I have run an additional test on DME bias.
The NACO approach plate for the KLAX ILS 6L says that the DME should read 1.6 nm over the threshold; the NACO approach plate for the KLAX ILS 24R says that the DME should read 1.9 nm over the threshold.
For background, both the ILS 6L and the ILS 24R use the frequency 108.5 MHz. Here are the plates:
First, I started on the threshold of 6L with base package and code from a day or so ago. The DME read -0.6 nm. I then started on the threshold of 24R, and the DME read -0.8 nm. The first reading is low by 2.2 nm, and the second is low by 2.7 nm.
Next, I edited by local copy of nav.dat.gz to change the DME bias for the ILS 6L (originally 1.000) and the ILS 24R (originally 1.000) to 0.000. The DME then read 0.4 nm on the threshold of 6L and 0.2 nm and on the threshold of 24R. The first reading is low by 1.2 nm, and the second reading is low by 1.7 nm.
Next, I removed the ILS 24R entry from my local copy of nav.dat.gz. The DME then read 1.6 nm on the threshold of 6L and 0.2 nm on the threshold of 24R. The first reading is exactly correct, and the second reading is low by 1.7 nm.
Finally, I restored the ILS 24R entry (with 0.000 bias) and removed the ILS 6L entry. The DME then read 0.4 nm on the threshold of 6L and 1.9 nm on the threshold of 24R. The first reading is low by 1.2 nm, and the second reading is exactly correct.
1. Robin Peel, despite the excellent work he does on the database generally, has clearly misunderstood the meaning of the "DME bias" field in the DAFIF. While the kind of DME bias he describes may exist, the numbers he has used were intended for something different.
2. KLAX does not seem to use any bias on its transmitters. Instead, it uses two separate DME transmitters for 6L/24R, colocated with the localizers. They must be set up to transmit directionally so that they do not interfere, the same way the localizers are, or else they can be switched on and off. I can further confirm that there are two separate DME transmitters on the same frequency by the fact that each broadcasts a different identifier: I-UWU for 6L, and I-OSS for 24R. These are the same as the identifiers for the two localizers.
3. FlightGear currently defaults to the closest DME with any given frequency, but that's not what we want for the ILS approaches with the same frequency for both ends of a runway -- we actually want to pick up the DME at the *far* end. That's why the readings were low for each threshold until I removed the opposite-direction approach, at which point they became correct.
Suggested Actions -----------------
1. Disable support for DME bias in FlightGear, at least until we find an accurate data source for it. Our current information is entirely wrong.
2. Find a way to make DME transmitters directional when there are two at the same airport on the same frequency (Curt already solved this problem for localizers, and I suspect that the code will be similar).
All the best,
_______________________________________________ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel