Shop our Airplanes Products Drone Products Sales
Thread Tools
Nov 23, 2012, 03:28 PM
Just another user
Dennis Frie's Avatar
As you have noticed, when using 100 % gyro weight, the gyro is pretty accurate on it's own. The difficult part is the magnetometer that's pretty damn sensitive to noise, current from wires etc. Unfortunatly some drift will always occur on the gyro, which have to be accounted for. 99% gyro-weight should be fine.

Pick whatever works best for you, but be aware that 0% = only magnetometer and 100% = only gyro. For tilt/roll 0 % = only accelerometer.
Sign up now
to remove ads between posts
Nov 23, 2012, 03:28 PM
Oxford Panic
AndyOne's Avatar
OK in a nutshell, here's what I learnt the hard way.

1 Most importantly, the sensor board MUST be located component side up and not as in the magnetometer cal screen.

2 On the GIU screen click on "show" menu and select advanced settings. in the top right hand corner some adjustment of the "gyro weight on pan %" may be necessary, 99% seems best.

3 In the accelerometer cal screen click on "help" menu and select help/how to, to see exactly how to position the sensor for calibration.

4 Regular resetting of the Arduino by pressing the on-board reset button seems to help

5 Select disconnect when exiting a calibration screen to release the serial link so it can be used by another cal screen.

Andy.
Nov 23, 2012, 03:34 PM
Just another user
Dennis Frie's Avatar
Quote:
Originally Posted by AndyOne
OK in a nutshell, here's what I learnt the hard way.

1 Most importantly, the sensor board MUST be located component side up and not as in the magnetometer cal screen.

2 On the GIU screen click on "show" menu and select advanced settings. in the top right hand corner some adjustment of the "gyro weight on pan %" may be necessary, 99% seems best.

3 In the accelerometer cal screen click on "help" menu and select help/how to, to see exactly how to position the sensor for calibration.

4 Regular resetting of the Arduino by pressing the on-board reset button seems to help

5 Select disconnect when exiting a calibration screen to release the serial link so it can be used by another cal screen.

Andy.
Always good to know to obvious things I haven't documented good enough... Documenting is not really my strong side.

Reset shouldn't be necessary - at least quite a few people have been able to make it work pretty good without drift.

Anyway, if you ever need to reset center, it's a lot better to connect a button between to be used as "set center".

The pin is defined in config:
// Button input:
#define BUTTON_INPUT 11

Just connect a button between the pin defined (in this case 11)) and ground. Will make it center immediately with no start-up delay
Nov 23, 2012, 06:34 PM
Registered User
Thank you so much! In few last posts I found almost all answers for my quetions and misunderstandings.
Last one - components up during mag calibration?
Nov 23, 2012, 07:48 PM
Registered User
subscribed
Nov 24, 2012, 12:47 AM
Registered User
P.I.Engineer's Avatar
did mine up today, that was super easy.

thanks for all the work.
Nov 24, 2012, 04:43 AM
Oxford Panic
AndyOne's Avatar
I wish mine had been easy, I did learn a lot in the process though.

A.
Nov 25, 2012, 12:09 AM
Registered User
P.I.Engineer's Avatar
check that, did some more testing.

my tilt works perfect, my pan, not so much.

when i pan, it pans but then moves back to center on it's own.

It shows the same thing happening on the plot.

it appears from the mag cal .exe that all the sensors are outputting correctly. So now i'm confused

Edit: just searched this thread, i'm not the only one with this issue.
Last edited by P.I.Engineer; Nov 25, 2012 at 12:25 AM.
Nov 25, 2012, 02:48 AM
Registered User
Quote:
Originally Posted by jrod155
check that, did some more testing.

my tilt works perfect, my pan, not so much.

.
Did you calibrate mag as shown in pictures?
try to calibrate it inverted, first 2 pictures componets up.
Nov 25, 2012, 04:17 AM
Oxford Panic
AndyOne's Avatar
Drifting back to centre is what happens if the board is upside-down.

A.
Nov 25, 2012, 05:49 AM
Registered User
Gabek's Avatar
Quote:
Originally Posted by AndyOne
Drifting back to centre is what happens if the board is upside-down.

A.

Or if the endpoint is too high!
Nov 25, 2012, 09:52 AM
Registered User
P.I.Engineer's Avatar
which way is up?

sensor board up or usb up?
Nov 25, 2012, 10:05 AM
LKJ
LKJ
Lars k Jensen
Quote:
Originally Posted by jrod155
which way is up?

sensor board up or usb up?
USB is on the controller board and it doesn't care, its is the sensor board vho need to have the component side up.

Lars.
Nov 25, 2012, 10:06 PM
Registered User
P.I.Engineer's Avatar
yeah, that's the way i have it, still drifting.

I'll mess with it more this week.
Nov 25, 2012, 11:29 PM
vpb
vpb
Registered User
Has anyone try Dennis Frie code on crius se board?


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Build Log Diy osd (Arduino and opensource) Dennis Frie FPV Equipment 2506 Feb 24, 2017 09:00 AM
Discussion PPM to PWM circuit for headtracker 38cjc38 DIY Electronics 18 Mar 04, 2011 06:43 AM
Question Adding additional PPM inputs to a headtracker/module setup. SebastianJ FPV Talk 0 Aug 19, 2009 07:38 AM
Question Adding additional PPM inputs to a headtracker. SebastianJ DIY Electronics 0 Aug 18, 2009 03:37 PM