Comments by "vk2ig" (@vk2ig) on "Mentour Pilot"
channel.
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
@BlackEpyon Sounds like you've had some interesting experience with radio frequency devices for networking, and your tale of investigations does credit - it appears that you've looked into it more than the average IT person would have.
A few points in response to your post:
1. Interference to out-of-line-of-sight antennas. Whenever an electromagnetic wave at radio frequencies encounters a conductor (whether it be a good or poor conductor), it will induce currents in the surface of that conductor. (The currents flow just under the surface too, and the depth to which they will flow below the surface is called "skin depth", and is inversely proportional to frequency.) These currents will flow along the surfaces of the conductor, and they will themselves radiate electromagnetic waves. This is what happened when you saw reflection of WiFi signals from metal lockers: the electromagnetic wave induced currents in the surface of the locker door, and those currents caused another electromagnetic wave to be radiated from the surface of the door. (Incidentally, it so happens that for good conductors the electric field component of the radiated wave is anti-phase to that of the incident wave, as the tangential electric field value at the conductor must be zero, so this gives rise to a 180 degree phase shift between the incident and reflected waves.)
If the conductor is curved, e.g. like the outer metal skin of an aircraft, those currents will flow along the surface around the curve. The currents will flow in every direction possible along the surface, radiating as they go. (Obviously the current density drops off as the distance from the incident wave increases.) So it's possible for electromagnetic waves to be radiated by a surface current and affect an antenna which is not line-of-sight to the antenna which gave rise to the incident electromagnetic wave.
A classic example of this effect is a conical horn antenna used at microwave frequencies (and by that I don't mean "microwave oven" frequencies): make the bicone angle too acute and currents will flow on the outside surface of the horn (and electromagnetic waves will be radiated from the outside surface of the horn), even though there is no external electromagnetic wave arriving from elsewhere to excite those currents. (Incidentally, that's not how you want a horn antenna to operate - you want to restrict the currents to the inside surface of the horn so that it launches a plane wave along the axis of the horn.)
2. Frequencies used by mobile phones. These aren't restricted to the 2.5 GHz ISM (or "hash") band where WiFi often operates. Google "LTE frequency bands" and look at all the frequencies!
3. Non-linear effects. With regard to surface currents; to compound matters, if the surface is not continuous, but is made of overlapping sheets (quite typical of aluminium-skinned aircraft), then the thin oxide layer between the sheets acts as an insulator, implementing a form of diode through which these currents will flow. This will give rise to some form of rectification. A basic rule is whenever a waveform is changed non-linearly (e.g. by rectification), then the frequency spectrum of that waveform will also change. So, you can get frequency multiplication, e.g. doubling, tripling, etc; and also intermodulation where two currents at different signals give rise to currents at sum and difference frequencies, or three frequencies give to even more complex sum and difference frequencies, and so on. Some of these spurious frequencies (as they're referred to) could be at the input frequency ranges of certain aircraft radio systems. Consider this in terms of all the LTE frequencies in 2 above, and a number of phones operating on different frequencies inside an aircraft cabin.
4. Is a short circuit always a short circuit? Aircraft metal structures are electrically bonded together, but this is not necessarily effective at radio frequencies. The bonding is mainly for static dissipation. The aircraft accumulates static charge as it moves through the air, and this shouldn't be allowed to build up differentially on separate parts of the aircraft and then discharge suddenly in the form or an arc, causing possible damage or disruption to electrical and electronic systems. Also, bonding helps ensure that lightning currents are passed around sensitive components such as bearings supporting moveable flight surfaces, etc. But the bonding straps have a finite length, and these will not look like electrical short circuits at radio frequencies. Consider a transmission line consisting of two parallel conductors which are not connected to anything at one end and connected to a radio frequency source at the other - if this is an odd multiple of 1/4 wavelength long, then the radio frequency source will see a short circuit, whereas if it is any multiple of 1/2 wavelength long then it will see an open circuit. So similarly with a single wire - the electrical length (in wavelengths) of a bonding wire can make a direct current or low frequency short circuit look like an open circuit at certain radio frequencies. Thus two pieces of metal bonded by a wire strap might actually act as two separate conductors at radio frequencies, i.e one piece of aluminium acts like a patch antenna even though it's part of the aircraft skin.
5. Out of band response of radio receivers. Yes, aircraft electronics are not designed to receive energy from mobile devices (operating at any of the LTE frequencies discussed in 2 above). But how good is the front-end filtering of the receivers based on the designer's assumptions of the proximity of nearby transmitters operating on other frequencies? Here are some examples I've seen:
- 14 GHz transmitter completely "flattening" a 12 GHz receiver. This was due out-of-band emissions from the transmitter being high enough in level to swamp the input circuits of the receiver. Solution: install a bandpass filter on the transmitter output which limited the signals at 12 GHz.
- A receiver designed for 7 GHz kept failing when a nearby 8 GHz transmitter was operated. The receiver had a very good lowpass filter at its front end designed to reject the 8 GHz signal. The problem was that the 8 GHz transmitter generated some broadband, very low level, spurious emissions all the way down to below 6 GHz. And the receiver designers hadn't counted on energy at that level down to 6 GHz and below coupling into the receiver - their lowpass filter was completely ineffective at 6 GHz. Solution: install bandpass filters (for the respective operating frequencies) on both transmitter output and receiver inputs.
- Transmitter on an orbiting satellite caused surface currents to flow in one of the metal panels forming the skin of the satellite body. The joint in the panel caused frequency multiplication which overloaded a receiver operating at a frequency five times higher than the transmitter output frequency! That's a very expensive mistake to make ...
None of these transmitters were bad transmitters - all transmitters generate some form of out-of-band spurious emissions, and often the transmitter designer has no control over the nearby environment (e.g. metal structures featuring joints). And the filters on consumer grade electronics are designed to be just good enough to meet the specification ... if the filters are over-designed, then that costs more money, increases the required transmitter power or level of the minimum received signal (all filters exhibit loss, and tighter filters are lossier), etc.
In summary, getting into trouble with radio frequency interference is pretty easy to do if all factors aren't considered - especially in multi-transmitter and multi-receiver environments.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1