I wanted to do long range telemetry so I got myself a dipole antenna for the modem.
I never could check the range because the plane couldn't get a gps lock anymore, with the modem powered on. With the modem powered off, I had a 3D GPS lock, 17 sats in seconds.
What can I do to get both long range telemetry and good GPS?
Very interesting topic, following
It's less expensive here :
But how to know which one is good between these two : http://www.fpvpro.com/store/receiver-antenna-for-dragon-link-30-cm-... and http://www.fpvpro.com/store/dragon-link-transmiter-antenna-high-per...
Not proven yet - Anton had good GPS fix with the SAME dipole, with the rest of the items shielded and powered separately. That already implies it is not the Dipole or its matching...Lets not confuse the issue here with unproven concepts.
Anton, you need to complete the test as I have asked, and all will be revealed...
Anton, do not get mislead yet. You have not answered my questions and we need to eliminate the items constructively. Your results were good or better with the SAME dipole, when you had the modem separated and shielded and powered from a separate BEC. It is not proven yet that the dipole is the problem.
Not proven, like the actual frequencies that the telemetry system is utilizing and where the third harmonic is in relation to the L1 and L2 GPS frequencies (1575mhz and 1227mhz), the L2 landing around 424mhz and fourth order L1 around 393mhz respectively. However, a badly designed antenna as was shown, being poorly coupled into it's source greatly enhances parasitic signal interference in the entire system, not just the module itself, The first logical place to start solving interference problems it at the source, especially when known, and from that standpoint the most common cause of mysterious interference is not actually what is radiated by the antenna's elements, but what doesn't get radiated effectively and (the energy must go somewhere) gets propagated back through the 'ground' wires in the circuit and the entire system becomes an inefficient antenna. That is the start, get the antenna working as an antenna, radiating as much of the signal as is possible, and all the extra shielding and baluns are suddenly a lot less important. From there, also mounting the antenna vertically will reduce the saturation within the GPS antenna's field, by altering the radiating pattern effectively off axis to the gps, and on axis to a vertical ground station antenna, Next, ensure there are no ground loops within the system, if they are unaviodable, use baluns there, but a properly designed star point ground will often do wonders to eliminate contamination. Ground loops, and even non ground (any wire forming a loop) becomes a crude transformer winding, converting any magnetic frequency within it's dimension into a current, that current is short circuited and produces sizeable current and so is easily re radiated quite effectively into wires running parallel to it, just like a transformer, it also adds itself to any voltage travelling trough the wire. This cross coupling make tracing the source of interference extremely difficult, as the source (loop) does not even need physical connection to the affected circuit. Every system I design has these fundamentals engineered in from the outset, it saves a lot of pain down the track, and improves performance all round. Rather than blaming hardware from the outset, I suggest correcting the known problems first, just my two cents worth.
Had problem similar to yours. Came up with a two solutions:
1) reduce the power output of the 433MHz transceiver. That's a poor man's fix.
2) replace dipole with ground-plane antenna (nose-down, ground-facing). That virtualy doubles 433MHz TX power (as it stops from being beamed into the sky), plus isolates sky-facing GPS from UHF frequencies.
Google for: "UHF ground plane antenna"
The more I read about impedance matching the more I'm certain that that is the problem. The new antenna has a coax cable longer than 1/8 wavelength. And that is a problem:
"A cable becomes a transmission line when it has a length greater than λ/8 at the operating frequency.
For example, the wavelength of a 433-MHz frequency is:
λ = 300/fMHz = 300/433 = 0.7 meters or 27.5 inches
A connecting cable is a transmission line if it’s longer than 0.7/8 = 0.0875 meters or 3.44 inches. All transmission lines have a characteristic impedance (ZO) that’s a function of the line’s inductance and capacitance:"
I don't know the impedance of the transmitter. Measuring it requires expensive equipment. Neither do I know the impedance of the antenna. If I find a friend with the right tools, I still have to build impedance matching circuitry. It this point it seems easier to buy a whole new modem system with matching antennas. I don't understand how youtube is full of antenna building videos and nobody meaures/matches impedance. Anyway, I ordered antenna raw materials and I am still looking for a lab with the right equipment.
I must also say that I haven't taken the original antenna to it's full range. I'll fly with that one and see how far I get.
Anton, I think you are over-complicating the issue. The issue of coax cable lengths is unimportant for this discussion - the fact that your coax is longer than 1/8 Lambda is not a problem. It can be any length desired and has no effect on the signal at all, other than losses in the cable, which will increase with length ( talking meters of length for significant loss). If the antenna impedance is a match to the coax impedance, normally 50ohms for this application, and if the transmitter output impedance is also 50 ohms, then all is matched and the coax length is irrelevant. The impedance of your transmitter should be around 50 ohms - that is a standard in the industry to which mostl transmitters and receivers that couple via unbalanced feeder, ie, coax cable, are designed to.
If you want to short-circuit the process then get an antenna that is known to work, maybe the one Alasdair proposes. If that does not work, then we can go back to the process of elimination.
If on the other hand, you wish to learn more about why it does not work, and more about the antenna and how it works, then maybe have a look at these blogs of mine - Antenna design has been a 30plus year career for me, on commercial and Military Aircraft, and for a small handfull of Amateur Radio Space Satellites..as well as large UAV's 150kg and larger..
The topics may be somewhat advanced, but there are elements that may interest you, and maybe even help!
John, the problem is that achieve the radiation pattern you allude to, you have to have a ground plane that is at least a 1/2 wave long per side, or 1/2wave diameter if a circle. You can make the ground plane from 3 or better, 4 wire elements, each 1/4 wave long. The size is somewhat unwieldy on the aircraft though - around 350mm or 175mm long wire elements...
Also, a ground plane may not help you in distant maneuvers, like a banked turn either way from heading out will likely black out your signal until you level off, unless you are at extreme altitude relative to distance. As Joe alluded to, keep it simple, unless you want the experience of it, spending fifteen bucks on a tuned dipole with a current balun is a lot easier and less risky, and likely lighter than trying to build and tune your own. Purchasing a different brand telemetry system with a factory supplied antenna will likely get you a super cheap rubber duck 8th wave monopole which will achive less range anyway.
I apologize for not reading the entire thread before commenting, but I wanted to put something out there...
Use copper foil tape for RF shielding, rather than aluminum foil. You don't need multiple layers of copper foil tape, although you can use multiple layers if you create a pseudo Faraday shield with them (insulating between each layer, and decoupling ground between them). Still, that would be overkill and likely not necessary in your case.