D-Star Repeater Performance Topics
D-Star Repeater Multi-band Antenna Configuration
D-Star repeater channel spacing discussion
Co-channel D-Star repeater spacing
High performance D-Star repeater concepts
The TX-RX single antenna combiner system offered seems to be a good idea.
However, there is another way to use 2 antennas and a tower-top amplifier to get even better performance (at added cost).
A few caveats exist that should be fully examined when actually planning such a system. Namely:
- Does the combiner bandpass sufficiently attenuate wideband TX noise?
- Are the antennas sufficiently well aligned to give the needed isolation?
Use a second frequency for the data input--note, you actually may not want the voice/data frequency pairs to be so close if using a bandpass combiner as shown in the diagram, though they'd work in a hybrid combiner.
The configuration above modifies the D-Star data repeater to have a separate RX output.
D-Star repeater channel spacing
Or...how much is good enough? 10kHz, 12.5kHz, 6.25kHz...?
First on using portable radios at greatly reduced power for test generators.
The 91AD with the power turned down may have a higher level of adjacent channel noise relative to the carrier than would a radio used normally at full power --and the repeater itself may well have a purer output than the portable, since the portable is expected to have weaker signals to other radios. There may be higher emissions of adjacent channel noise from a portable vs. a repeater/mobile even at full power by design (miniaturization & cost/power savings consequences).
Second, at what distance would you see on-channel signals -70dBm or greater at a mobile unit? Maybe say 3 miles or less. And then if you're on an adjacent repeater with sufficient signal strength, it may still not bother you at all.
But what if it's -60dBm? Then it's say less than a mile, and the area bothered is truly tiny.
The uncertainty may lead to a very or even vanishingly small area of disturbance for sufficiently strong desired signals.
This effect of "hole punching" occurs similarly with 12.5/25kHz systems (and AM/FM broadcast, etc.), and we take this into account already--we don't want to and do not today waste spectrum by spacing channels too widely for the exceptional case of someone accessing a long-distance repeater while almost literally in the shadow of an adjacent channel repeater tower. And recall that the digital system makes these effects less noticeable because of the high S/N experienced over a wide variety of C/N, i.e. the S/N "gain" from the digital vs. analog circuit.
So we may see that at 10kHz spacing, D-Star presents nearly a noise-limited scenario, even when close to the tower. That seems potentially wasteful to me.
At 5/6.25kHz D-Star repeater channel spacing, we start to come above the thermal noise--NOT a bad thing out of hand, and we need to examine the test-transmitter purity more closely. Efficient spectrum planning is interference-limited, so just like with analog systems, we work to QUANTIFY and consider the effects on those very close to the repeater, but can't afford to make 100% of stations everywhere thermal noise-limited.
So in short, what I would conjecture from this is that 10kHz is an UPPER BOUND that we should not space wider than, and really we should usually be spacing closer than that such as 5 or 6.25kHz.
D-Star co-channel repeater spacing
There is not still 100% clarity (as of Sept. 2007) in the community on how the D-Star repeater controller and mobiles handle the Rpt1 traffic, especially in regards to co-channel repeaters.
My argument has been in regards to co-channel spacing, that the interference vs. coverage contours must be understood as part of the process, along with other concerns such as necessary C/N, D/U etc.
Below is a look at mobile-->Repeater coverage. First, in green, desired coverage, and then two red undesired coverages.
The hypothetical system is a 2 meter repeater with a 300ft antenna with net 0dBd RX system gain, and a threshold of -113dBm to account for site noise etc.
Note how the coverage area greatly increases as the time and location qualifiers decrease.
70% time, 90% location
50% time, 50% location
10% time, 25% location
