There ought to be a market for a low cost ILS tester.
I started developing one a year ago and spent a lot of time going through the countless patents (which are all lapsed now). It is pretty easy to do it. The IMHO necessary attenuator adds complexity but nowadays there are some nice chips for it. But then found somebody doing one for about $1500 which does DME etc too, and decided that would be more work than I wanted to take on.
So I periodically give 20 quid to a local guy with an IFR4000…. I would buy an IFR4000 myself but everybody knows how much the used ones are worth 
then either you were vectored too close for a proper approach
Unfortunately that happens pretty often… so all that is left to monitor is the ROD on the GS (glideslope) versus your GPS (or DME) GS (ground speed).
A good check is to always verify that there is movement of the GS indication. If you intercept from below the GS and don’t have a fly up indication followed by the GS centering at the appropriate approximate distance from the runway, then you should be extremely suspicious and miss the approach. The sequence should include GS alive, fly up indication followed by GS centered at the correct distance from the runway. If your first indication is GS centered, then either you were vectored too close for a proper approach, you missed the GS alive and movement, or there is a system problem in your aircraft or in the ground equipment. Any of these justify missing the approach.
I thought ILS signals were monitored on the ground as well, with dedicated (and calibrated) monitoring tools?
Yes … but watch the video.
I thought ILS signals were monitored on the ground as well, with dedicated (and calibrated) monitoring tools? If the signal fails or is outside parameters, the ILS is supposedly shut down, or at the very least ATC is informed?
That doesn’t absolve the pilot from any and all responsibility to monitor and cross-check things of course.
The SBO out failure mode is pretty bad, there’s nothing in the signal to detect it, it looks perfectly like an on-glide signal. The only way to detect it is via cross checks to other glide information sources, be it the FAF check (but then you have to look for the glide slope indicator approaching from above), or DME/baro height, or GPS derived glide slope…
I recall seeing this (or very similar) video of this incident, several years ago.
The value doesn’t diminish though. The classic avionics have poor error detection and cross checks are vital. With GPS this sort of thing is much less likely.
I do the double check with a second ILS indication in flight and as I taxi out, I look for the GS swings on the ILS frequency and as I cross the runway, I want to see the localizer center mid runway
That’s clever but needs a departure airport with an ILS 
I at least tune in both ILS receivers for any ILS approach. And set up the indications on EHSI1 (NAV1) EHSI2 (NAV2) KI204 (NAV2).
The VOR check is fairly useless although I suppose it could pick up the loss of the LOC receiver.
I once deliberately tested an ILS with this and the autopilot was quite happy all the way down – the plane flew quite fast though
I posted this on the BeechTalk forum on the same topic and the post is my work.
I experienced a similar failure, but the failure was in the avionics. The aircraft had a single GS from a KX155/KI209. There was no flag and both the GS and localizer were centered. The customer had an Stec 60-2 for an autopilot. The conditions had been below minimums at the airport, with RVR increasing to minimums. We were in a holding pattern at 6000 feet and as the conditions improved to above minimums, we requested an approach and were vectored to final well above the intercept altitude and in a steep descent around 1000 fpm. The ILS was tuned and ID’d and we were given an approach clearance close to the localizer center line. As we were turning inbound, the GS was on center and no flag was displayed. The pilot engaged the autopilot Nav mode and it indicated it had captured the GS. The descent rate on final remained too high at 1000 FPM. We were also too far from the airport and should be below the GS, but we were getting the on GS indication. It took about 30 seconds before it was obvious something was wrong, and I told the pilot to disengage the autopilot and commence a climb. He did not know what to do, so I took over and disengaged the autopilot and commenced a climb. By that time the controller was saying we had a low altitude alert, which I acknowledged that we were climbing and abandoning the approach. After the climb out we got vectors and were asked if we wanted another approach. I declined, indicating we had some sort of equipment failure and we headed back to our departure airport which was marginal VFR.
Back at the home airport, I asked for and got an ILS although we were in VMC at the time. This time we were level, with the localizer tuned and ID’d. The GS was in the center and no flag shown, same as before. The difference is when I engaged the autopilot as a test, we remained level, a more mundane failure. So I switched off the autopilot and landed VFR.
On the ground, there were no swings on the GS, it remained centered without a flag all the way back to the hangar. We talked about it for a few minutes and I asked the owner if he had had any avionics work done recently and he said he just had his GPS repaired. It was mounted just above the KX155, so I had an idea. I removed the KX155 and re-seated it. Then I turned on the avionics and immediately got the GS flag pop into view along with the vertical CDI moving up and down. We did a test flight and the GS worked the way it was supposed to. The reason I re-seated the KX155 was because it is common for an avionics tech to remove the avionics below one being repaired when the fit is tight as the pall latch can have a slight mechanical interference as the higher unit is inserted. This is what must have happened and the KX155 was not well enough seated to provide connections to all of the signals, in particular the GS flag and GS +/- signals.
Ever since then, I do the double check with a second ILS indication in flight and as I taxi out, I look for the GS swings on the ILS frequency and as I cross the runway, I want to see the localizer center mid runway. This drove home to me that the 30 day VOR check does not in any way confirm that the localizer or GS are functioning. It is another reminder that it pays to intercept the GS from below and have a fly up before autopilot engagement. If the needle is in the center and not doing anything, the autopilot will continue to do what it was doing prior to engagement.
I was wondering if you would see the glide slope coming in from above with an erroneous GS?
In that particular case (SBO missing), no. You would get a perfectly on glide indication as soon as the signal is strong enough to “lift” the flag.
I don’t think there is any way to clearly detect a well spoofed glideslope, if you don’t have a distance indication of some sort.
A 4 deg one will result in an excessive speed when tracking it but how many pilots would notice that, in time?
That said, a GS transmitter failure which would present a glideslope at say 2 or 4 degrees (where it is meant to be 3) would be an unusual electronics failure. And if the ILS monitoring is working that would pick it up. In the case of the above video the monitoring was INOP which is really nasty and I am surprised a passenger flight went there in IMC, but I guess it is within their AOC manual.
For a bit of fun, see this. That could easily kill somebody. Often, with avionics, out lives do hang on a thread, which is one reason why I always set up the GPS, in OBS mode, to depict the LOC, when flying an ILS.
