I can't remember who exactly I gave my money too - he was printing the boxes but had run out. It was only £3 so I am not exactly crying over it.
DIY Head Tracker For A Tenner
- Thread starter Dead Fred
- Start date
Rather than drill-out a huge hole in my Maplin case (so you can reach the button), instead I bought a very tall button, so that the hole only needed to be 4mm diameter (it also looks more professional).
The one I bought was labelled 6*6mm with an 8mm button:
Which I bought from here for 99p:
http://www.ebay.co.uk/itm/Quality-M...-Miniature-Mini-Micro-Small-PCB-/180732232689
When that buy-it-now eBay 'auction' has expired, you can look for others selling something similar. Just make sure the "pin pitch" is 6mm*4.5mm.
And while my button is 8mm high, one slightly less tall (e.g. 6mm) might also work, since my button protrudes nearly 4mm outside of the case.
Yes a longer switch is a possible change, not sure I would call it more professional, more a personal choice.
I like the sub flush fitting, but each to his own, not telling anyone that my way is the only way. Also I bought 10 of the shorter switches for the same cost.
I've got some REALLY WEIRD problems with my EDTracker:
1. It's quickly stops working (usually stuck at a large angle) UNLESS I have the EDTrackerUtilV2 running! This happens both in Elite Dangerous AND with Windows' joystick utility.
2. I've got basically no (sideways) head rotation (which luckily isn't used by Elite Dangerous anyway). Some kind of response is visible in EDTrackerUtilV2, but it seems a bit erratic (and goes back zero-ish on it's own).
(3. Elite Dangerous' headlook seems to be non-linear, even though I set EDTracker to be linear.
EDIT: Looks like EDTrakcerUtilV2 was lying! Changing it to exponential & then back to linear fixed this!)
Could any of these be caused by a bad solder joint, or a bad board? Or something else?
1. It's quickly stops working (usually stuck at a large angle) UNLESS I have the EDTrackerUtilV2 running! This happens both in Elite Dangerous AND with Windows' joystick utility.
2. I've got basically no (sideways) head rotation (which luckily isn't used by Elite Dangerous anyway). Some kind of response is visible in EDTrackerUtilV2, but it seems a bit erratic (and goes back zero-ish on it's own).
(3. Elite Dangerous' headlook seems to be non-linear, even though I set EDTracker to be linear.
EDIT: Looks like EDTrakcerUtilV2 was lying! Changing it to exponential & then back to linear fixed this!)
Could any of these be caused by a bad solder joint, or a bad board? Or something else?
Last edited:
Thanks Neil. Delay in responding as our electricity has been off all day. It happens roughly twice a year while they again fix the power cables in our area. I am in the same state but I have restarted the whole flashing etc process. I'll probably have to complete it all tomorrow and I'll check the device showing up as a game controller. Bit of a trial when one is not sure of what is supposed to happen and what I am doing wrong or what Windows 8.1 is not liking. I do get lots of pretty lights flashing - like Christmas...
First off, thank you for creating this project!
I’m a tinker, so I like to build my own when it comes to projects like this one. I got all the parts a few days ago and soldered it together. My soldering skills are very rusty and my eyesight is poor.
I mention this as I had some problems. I was experiencing random disconnects and issues with the tracker connecting to the EDTrackerUtilV2 (kept getting COM errors). It was so bad I thought for sure I had either a bad part or did something wrong. Oddly enough, I had no problems flashing and calibrating the tracker, it was only during use that things seemed broken.
For my board, I noticed that when the tracker does not connect the green light is off on the bottom. When the tracker is working fine, the green light is always on, the yellow light blinks, and both red lights are always own.
Since I have very poor eyesight I found my 3x reading glasses (normally 1.25x works) and noticed I had really fouled up a few of the pins when soldering. Touching up those pins solved the random disconnects and connecting to EDTrackerUtilV2. So far everything is working great.
I had another problem with Windows 8.1 giving me errors with the tracker drivers. After every reboot the tracker seemed to lose it drivers. In other words, I got the yellow exclamation point for the tracker. Installing the Arduino 1.5.6r2 driver would do nothing. I had to let windows search for a driver and it installed USB Composite Device. Once done the tracker worked fine until the next reboot. The driver (arduino-1.0.5-r2-windows.exe) reaper7 posted about solved that issue.
I made a very rough box for the tracker out of a Tic Tac box. It is ugly, but functional. The box is velcro to my headset. Overall, I am very happy with this setup.
I’m a tinker, so I like to build my own when it comes to projects like this one. I got all the parts a few days ago and soldered it together. My soldering skills are very rusty and my eyesight is poor.
I mention this as I had some problems. I was experiencing random disconnects and issues with the tracker connecting to the EDTrackerUtilV2 (kept getting COM errors). It was so bad I thought for sure I had either a bad part or did something wrong. Oddly enough, I had no problems flashing and calibrating the tracker, it was only during use that things seemed broken.For my board, I noticed that when the tracker does not connect the green light is off on the bottom. When the tracker is working fine, the green light is always on, the yellow light blinks, and both red lights are always own.
Since I have very poor eyesight I found my 3x reading glasses (normally 1.25x works) and noticed I had really fouled up a few of the pins when soldering. Touching up those pins solved the random disconnects and connecting to EDTrackerUtilV2. So far everything is working great.
I had another problem with Windows 8.1 giving me errors with the tracker drivers. After every reboot the tracker seemed to lose it drivers. In other words, I got the yellow exclamation point for the tracker. Installing the Arduino 1.5.6r2 driver would do nothing. I had to let windows search for a driver and it installed USB Composite Device. Once done the tracker worked fine until the next reboot. The driver (arduino-1.0.5-r2-windows.exe) reaper7 posted about solved that issue.
I made a very rough box for the tracker out of a Tic Tac box. It is ugly, but functional. The box is velcro to my headset. Overall, I am very happy with this setup.
Can't even flash the calib
Hi
I've just built up an EDTracker today (ATmega32u4 and MPU6050).
I've plenty of experience with micros and soldering so I thought this would "just work".
However I can't even get the Calib image into it!
The drivers load and it identifies as an Arduino Leonardo on COM6.
But when I try the GUI I get repeated connection errors on COM6.
If I hit the FLASH button it says "flashing" but it never connects and the GUI head never turns. So it never identifies it as programmed even if I power cycle it.
I've tried this on my home and work PCs. Both run Windows 7 x64. Both have Java 7x. I've tried different cables and ports on each.
I've also tried loading up the Calib sketch in the Arduino GUI but it fails to upload with errors in mpu_set_dmp_state(1);
Has anyone seen this? Do I just have a bad Micro?
Thanks
Marc
Hi
I've just built up an EDTracker today (ATmega32u4 and MPU6050).
I've plenty of experience with micros and soldering so I thought this would "just work".
However I can't even get the Calib image into it!
The drivers load and it identifies as an Arduino Leonardo on COM6.
But when I try the GUI I get repeated connection errors on COM6.
If I hit the FLASH button it says "flashing" but it never connects and the GUI head never turns. So it never identifies it as programmed even if I power cycle it.
I've tried this on my home and work PCs. Both run Windows 7 x64. Both have Java 7x. I've tried different cables and ports on each.
I've also tried loading up the Calib sketch in the Arduino GUI but it fails to upload with errors in mpu_set_dmp_state(1);
Has anyone seen this? Do I just have a bad Micro?
Thanks
Marc
If you check 15 posts back, you will see that I posted with a very similar problem, and eventually got it to work (I'm not quite sure how though):
http://forums.frontier.co.uk/showpost.php?p=716912&postcount=1951
I've got a similar problem (although it only happens when NOT using the EDTrackerUtil), and I was rather suspicious of at least one solder joint, so I guess I'll give that a go...
EDIT: Didn't help. I also checked the electrical connections between the two boards as best I can, and they all *appear* to have a low resistance... EDIT: I might be mistaken - need to check further...
Last edited:
I've just bought the 'GY-5150 / MPU-9150' board off amazon.
this has a magnetometer (compass) incorporated into the chip as well as the accelerometer and gyroscope that is found in the MPU 6050. it was more expensive but it has been mentioned that there may be a way to use the compass to help stop any drift issues.
does anyone know if there has been any development work done / any progress made on this?
If not that's fine, but in any case i'd be interested to find out any theories as to how this functionality could be incorporated.
Edit: I'm a bit late to the party so to speak, so sorry if this has alreday been covered
this has a magnetometer (compass) incorporated into the chip as well as the accelerometer and gyroscope that is found in the MPU 6050. it was more expensive but it has been mentioned that there may be a way to use the compass to help stop any drift issues.
does anyone know if there has been any development work done / any progress made on this?
If not that's fine, but in any case i'd be interested to find out any theories as to how this functionality could be incorporated.
Edit: I'm a bit late to the party so to speak, so sorry if this has alreday been covered
Last edited:
I bought one of those too, since I was pretty sure the current attempts to "fix" the yaw drift issue were kludges that wouldn't work too well... and based upon my experience so far, it seems that I was right (I have to reset pretty often, despite calibrating it for 30 minutes). I do intend to calibrate it again though, now that I've fixed my EDTracker problems (see other posts).
Pitch & Roll don't drift (much) because they have a 'point' of reference (downward pull of gravity). Yaw has no such reference, so it can drift forever... But a magnetometer (compass) should provide such a point of reference, meaning that (once implemented) Yaw can never drift too far. Potential problems: There is a youtube video (unrelated to EDTracker) showing the magnetometer is very noisy, but IMHO that can easily be solved by averaging it's output. But that will add a delay to the magnetometer, so any anti-drift corrections will need to be slow/small (but they should be anyway so no problem).
Unfortunately using the magnetometer is a low priority AFAIK.
Last edited:
Now fixed! This was EITHER a faulty solder joint OR a tiny tiny piece of solder "spray" that had landed on the circuit board (and was touching a component or two).
This was partly my misunderstanding (I was expecting the EDTrakcerUtilV2 head to rotate - it doesn't), but even so I think there might still have been a problem related to the above fixed issue. In any case it works fine now when I look at the Windows joystick utility.
So everything working great now
You'll find that when the board is switched to flashing mode that the com port changes. Open the device manager and leave it open when you try to flash and you'll probably find an unknown device show up. I had to force-install the drivers.
If you bought the kit from hobbycomputers then you'll have a sparkfun pro micro clone and you'll be better using these drivers for flashing https://learn.sparkfun.com/tutorials/pro-micro--fio-v3-hookup-guide/installing-windows
Now fixed! This was EITHER a faulty solder joint OR a tiny tiny piece of solder "spray" that had landed on the circuit board (and was touching a component or two).
This was partly my misunderstanding (I was expecting the EDTrakcerUtilV2 head to rotate - it doesn't), but even so I think there might still have been a problem related to the above fixed issue. In any case it works fine now when I look at the Windows joystick utility.
So everything working great now
Phew! Good news
glad to hear it Rob has some test code and has found exactly this re: the noise. So it's not at a useful point yet, but it's on our list. I'm just finishing off the online shop to help ease the order process a bit, then I can turn my head to other things, maybe looking at the magnetometer integration with Rob.
Excellent, i for one will look forward to that
hobbycomponents only had a very small number of sparkfun boards and now seem to be doing their best to source other arduino pro micro boards.
But you will have an issue with using the sparkfun one if you have windows 8
In case it helps anyone else, in addition to the DIY PCB & the DIY Head Tracker Bundle, I used the following things to build my EDTracker:
http://www.maplin.co.uk/p/antex-25w-soldering-iron-type-xs-pvc-cable-fr12n (maybe not the best, but I already had it)
http://www.maplin.co.uk/p/antex-soldering-iron-stand-st4-fr20w (goes with the above, not essential but helpful to avoid burning your table/etc & avoid starting a fire!)
http://www.maplin.co.uk/p/lead-free-silver-solder-tube-n51aw (lead-free, with silver to make it melt quicker. Lead-free is a very good idea as you may breath-in the solder fumes, and lead is really bad for your brain.)
I tried using the following to help hold the board in place (with the sticking-out pins), but it was NOT very helpful:
http://hobbycomponents.com/index.php/solderless-400-point-breadboard.html
Instead I think something like the following would work MUCH better:
http://www.maplin.co.uk/p/helping-hands-with-magnifier-n30ch
(Just make sure to put something soft between the metal clips & the board - felt would be ideal. And note you can get one much cheaper on Amazon.)
I used the following box:
http://www.maplin.co.uk/p/hammond-m...se-abs-box-1551-series-black-50x35x17mm-n78bq (black)
http://www.maplin.co.uk/p/hammond-m...ose-abs-box-1551-series-grey-50x35x17mm-sc78k (grey)
Beware the box & official EDTracker PCB need modifying, as described in the following guide:
http://www.edtracker.org.uk/index.p...ownload=18:edtracker-maplin-project-box-guide
(if the guide doesn't download then try looking for it on this page: http://www.edtracker.org.uk/index.php/downloads/category/1-hardware )
I bought myself a much taller (8mm high) button:
http://www.ebay.co.uk/itm/Quality-M...mponents_Supplies_ET&var=&hash=item2a147b03f1
But if you want to use this then you canNOT precisely follow the above guide. At the very least you need to drill a hole for the button BEFORE you try putting the board in the box & cutting a hole for the USB cable. I drilled a 3mm hole (using a 3mm 'spikey' wood drill), and then enlarged it to a 4mm hole (using a normal 4mm drill) once I had checked the 3mm hole lined-up.
Also, I strongly advise cutting the corner out of the official EDTracker board (PCB) before soldering it. Contrary to the guide, I cut a 3mm*4mm hole out of the corner (not 4mm*4mm), as that gave a tighter fit. I used a junior hacksaw, and then (IIRC) filed it down very slightly. As a result, I didn't need to put ANY padding inside the box
Make sure you have some medium-sized sticky labels & a permanent marker, as they were invaluable for marking where I needed to cut & drill the case! And a good craft knife will be VERY helpful for the plastic parts that a junior hacksaw cannot cut. A file is also very handy. As is a good ruler (with 1mm markings).
And as suggested by the guide, I did not cut ANY of the leads. That worked pretty nicely.
Once finished I used some stick-on Velcro to attach the box to my headset.
When handling the PCB (with Arduino/etc boards attached) I took anti-static precautions. An anti-static wrist-strap would be best, but periodically touching a metal radiator or tap would probably be sufficient.
A long (1.8m+) USB cable is also a good idea:
http://www.amazon.co.uk/USB2-0-Plug-Micro-B-Speed-Cable/dp/B003DQWSDK
(of course you may need an even longer cable than that, depending on where your computer is)
While it's probably better to get someone with prior experience of soldering to do it, the following guide might be helpful:
http://www.instructables.com/id/How-to-solder/step4/Soldering-components-onto-a-circuit-board/
If you don't want to cut a corner out of your PCB, and don't mind needing to heavily pad your case, then one of the following slightly larger boxes SHOULD work (untested):
http://www.maplin.co.uk/p/hammond-m...se-abs-box-1551-series-black-60x35x17mm-n79bq (black)
http://www.maplin.co.uk/p/hammond-m...ose-abs-box-1551-series-grey-60x35x17mm-sc79l (grey)
P.S. I personally advise upgrading your DIY Head Tracker Bundle to include the MPU9150, as I suspect it will greatly help with Yaw drift once they add code for the magnetometer (aka compass). But it does increase the bundle cost by £10, and there is no guarantee it will ever provide an advantage, so it's your decision!
P.P.S. Once I am sure no more modifications are required, I plan to put a blob on Araldite over the board's USB socket, to prevent it from coming loose with repeated (un)plugging of the cable. (That is apparently a common problem.)
http://www.maplin.co.uk/p/antex-25w-soldering-iron-type-xs-pvc-cable-fr12n (maybe not the best, but I already had it)
http://www.maplin.co.uk/p/antex-soldering-iron-stand-st4-fr20w (goes with the above, not essential but helpful to avoid burning your table/etc & avoid starting a fire!)
http://www.maplin.co.uk/p/lead-free-silver-solder-tube-n51aw (lead-free, with silver to make it melt quicker. Lead-free is a very good idea as you may breath-in the solder fumes, and lead is really bad for your brain.)
I tried using the following to help hold the board in place (with the sticking-out pins), but it was NOT very helpful:
http://hobbycomponents.com/index.php/solderless-400-point-breadboard.html
Instead I think something like the following would work MUCH better:
http://www.maplin.co.uk/p/helping-hands-with-magnifier-n30ch
(Just make sure to put something soft between the metal clips & the board - felt would be ideal. And note you can get one much cheaper on Amazon.)
I used the following box:
http://www.maplin.co.uk/p/hammond-m...se-abs-box-1551-series-black-50x35x17mm-n78bq (black)
http://www.maplin.co.uk/p/hammond-m...ose-abs-box-1551-series-grey-50x35x17mm-sc78k (grey)
Beware the box & official EDTracker PCB need modifying, as described in the following guide:
http://www.edtracker.org.uk/index.p...ownload=18:edtracker-maplin-project-box-guide
(if the guide doesn't download then try looking for it on this page: http://www.edtracker.org.uk/index.php/downloads/category/1-hardware )
I bought myself a much taller (8mm high) button:
http://www.ebay.co.uk/itm/Quality-M...mponents_Supplies_ET&var=&hash=item2a147b03f1
But if you want to use this then you canNOT precisely follow the above guide. At the very least you need to drill a hole for the button BEFORE you try putting the board in the box & cutting a hole for the USB cable. I drilled a 3mm hole (using a 3mm 'spikey' wood drill), and then enlarged it to a 4mm hole (using a normal 4mm drill) once I had checked the 3mm hole lined-up.
Also, I strongly advise cutting the corner out of the official EDTracker board (PCB) before soldering it. Contrary to the guide, I cut a 3mm*4mm hole out of the corner (not 4mm*4mm), as that gave a tighter fit. I used a junior hacksaw, and then (IIRC) filed it down very slightly. As a result, I didn't need to put ANY padding inside the box
Make sure you have some medium-sized sticky labels & a permanent marker, as they were invaluable for marking where I needed to cut & drill the case! And a good craft knife will be VERY helpful for the plastic parts that a junior hacksaw cannot cut. A file is also very handy. As is a good ruler (with 1mm markings).
And as suggested by the guide, I did not cut ANY of the leads. That worked pretty nicely.
Once finished I used some stick-on Velcro to attach the box to my headset.
When handling the PCB (with Arduino/etc boards attached) I took anti-static precautions. An anti-static wrist-strap would be best, but periodically touching a metal radiator or tap would probably be sufficient.
A long (1.8m+) USB cable is also a good idea:
http://www.amazon.co.uk/USB2-0-Plug-Micro-B-Speed-Cable/dp/B003DQWSDK
(of course you may need an even longer cable than that, depending on where your computer is)
While it's probably better to get someone with prior experience of soldering to do it, the following guide might be helpful:
http://www.instructables.com/id/How-to-solder/step4/Soldering-components-onto-a-circuit-board/
If you don't want to cut a corner out of your PCB, and don't mind needing to heavily pad your case, then one of the following slightly larger boxes SHOULD work (untested):
http://www.maplin.co.uk/p/hammond-m...se-abs-box-1551-series-black-60x35x17mm-n79bq (black)
http://www.maplin.co.uk/p/hammond-m...ose-abs-box-1551-series-grey-60x35x17mm-sc79l (grey)
P.S. I personally advise upgrading your DIY Head Tracker Bundle to include the MPU9150, as I suspect it will greatly help with Yaw drift once they add code for the magnetometer (aka compass). But it does increase the bundle cost by £10, and there is no guarantee it will ever provide an advantage, so it's your decision!
P.P.S. Once I am sure no more modifications are required, I plan to put a blob on Araldite over the board's USB socket, to prevent it from coming loose with repeated (un)plugging of the cable. (That is apparently a common problem.)
Last edited: