Thanks heaps for your feed back. I am delighted to have your help. I am currently stuck writing a complex document for a dispute I am having about my house.
It should be over in a few days.
The main thing I am thinking is how to display the entries so that people can gain instant knowledge and use that earlier entry to improve their entry.
ie some one does a great gimbal and some else takes that and makes a motor for that gimbal.
I would give the prize to both entries.
āThe main thing I am thinking is how to display the entries so that people can gain instant knowledge and use that earlier entry to improve their entry.ā
Motivation is the key in any competition, options:
Mental stimulation ovelaps, but nobody will reveal what s/he thinks is the best if thereās a 200 dollar price involved. Also, to test antennas youāll need equipment, so sending a second receiver is moot as a price.
Outernet and satellite signal reception is interesting for people, I see from my webpage stats, more than 10,000 have read the easy outernet antenna post, so itās clearly a current and exciting aspect of radio.
Some of the categories listed above are only interesting to a limited set of people, complicating th=e competition.
As for the learn from mistakes, I think that the competition could be run in two stages, with only ONE category, entrants got 168 hours and ten dollars to build the best antenna. Other categoies for mobile ops etc will not bring in sufficient entries, no general appeal.
So ONE category, 10 dollars, 168 hours, get the best signal to noise ratio with an Outernet kit. Best Five gets loads of money.
Next stage, build and experiment with designs above, make them better.
Before it comes up Iāll not enter the competition, Outernet is already supporting me (Syed if youāre reading this it would be good to be able to reach the office, or talk to the company in a different channel, please. PMā¦), Iāve tested different antennas and will make them public for free, my joy is getting Outernet reaching a wider audience.
Please let me know what might be required in the future on my part.
Iāve been trying to get an antenna pointing system working using an MPU9150 (MPU-9150 | TDK), arduino and 3x 9g servoās.
Iāve been waylaid for the past few weeks so not too much progress lately :-(.
Once you get the raw gyro, accel and mag data, you can fuse them to an absolute pose using those algorithms:
I have found several sensor fusions that state that they can find magnetic with no twisting, but have not had stable out put from any of them. You call this one ok?
Most find gyro, or accel stable, but do not put out a good heading. And all I need is a heading.
-C
I want to solve a heading to point then antenna. Elevation can be solved by the turn of a screw every 20-30 degrees north or south. Gravity will make the antenna stable enough. If I can point in a 15 degree sector the outernet signal should download.
-C
I am well beyond trying to get good outputs from the sensor.
If you take care to avoid electromagnetic influences and calibration then the heading is reasonably repeatable. Have you seen how stable drones are nowadays?
I donāt see why you wouldnāt make the thing fully automatic, no need for manual mechanical adjustment due to latitude. It should regularly read an input string with its current/updated position and the desired azimuth and elevation of the target.
My challenge is now more mechanical - a good compact design for the servo mounts - but I keep getting confused with the orientations 
Unfortunately, the XIO code doesnāt have magnetic calibration at all.
For a compass to work well, you need to do soft and hard iron correction.
That means eliminating any static magnetic fields and calibrating out directional dependencies.
If your compass is calibrated right, the magnitude of the 3D magnetic vector needs to equal the magnetic field strength at your location in all possible directions.You can look up the to be expected strength here:
https://www.ngdc.noaa.gov/geomag-web/#igrfwmm
I have only used the method where you need to swivel the sensor around and then do a sphere fitting to eliminate static fields.
The open source RTIMULIB library has full calibration functionality. It is well proven and is used in many drone applications. Like I said getting sensor output is not a problem.
Of course, care will need to be taken when installing the sensor.
I will go look at that library, thanks!
-C
With all of the kits getting packed up this week I look forward to a whole new set of Outernet receivers coming online in the next month.
With that in mind I would like to bump and find out where we are on the antenna design competition.
-C
I built three patch antennas, adjusting feed points and inverting to LHP for use with dish I was able to achieve this:
Yes, Iām trying to pack about 100 units per day. Fingers crossed we get our final protos back next week and then a few days for validationā¦and then on to production!
gotta love that antenna mount
very nice! 
That is not bad, considering that this kind of grid is not suitable for circular polarisation.
You are effectively only using linear polarisation. Could have used a dipole+reflector
instead of the patch.
ā¦I do tend to lose signal from the bird when the burgers are done
True, but who cares of losing 3dB on polarization when he still get 17dB SNR? 
Also a 18dBi Yagi would be nice, acting as 15dBi antenna for RHCP. I it would exists for L-Bandā¦
For 2.4 GHz wifi those quite popular and cheap: http://wtfmoogle.com/wp-content/uploads/2012/07/DSC00023.jpg
Sure, but I consider it a bit of a waste of time to construct a patch with different polarisation, and then
placing it in front of a grid only relecting linear polarisationā¦ it is in fact worse, because the other
polarisation component will go right through the grid and pickup interference from behind. A linear
antenna would perform better.
However, when another type of reflector is used, it is the correct approach.
(you cannot put the stock antenna in front of a reflector)
