mdoerr, do you mean to imply that localiser receivers reverse their indications in certain situations? Are you referring to Backbeam Localisers? Except for those I have never heard about reversed indication.
It is my understanding that localisers not used for backbeam approaches (= by far most of them) have to be shielded in the direction "behind" the front localiser direction, in order not to emit any "false" or "reversed" signals.
I did not quite understand the chain of events in the Condor accident from the very short Wiki text. Was the accident caused by receiving a backbeam localiser signal?
But Peter LPV exist outside the UK.
And +1 just means I agree. Nothing to do with twitter.
When LPV becomes operationally relevant in Europe, I will worry about it 
It will be a while. Just as well - a good £20k, minimum, to re-do my panel with comparable functionality. The bigger problem is who is going to do it, after this experience in the hands of one of UK's "most recommended" shops...
Interesting about false LOCs. I agree there should not be any at anywhere near the normal signal strength. I am sure there are sidelobes; these are impossible to avoid unless, AFAIK, you are into microwaves and using parabolic etc antennae (MLS?). All the time you are radiating with dipole arrays there will be sidelobes, albeit at ever reducing signal levels.
The ILS approach must assume that one will be approaching it in "reasonable" ways. I would expect that if you try to do a LOC intercept 3nm out, you might pick up a sidelobe.
So if e.g. you are flying in the vicinity of some ILS airport, with the autopilot in APR mode, you may well get "pulled over" by one of them. The Sandel SN3500 EHSI won't issue valid FCS control signals unless the indication is vaguely plausible (e.g. if you are flying in the wrong direction it will assume a BC approach) but a normal HSI won't care. And either one will generate valid FCS (autopilot) outputs if the signals are plausible. It can't read your mind.
Likewise, the LOC receiver+antenna sensitivity must be designed with the assumption of a "reasonable" intercept i.e. one at a distance great enough to not pick up the sidelobes.
It's like ATC are not supposed to vector you to an ILS inside the GS intercept, but they do it sometimes (I've had it a few times).
I did not quite understand the chain of events in the Condor accident from the very short Wiki >text. Was the accident caused by receiving a backbeam localiser signal?
The aircraft was destroyed when it flew into the side of a hill while carrying out an ILS approach to Runway 35 at Adnan Menderes Airport., Izmir. The point of impact was some 10.5nm to the West of the airfield at the 2,750ft. level. The accident happened at night and in poor weather with heavy rain. On arrival at Izmir, the aircraft had been cleared direct to the CU NDB located on the extended centreline of the runway some 5nm from the runway threshold. The flight subsequently reported passing CU outbound; however, after passing the NDB, both pilots 'switched to the ILS' and therefore were unable to confirm their position during the procedual turn. The aircraft was outside the +35deg sector of the ILS centreline and the crew apparently followed the wrong side beam. Although concerned about the unexpected ILS indication, they continued with the approach.
According to an ICAO Summary, causal factors included the presumed NDB passage, the wrong use of the VOR and ILS, overconfidence of the pilot and considerable inactivity of the co-pilot.
There was no radar service at that time.
I have looked up LTBJ and there are no less than five different ILS approaches to runway 34.
There is also no "CU" NDB anymore - well not anywhere near that place.
the aircraft concerned was a B737-200 with standard sixpack. In my opinion the accident had never happened with the newer model 737-300 (IRS & electronic map equipped) that already was in use in those days - one of the pilots would have noticed not beeing on centerline but following a side lobe of the ILS by just looking at the map of the EHSI. Shortly before impact they must have discovered their fault.The aircraft was in a go-around-config when it hit the hill. The NDB "CU" was probably not showing any usable bearing because of too much static in the air (CBs & RASH).
On older localizers you had side beams, depending on antennae configuration. Some had 30 degree sidebeam on both sides.
My understanding of the false glideslopes is that they occur in multiples of 3 degree angles, with each adjacent false glideslope reversing the sensing. So the one at 6 degrees is reverse sensing, the one at 9 degrees is correct sensing.
I was just talking to someone about what would happen if one was flying an ILS on the autopilot and intercepted one of the higher glideslopes.
Taking what @ncyankee wrote immediately above:
If you intercepted the first false one, normally the 6 degree one, the GS tracking control loop would be reversed and the autopilot would drive the pitch axis downwards, indefinitely, so you would crash pretty fast. I actually tried that some years ago (outside CAS and a long distance away from the said airport and away from any procedure paths) and sure enough very quickly the VS was -1000fpm and increasing fast.
If you intercepted the second false one, normally the 9 degree one, the GS tracking control loop would be the right way up but the VS required to track it would be 3x the normal one i.e. about -1500fpm. I never tried doing that
The 6 degree case might be a plausible way to crash on an ILS if in IMC and not monitoring things.
Peter wrote:
If you intercepted the first false one, normally the 6 degree one, the GS tracking control loop would be reversed and the autopilot would drive the pitch axis downwards, indefinitely, so you would crash pretty fast. I actually tried that some years ago (outside CAS and a long distance away from the said airport and away from any procedure paths) and sure enough very quickly the VS was -1000fpm and increasing fast.
In this case, it should shortly reach the main glideslope, acquire negative feedback instead of positive, and stabilise there unless the VS is so high that it overshoots it right into the ground. On the other hand, a reasonably designed system should probably limit the VS in approach mode.
