My understanding is different. The principle of DME operation is that the aircraft sends pulses to the DME, the DME replies, and the time taken is used to determine slant range. The time delays are small, one thousandth of a second of timing error would produce a range error of 80 odd NM.
The process recognises that retransmission from the DME ground station would not be instantaneous, and even a small delay while the wiggly amps were working down the wires would cause a predictable over read of range. To allow for this the DME ground station holds onto the signal before retransmitting for a fixed period of 50 microseconds, this delay being taken account of in the range calculation back at the aircraft so that a true range is indicated. To make a DME under read by half a runway length, a bit less than a NM, the DME only has to reduce the fixed time delay by 10 microseconds or so. When the aircraft unwittingly compensates for the standard delay you get an under read.
The ideal used to be to position the DME between two thresholds so that it was zeroed at both ends. When this is not possible it is zeroed to the main instrument runway and the approach plate on the opposite runway reads something like 'DME reads 0.2NM at the threshold'. I understand some modern equipment has the capability to change the timing correction on the DME as the ILS is changed from one runway to another.
I also understand that DMEs associated with an ILS can only be used in the localiser coverage area and up to 25,000ft. If the FMS is programmed only to use the DME in the coverage area, that is to say on the approach side, that would explain a 'virtual position' at the threshold.
I can't reference this, its accumulated from snippets of data. I would be interested if anyone can confirm or deny?
Last edited by Alex Whittingham; 4th February 2005 at 14:30.