Just checked the code and...
wups... It's not being saved in EEPROM.

That's my mistake, sorry.

if you don't use tilt...

In DIY_headtracker.ino around line 636 you should be able to find:
tiltFactor = (float)(EEPROM.read(9)+(EEPROM.read(10)<<8)) / 10;

Just add this
rollFactor = (float)(EEPROM.read(9)+(EEPROM.read(10)<<8)) / 10;

And the tilt-factor will control both tilt and roll (the headtracker must be restarted (power-off and on again) for the saved settings to take effect when done like this).

Originally Posted by Dennis Frie

Looks really great, and I'm happy that hear that the headtracker also works IRL

I'll just find the roll-gain for the headtracker...

Thank you Dennis!

You give us great support!
For something you gave us for free!

Please accept donations.
Give us a bank account number, or any way i can donate.

Thank you!

Gábor

Originally Posted by kristaps_r

Sorry for spam.

Just got another 6 DIY HT boards for Fatshark Dominators.

If you are interested PM me.

Kristaps

I got a board from Kristaps a few days ago. I added the IMU and put it all inside my Dominator goggles. If fits perfectly.
After calibration it works beautifully with my Futaba FG 12. No drift.

/Pete

Very NICE HT,
Thanks for taking the time to do this !!

Originally Posted by don9f

which PPM inputs to use and which to ignore when the Trainer switch is operated......for example to use channels 7 & 8 for pan & tilt and ignore 1 to 6.

Unfortunately my DX7 cannot do this. Only the way is DX7->arduinoHT->LRS
I know that version 0.8 has ppm in. Please, explain how to configure it in th sketch.

I have the whole thing set up on a breadboard now so i can play around a little. The weird thing so far is: the scl and sda need to be pulled up to a certain voltage for the gui to even notice the input, but even then it's all 3000's or 90's (exactly as described by Stan). However even with both the sda and scl leads completely disconnencted it will just keep getting these values.

So with the whole thing hooked up and 'running'
-if I disconnect the scl pull up resistor: no more input
-if I disconnect the sda pull up resistor: it keeps plotting the 90s or 3000s
-if i disconnect either or both the sda and scl: it keeps plotting the 90s or 3000s

How many ms is the delay if i use HT as ppm in-out?

What does it mean?
#define PPM_IN_MIN 1000 // 0.5 ms
#define PPM_IN_MAX 4000 // 2 ms

Originally Posted by how_eee

How many ms is the delay if i use HT as ppm in-out?

What does it mean?
#define PPM_IN_MIN 1000 // 0.5 ms
#define PPM_IN_MAX 4000 // 2 ms

Short answer: MAX 1 frame! (around 1/50 second).
Average around 1/100 second
Min: The pulse-width

The latency is really as short as it can be. As soon as the PWM signal for a channel is read, its value is stored in a buffer (note that not the entire PPM signal must be read before channel-values are saved). The routine generating the output PPM-pulse will read from the buffer. In worst case the output-channel will be generated and just after a new signal/position is being received. That will give ~1 frame delay. In best case the new position will be received and immediately used. With different input-channels and output-channels the frame-length must be different and a small delay impossible to avoid.

Bur you shouldn't be able to notice any delay.

The min and max value is used to set the interval in which a pulse is detected as a channel.

Originally Posted by thefamoushat

I have the whole thing set up on a breadboard now so i can play around a little. The weird thing so far is: the scl and sda need to be pulled up to a certain voltage for the gui to even notice the input, but even then it's all 3000's or 90's (exactly as described by Stan). However even with both the sda and scl leads completely disconnencted it will just keep getting these values.

So with the whole thing hooked up and 'running'
-if I disconnect the scl pull up resistor: no more input
-if I disconnect the sda pull up resistor: it keeps plotting the 90s or 3000s
-if i disconnect either or both the sda and scl: it keeps plotting the 90s or 3000s

Sorry, but I don't have any brilliant ideas or suggestions. In the debug-info, do you get the adress from the sensors?

Originally Posted by Dennis Frie

Sorry, but I don't have any brilliant ideas or suggestions. In the debug-info, do you get the adress from the sensors?

where would i find the debug info?

Originally Posted by thefamoushat

where would i find the debug info?

See post #577, that may help you, although I have never tried it myself.....

Cheers Don

Originally Posted by Dennis Frie

Sorry, but I don't have any brilliant ideas or suggestions. In the debug-info, do you get the adress from the sensors?

Hi Dennis,

Let me ask if it is possible to capture end and middle points from the tx?

It would be useful when you use more than one plane.

Gábor

Originally Posted by Gabek

Hi Dennis,

Let me ask if it is possible to capture end and middle points from the tx?

It would be useful when you use more than one plane.

Gábor

It's all a question about software
If you use PPM-in to detect the center-position it should be possible. But it's not a feature that's implemented. Shouldn't take much to write if you wanna gave it a try

Just found this thread ...(looking for head tracker to work on JR ) and scimmed over thread and would like to hear from imagebuff or cactus or any other person that has tried to work this with JR equipment ..

I understand I don't have the trainer switch capabilities or return to center position if lost signal ... but that is o.k. as the work I want to do with this HT system is buiding inspection and I fly DJI controlled Hexa copter Line of site ,,while my assistant wears the googles . So being on full time is no worry for me . But it would be nice to not have to have the assistant yell turn left ,,, now lower ,,,, no higher now .

By the way Thanks Dennis for the work you have done getting a great steady tracker system .

Originally Posted by Dennis Frie

It's all a question about software
If you use PPM-in to detect the center-position it should be possible. But it's not a feature that's implemented. Shouldn't take much to write if you wanna gave it a try

Well i think it is easyer for me to make plane memory.
I have newer worked with PPM before.

Will see.

Thank you Dennis.