One problem with NDBs is maintaining the equipment. Apparently a friend of a friend is involved in maintaining all this ancient technology (VORs also). Somewhere there is an old guy with an actual electronics workshop who knows how to fix the boards when they go wrong, presumably using his stock of germanium transistors and twin triodes. They have a few spare boards, so when one dies they swap the board and give him the old one to fix.
Of course you COULD build an NDB or VOR with a Raspberry Pi and $20 of components off eBay, but it wouldn’t be certified etc etc etc.
The last time I flew an NDB approach using the ADF was on my instrument checkride in 2002. I was pleasantly surprised that it turned out as well as it did – the runway was visible through the windshield rather than one of the side windows. I don’t think there are any NDBs left within 200 miles of where I fly now.
Yes; that “man with a van full of 1970s circuit boards” is what you get with a lower priced NDB or DME service contract 
VORs cost a lot more and I believe they are still supported with somebody making new circuit boards. DMEs too at large airports; I would bet they are regarded same as ILS for which there is total service support and the whole system is still made (I believe Thales makes this stuff). Philips used to do VOR support for many years.
I think NDBs are still used because
Almost nobody actually tracks an NDB when flying for real. Private pilots usually set up a GPS OBS or whatever track, and the bigger stuff flies a synthetic track on the FMS (usually this includes a synthetic glideslope too). But, from conversations with airline pilots, most European AOC operators need the NDB to not be notamed INOP and to ident OK, even when they are not using it at all; the charade must be maintained and nobody can say the emperor has no clothes
NDBs are accurate for long distance enroute nav, and in good conditions and with top equipment they are as accurate as VORs. The problems start when using them close-in (e.g. approaches) and then the system becomes vulnerable to local effects e.g. weather, TS, or a coastline if it runs assymetrically i.e. not perpendicular to your current track. As an example of the last one, the EGKA NDB IAP (inbound track 020, coastline running 090-270) is way off between about 2nm and 5nm. The RMI needs swings tens of degrees off, briefly. On a CAA IR checkride you are supposed to track it, and then you can end up not in a position to land. I once flew the EGHH NDB approach with a freelance examiner, as practice, and despite flying it perfectly I then had to do a huge dogleg at minima which would never be allowed on a commercial flight.
I too had an RMI installed (dual needle) when I got my TB20 in 2002. A great instrument; must better than a CDI. Almost never used it and eventually replaced it with the 3.125" version of the EDM700 which is much more useful
Even if everything is working correctly, a “classic” NDB approach with the NDB say 5 nm off the field, serving simultaneously as IAF and FAF with a PT in between, just do the math… in the unlikely event that your compass is accurate to 1° and you can also fly to that accuracy, on average you will be 400’ off the runway when you break out. Given a real pilot and real equipment, it could easily be 1000’ or more.
an NDB is the cheapest way of making an airport legal for IFR and thus legal for AOC flights (hence a lot of them around e.g. Croatia)
This of course is assuming you hold to the normal European xAA view that GPS/RNAV/LPV (for the latter, cross yourself three times and immediately attend confession) are the work of the devil. Otherwise, nothing could possibly be cheaper since absolutely no ground equipment is required.
Indeed, although a lot of bigger passenger jets can’t fly GPS approaches.
I am told by a 787 pilot (used to post here) that it cannot fly LPV.
But all can fly NDB on the FMS.
That is funny actually – and quite a pity that such a plane with expensive equipment cannot fly LPV.
. Here is another example of the (un)precision of NDB we all know. LFMV ILZy17 approach, you have outbound leg from CM ndb on route 003°.
I am on route of the GPS overlay, but needle says i’m on radio 020.
I think that people that fly NDB approach in mountain area are quite courageous.
I did spend quite some time in training flying QDM and QDR’s and a few NDB approaches. On approach at minimums I was allowed to remove my ‘foggles’ and the runway was quite a bit on my lefthand side. My instructor said: ‘well, it IS a non-precision approach’. I never flew one since. My examinator told me before the check that he didn’t do anything with the ADF, to my relief.
Peter wrote:
NDBs are accurate for long distance enroute nav, and in good conditions and with top equipment they are as accurate as VORs.
Do you mean that e.g. coastal effect and thunderstorms have less effect on ADF accuracy the further away you are from the station!?
Am I remembering correctly, but don’t NDB’s also suffer from ionospheric effects which in particular leads to them being inadvisable for use at night especially at a distance?
VORs are still being made, with modern technology, with the units fitting into typical 19inch telecom racks (just like what you will find in any company’s server room). As such, a VOR is easily servicable with modern parts. It’s likely if a VOR goes down they’ll just switch out the faulty component from the rack and get it back working again in short order, just as you might do with a faulty server in a data centre.
I’d imagine most ADFs now are at least 1980s technology, so no germanium transistors, and are likely not to be that difficult to repair with in-production kit. It’s just a low powered AM transmitter after all.
gallois wrote:
Am I remembering correctly, but don’t NDB’s also suffer from ionospheric effects which in particular leads to them being inadvisable for use at night especially at a distance?
From what I remember from ATPL theo. and a bit of IR, it is actually better during the night thanks to ionospheric reflexion, but does suffer a lot of CB, mountain reflexion and coastline which deviate the needle. Lighning does attract needle.
