EDTracker Bundle (HCKITS0021, 21B, 21E, 21F, 21G, 21H & 21I)
-
- Posts: 4
- Joined: Mon Jan 11, 2021 11:17 pm
Re: EDTracker Bundle (HCKITS0021, 21B, 21E, 21F, 21G, 21H & 21I)
Sorry I was quite busy recently and didn't get a chance to mount it away from the head set. Mine is quite permanently mounted to a head set. I am waiting for a new Arduino to arrive then I will make another one and mount it on a head band away from any speakers.
From your picture I can see you have a really nice spheres without any distortion and your points are really evenly distributed. I have a few questions about your magnetometer calibration process:
1) at what position to you set the sensitivity slider when you do your calibration? I usually put it at about 60%. Does it really matter where you put the sensitivity slider?
2) how many points do you use? I usually to all 2000 points.
3) how do you end up with such a nice distribution? Is there a method or special way of moving the tracker to get this nice distribution? I have already watched some videos how it should be done. First I rotate it about all 3 axis 180 degrees each way, then I move it in a figure of 8 and in other random directions until I get 2000 points.
4) how fast do you rotate it during calibration? Is the slower the better or is speed irrelevant?
There must be something wrong about magnetic field or my calibration, because I now have 4 MPU9250 and they all show the same issue of head coming back a bit after rotation about z axis.
From your picture I can see you have a really nice spheres without any distortion and your points are really evenly distributed. I have a few questions about your magnetometer calibration process:
1) at what position to you set the sensitivity slider when you do your calibration? I usually put it at about 60%. Does it really matter where you put the sensitivity slider?
2) how many points do you use? I usually to all 2000 points.
3) how do you end up with such a nice distribution? Is there a method or special way of moving the tracker to get this nice distribution? I have already watched some videos how it should be done. First I rotate it about all 3 axis 180 degrees each way, then I move it in a figure of 8 and in other random directions until I get 2000 points.
4) how fast do you rotate it during calibration? Is the slower the better or is speed irrelevant?
There must be something wrong about magnetic field or my calibration, because I now have 4 MPU9250 and they all show the same issue of head coming back a bit after rotation about z axis.
-
- Posts: 1
- Joined: Fri Jan 29, 2021 2:52 pm
Re: EDTracker Bundle (HCKITS0021, 21B, 21E, 21F, 21G, 21H & 21I)
Sorry for the following ridiculously naive questions, I've just received my bundle today having only ever previously having had the pre made and sealed Ed-Tracker Pro which was plug and play.
1) What is the button for? - EdTracker never had one (visible) , Do I need to cut a hole in the casing for this?
2) After running a new calibration, will my existing ED Tracker software and drivers work with this? the GUI is older but have never had any issues with it and all curves are already set through OpenTrack, which I have set on auto-launch.
** Side note in relation to previous post - I had an issue with my EDtracker recently bumping at the extreme head motions as described in the previous comments, whilst I have no knowledge of internals or of this product etc this problem arose after 2-3 years of use, turns out my cable was damaged, swapped it out, re-ran calibration - problem stopped - Long shot/coincidence, but maybe worth checking.
Thanks for any help
G
1) What is the button for? - EdTracker never had one (visible) , Do I need to cut a hole in the casing for this?
2) After running a new calibration, will my existing ED Tracker software and drivers work with this? the GUI is older but have never had any issues with it and all curves are already set through OpenTrack, which I have set on auto-launch.
** Side note in relation to previous post - I had an issue with my EDtracker recently bumping at the extreme head motions as described in the previous comments, whilst I have no knowledge of internals or of this product etc this problem arose after 2-3 years of use, turns out my cable was damaged, swapped it out, re-ran calibration - problem stopped - Long shot/coincidence, but maybe worth checking.
Thanks for any help
G
Re: EDTracker Bundle (HCKITS0021, 21B, 21E, 21F, 21G, 21H & 21I)
@Witorwitor
Nope, no special method, we just randomly rotate it by hand on the end of a USB cable until we've got enough samples in each orientation. We don't bother doing a figure of 8 though. IF you keep trying and keep ending up with distorted spheres this is usually a sign of magnetic interference.
Yeah again, I really can't speak for the ones that weren't purchased from us but you really shouldn't be seeing that issue with all 4. It suggests something else is going on. The offer is always open to get the one from us back so I can take a look myself if you still have no luck.
It's set to just at the point you start getting samples even when it's completely stationary on a desk. Just as Brumster does it in the calibration video. It will differ slightly from EDtTacker to EDTracker but 60% sounds about right but I don't think it matters too much. Obviously if you have it set to far in either way it will take you either forever to get enough samples or you'll get to the 2000 limit before you can rotate it in all directions. Anyway in between is fine in my experience.1) at what position to you set the sensitivity slider when you do your calibration? I usually put it at about 60%. Does it really matter where you put the sensitivity slider?
When we're doing batch testing we just do the minimum of 500 points because it's enough to confirm the sensor is working and get stable readings but the more the merrier.2) how many points do you use? I usually to all 2000 points.
3) how do you end up with such a nice distribution? Is there a method or special way of moving the tracker to get this nice distribution? I have already watched some videos how it should be done. First I rotate it about all 3 axis 180 degrees each way, then I move it in a figure of 8 and in other random directions until I get 2000 points.
Nope, no special method, we just randomly rotate it by hand on the end of a USB cable until we've got enough samples in each orientation. We don't bother doing a figure of 8 though. IF you keep trying and keep ending up with distorted spheres this is usually a sign of magnetic interference.
This is a bit subjective but just at a reasonable rate when rotating it with your hands. Not particularly fast or slow. So long as you've got plenty of samples in different orientations that's fine.4) how fast do you rotate it during calibration? Is the slower the better or is speed irrelevant?
There must be something wrong about magnetic field or my calibration, because I now have 4 MPU9250 and they all show the same issue of head coming back a bit after rotation about z axis.
Yeah again, I really can't speak for the ones that weren't purchased from us but you really shouldn't be seeing that issue with all 4. It suggests something else is going on. The offer is always open to get the one from us back so I can take a look myself if you still have no luck.
Comments made by this poster do not necessarily reflect the views of Hobby Components Ltd.
Re: EDTracker Bundle (HCKITS0021, 21B, 21E, 21F, 21G, 21H & 21I)
@savagesquirrels
It is for recentering the EDTracker. It's just an alternative option to assigning a key. You only have to cut a hole if you want to use it but if not just trim it so that it fits inside the case without getting pushed down and causing it to permanently recenter.
All the calibration process is for is to calibrate the accelerometer and magnetometer within the MPU9250 sensor. It will then store these new calibration results within the Pro Micro. It won't change any firmware so if it's working before the calibration it will work fine after.
1) What is the button for? - EdTracker never had one (visible) , Do I need to cut a hole in the casing for this?
It is for recentering the EDTracker. It's just an alternative option to assigning a key. You only have to cut a hole if you want to use it but if not just trim it so that it fits inside the case without getting pushed down and causing it to permanently recenter.
2) After running a new calibration, will my existing ED Tracker software and drivers work with this? the GUI is older but have never had any issues with it and all curves are already set through OpenTrack, which I have set on auto-launch.
All the calibration process is for is to calibrate the accelerometer and magnetometer within the MPU9250 sensor. It will then store these new calibration results within the Pro Micro. It won't change any firmware so if it's working before the calibration it will work fine after.
Comments made by this poster do not necessarily reflect the views of Hobby Components Ltd.
Re: EDTracker Bundle (HCKITS0021, 21B, 21E, 21F, 21G, 21H & 21I)
Is there a version of the software that doesn't call home as their website is down?
Re: EDTracker Bundle (HCKITS0021, 21B, 21E, 21F, 21G, 21H & 21I)
Not that I'm aware of I'm afraid. We have contacted EDTracker about this and are still hoping for a response. If we don't get one we may have to look at modifying the GUI to point to our server.
Comments made by this poster do not necessarily reflect the views of Hobby Components Ltd.
Re: EDTracker Bundle (HCKITS0021, 21B, 21E, 21F, 21G, 21H & 21I)
I'm a massive fan of EDTracker and when I purchased a replacement unit sometime last year I was dissapointed to experiece the same issue as in the video above. After spending many hours trying to overcome this I put it to one side and resigned to the fact it was probably a faulty sensor and bad luck. I purchased another recently and having assembled and tested yesterday I could'nt believe it was doing the same thing again, which is really frustrating as I know they can work with perfect stability.Witorwitor wrote: ↑Tue Jan 12, 2021 1:10 amWhen I move my head left-right (yaw) the head in GUI always comes back a bit. I have tried to calibrate it many times following Broomster videos, but no mater how many times I try I can not get rid of that issue.
The issue is show on this video:
https://www.youtube.com/watch?v=j6MNSWrxaIY
Anyway, I have the unit mounted on top of my headpone's head band with the USB socket to the right, but whilst having another look today out of persistance, I noticed if I changed just the mounting position setting in the GUI as 'Top Left' the unit performs perfectly (albeit with pitch now reversed), but that is easily corrected in Opentrack. I also retried the previous unit I purchased last year and that also works perfectly stable with the 'Top Left' setting.
Whilst I can't explain why this overcomes the stability issue, i'm glad to have found a work around, so worth a try!
PS I'm using GUI version 4.0.4.0.
Re: EDTracker Bundle (HCKITS0021, 21B, 21E, 21F, 21G, 21H & 21I)
Hi @all,
i recently bought an EDTracker DIY-Kit. I soldered it yesterday and setup was a breeze thanks to all the videos and this forum. However, I get the same issue with yaw wandering back to center as described by other users in this thread. Did all 2000 points, the spheres are without any spikes and calibration and testing was done without a headset, just by turning the enclosure and I tried different usb-cables.
Is there a solution to this problem?
Kind regards,
- mo
i recently bought an EDTracker DIY-Kit. I soldered it yesterday and setup was a breeze thanks to all the videos and this forum. However, I get the same issue with yaw wandering back to center as described by other users in this thread. Did all 2000 points, the spheres are without any spikes and calibration and testing was done without a headset, just by turning the enclosure and I tried different usb-cables.
Is there a solution to this problem?
Kind regards,
- mo
Re: EDTracker Bundle (HCKITS0021, 21B, 21E, 21F, 21G, 21H & 21I)
Is it a slow drift back to the centre is does it quickly snap back. Roughly how many seconds does it take?
Comments made by this poster do not necessarily reflect the views of Hobby Components Ltd.
Re: EDTracker Bundle (HCKITS0021, 21B, 21E, 21F, 21G, 21H & 21I)
Let's say I turn the tracker to a 45° angle (linear / off center), then it takes about 3-5 seconds for returning to about 30° and after that it reduces the angle continuosly but slower (1° to 2° per roughly 10 seconds). If I return then to center it overshoots to about 20° to the other side and returns to center in the same manner.
It looks like the compensation from the gyro is somehow correlating to the drift. After the acceleration of the turn, the gyro is reducing its values to zero and while it is doing this, the drift is faster then when it reaches zero. I am using the firmware which was on the arduino and the GUI V4.0.4.0.
Hope this helps, I could make a video of it if neccessary.
- mo
It looks like the compensation from the gyro is somehow correlating to the drift. After the acceleration of the turn, the gyro is reducing its values to zero and while it is doing this, the drift is faster then when it reaches zero. I am using the firmware which was on the arduino and the GUI V4.0.4.0.
Hope this helps, I could make a video of it if neccessary.
- mo