Yes, agree with David. The only difference you might notice is the RC in options are now across all channels. All of your settings will transfer over when you update. Go for it.
1 Like
Just took my rover out for a trial in a car park down the road from me. It enabled me to use the long white lines in the car park to plot a waypoint straight down the centre line & see how tight it keeps centered.
To be fair the rover did a good job but at some points it was showing “Error compass variance” and also the EKF was showing red for “Compass”
Not sure why this is but it seemed to happen allot. GPS for both modules was great, showing combined satellites detected of around 34 satellites & HDOP at 0.50 which is also really impressive. Just a shame about the errors showing up with compass variance & EKF on the HUD which seemed to put the rover slightly off the line
Disable compass 2 (and 3 if there is one).
Already did that right from the start of setting up the external compass. Some people have said on other forums to unplug the i2C connection for the gps. Not sure if that would help tho.
No difference if they are disabled anyway. This means there is an error between the compass and the flight controller gyro. It can be caused by a misaligned external compass, wrong orientation setting or a bad calibration. How did you do the calibration?
1 Like
It’s probably compass interference possibly from metal in the ground. A calibration alone is actually unlikely to resolve the problem. Moving the external GPS/compass unit higher might help.
Turning off all the compasses probably won’t work well because then the vehicle won’t know it’s starting direction. Feel free to try it of course and if you do, please report back!
2 Likes
Randy took my theory. If the surface of the car park is concrete, it will have a matrix of steel rebar which could give a compass a fit.
1 Like
@dkemxr the compass was calibrated as follows. I held the entire rover with everything in its place and rotated it like in this video:
When I do this two boxes that show the colourful sphere rotate. The 1st compass shows reading in the minus for all 3 axis. They are all -100 to -300. For the 2nd compass the readings are high, in the 1,000s which I thought was odd but I have tried calibrating a dozen times and each time it shows the same result.
I even added magnets at one time to try to fix the high offset but the same results appeared:
I have positioned both external compasses at the front of the rover to avoid any interference. Wether it’s to the front of the rover or making them elevated upwards would that make a difference? I’m thinking it’s the pixhawk that’s showing the high readings not the external compasses. Here’s a pic of the rover & shows the position of the external compasses:
This has been also happening on grass & has always shown up at some points whilst the rover has been in my garden away from any metal in the ground or sources of electricity apart from the rover. Not sure if that’s the issue & think it’s more of an internal problem I’m having.
I appreciate all your feedback & hopefully it’s a simple fix, either that or I might have to buy another pixhawk if it’s potentially faulty
this is a live calibration, its not as accurate as the on board calibration, try a on-board one
1 Like
Thank you will try this again. When I use on board calibration the top compass completes & the second one resets from the beginning and keeps on resetting.
Maybe I need to use the relaxed setting and it might work then
1 Like
If you have it disabled it won’t matter but yes, you can use relax and it might pass. If the offsets are very high that compass will likely give you trouble anyway if it’s active.
You mentioned that you were seeing the same compass error during missions in you backyard during turns. Is that right or only in the parking lot?
Magnets? Ha, what a bad idea. Sounds like a YourTube find.
Space the GPS module with the active compass up with a mount so it’s as high as it will go in that box. This has already been suggested. You may as well elevate both of them.
After doing several of the same missions in a row to determine course repeatability, which was excellent, I think I need an enclosure like yours for the electronics Mus. I’m driving thru dense snow, which the crawler just plows right thru, and it’s getting wet inside. All of the stock Crawler components are waterproof but all the add ons are not. This will take some doing.
This is the one I purchased & I’ve used it in the rain, had no problems:
Whitefurze https://www.amazon.co.uk/dp/B01GS5P2NA?ref=ppx_pop_mob_ap_share
Yes your right I was seeing the same error whilst in my garden. Here’s a silly question but I’m not sure about this. The compass for direction/heading is the pixhawk right?
And the external GPS modules are just for location not for which way the rover points.
When I unplug the i2C cable the rovers position in which way it’s pointing & what it shows on MP is spot on 100% accurate. When I plug it back in the direction on MP moves slightly off. Why would that be happening?
Also I will do what you’ve all suggested & elevate the external gps modules & do an on board calibration. Hopefully that will help
The compass for direction is dependent on your parameters. it depend which ones are being used.
If there is an external compass it will always be assigned as Compass 1. If you unplug it then the Pixhawk internal compass will be assigned as Compass 1. You can see this on the Compass Device ID’s. As far as I know the compass offsets do not follow the compass device but which one is assigned as 1,2, or 3 (could be wrong here) so if you were to disconnect the external compass you would need to do another compass calibration.
In my experience with several different Flight Controllers with internal compass’s, except for one used on a Plane, the internal compasses have been trouble. Certainly that’s been true on multirotors and rovers.
So it’s possible to use the external gps module for direction as well as location. As far as I’m aware everything else is disabled apart from my external gps module but when I pick it up and move it the Rover icon on MP doesn’t follow. When I pick up the Pixhawk it moves accordingly. So right now the pix is the compass and GPS is being used just for GPS positioning. I’d like the external GPS module to do both direction & location
I don’t think my GPS module has a compass🤦🏻♂️
Or am I being completely thick as s**t here and the gps is a compass as standard?
It has a compass and again if it’s connected to the I2C port it will be used as the primary compass. Moving the GPS/compass module around on it’s own is meaningless. This has been discussed before. The FC expects to see Z-axis gyro movement when the compass is rotated and when it doesn’t there is something wrong.
Juts raise the modules as high as you can in the box and do an On Board Calibration. Had we known before yesterday you were monkeying around with the old Live Calibration (get that from YouTube?) there would have been comment.
I know you hate YouTube😂 I will follow your instructions and report back with hopefully some good news. Thank you for explaining what I need to do👍🏼
When the external gps module is connected to the i2C port and when I pick up the compass it doesn’t move regarding the direction on MP.
So something isn’t right.
I don’t hate YouTube it’s just there is so much bad information on there.
So you are just rotating the GPS/compass module by itself, the Pixhawk is not rotating with it? If so read again what I posted above. What you may be missing is the fact that the directional display in the MP HUD is not a direct indicator of compass output but a fusion of sensors that includes the compass if it’s/they are active. That’s why with no compass active (uncheck ALL the boxes and try it) it will still show directional change when the Flight Controller is rotated.