have them on. and yes its exactly like yours and should be everything is the exact same

BTW you never did mention what was the fault of your 30 second issue. was it RX?
Jul 01, 2011, 09:55 PM
Registered User
AlPackin's Avatar
Quote:
Originally Posted by Vantasstic

Okay, found this: http://wiki.openpilot.org/download/a...cccopter03.jpg Not the best picture, but you could probably map out the connections via the picture.
I can't even see the receiver in that pic much less the connections
I don't mind soldering direct, is there some more info on how?
Ed


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion The Aussie RCG Coffee thread (aka "we don't drink instant") glydr Australia 164 Sep 08, 2012 03:33 AM
Discussion NEW? -Openpilot flight controller tc3wins Multirotor Drone Talk 229 Mar 02, 2012 05:14 PM
Discussion Share your art! PinkYakRC Painting and Finishing Tips 13 Apr 27, 2011 07:04 PM
Discussion RCG 07 JR Indoor Live Coverage Thread Jim T. Graham JR Indoor Electric Festival 19 Nov 04, 2007 01:36 PM
Discussion My first RCG flame thread WBE Chit Chat 6 Sep 04, 2007 09:17 PM