I have added an equation to the Raspberry-PI to automatically change the Antenna Array Radius as the frequency changes, and I have managed to achieve almost 0.5.
At least the first four points are just the physics of the problem. I share some of your frustration, mine is still problematic (especially re power, i ended up using pi5 power supplies which are 5.1V
Your setup does not appear to have a groundplane under the antennas. They’ll work much better with a metal at least a quarter wavelength surrounding the whips on all sides.
The first image appears to be 1/4 wave verticals with no ground planes, but it is hard to tell. Vertical 1/2 wave dipoles in this configuration would be much better - no ground plane needed. In the second picture it appears there might be a metal groundplane, but all of that metal framework of the tower sticking up above and surrounding the antennas is going to potentially cause accuracy problems. You need to get your whip antennas up in the clear with a good groundplane or use dipoles up in the clear.
I’m thinking of doing this KrakenSDR Direction Finding remotely, so I don’t want to mess with the Antenna at all. I want to keep both VHF and UHF the same. So I’m thinking of doing the Heading remotely. But the Direction is wrong according to the Frequency. The error should be Serial. But the DOA error is not Serial, so it was not convenient to format the Heading.
The antennas are usable, but the problem is that it is inconvenient to move the antennas around. I don’t move the antenna, but I have to change the heading according to the frequency. I want to install it at the height of the tower where I want to place it. Please give me some advice on how to install it.
The KrakenSDR cannot reliably use antenna cables longer than 10 meters due to potential signal degradation and phase inaccuracies. For optimal performance, it is recommended to use shorter cables to maintain signal integrity and accurate phase synchronization between antennas.
Different frequencies are going to have different multipath distortions, especially when interacting with the metal on that tower. I believe that is the effect you are seeing.
The cable length can be any length, as long as the cables are phase matched. For shorter cable runs it’s possible to simply make them exactly the same length, and as long as they are all from the same cable batch it will be phase matched close enough to work well.
I’ve used those antennas before during early development testing with the Kraken, and found that their cable lengths could vary wildly up to several centimeters. Are you confirm the cable lengths are matched with a VNA?
I definitely would be using cable off the same reel if at all possible. Putting a connector on one end, and then checking with a VNA and trimming lengths would be the best way to match the cables I suppose. Mine are all run inside a spiral wrap so they lay identically, none gets bent more than the others etc.
The antenna cable length and type are the same. I tested it and there are some MHz frequencies where the direction is completely correct. At other MHz, the direction is incorrect but the Heatmap is obtained correctly by changing the heading. Therefore, I am thinking of changing the heading according to the frequency. Please advise how difficult it will be.
Kraken-1 and Kraken-2 Pi-5s on the top of the tower can be seen remotely with Mobile Bill, but Krakensdr cannot run.
Today, I bought 65 W adapters for Kraken and installed them for 2 sets of Kraken, but I still can’t find a sufficient power solution.
They are working hard on the tower, but they can’t place the antenna properly.
Good decisions or suggestions are needed.
The main requirement of Kraken is not a direction finding system, nor software, nor network, but about 80% of the power requirement is the antenna part and 20% is the antenna part.
I don’t think you want to do this. The effect you are seeing is most likely due to multipath. So in one particular location the multipath corruption from one frequency will be different to the next frequency. This can change depending on the angle the signal is coming into, so it’s probably best not to try and compensate, as compensating for one direction, may make another direction worse.
Do you have seperate power supplies for Kraken and the Pi? If you do, you should have sufficient power. Unless you are using a really long power cable.
For the antennas, I honestly think all that metal from the tower blocking the antennas is going to cause major refraction, reflection and multipath issues for you. Unfortunately, I don’t really have a good solution for that unless you can get it to the top of the tower.
