Deviation
As with voice transmission, too much or too little deviation is not ideal. AX.25 packet using AFSK at 1200 baud (i.e. APRS on VHF or UHF) needs pre-emphasis at the transmitter and de-emphasis at the receiver just like voice mode does, and for the same reason (triangular noise spectrum of raw demodulated FM).
Digipeaters
DWAIT
This is a parameter hidden away in the TNC (or equivalent) configuration. It might be alternatively be referred to by the combination of PPERSISTENCE and SLOTTIME. With traditional BBS oriented packet, the idea is for users to politely share the channel, however with APRS being a one-to-many scheme it is to our advantage for beacons to spread out rapidly through the network and it uses up less timeslots overall if digis immediately repeat what they hear, and FM Capture Effect is used to overcome what one might think would have been interference.
For an APRS digipeater, DWAIT=0 is the correct value, alternatively SLOTTIME=1 and PPERSISTENCE=255 are correct values where those parameters are used instead.
iGates
Transmitter Power
What power does an iGate transmitter need to be using when it is gating beacons of all NZ from the internet to RF? Perhaps it only needs enough to be only just heard by the nearest hilltop digipeater, so that RF users including distant trackers are still able to be heard by the digi and not frequently denied access due to their weaker signals being overwhelmed by very strong and very frequent signals from the iGate.
Txdelay
This is the time delay needed between the PTT line being asserted, and the transmitter to have developed power on the desired frequency, adequate for others to receive you properly. Some handheld radios need around 300 milliseconds, some radios designed for data use might need only 10 or 20 milliseconds.
Note that the units used in the configuration program might not be milliseconds, often they are 10's of milliseconds, so a value of "20" might correspond to 200 milliseconds. In the case of a tracker, the units might be some weird number related to a microcontroller's timer.
In practice it is wise to determine the best value of Txdelay. Too short and you won't be received properly by other stations, and too long will be wasteful of precious on-air time.
To find the right value, start with a small value, say 50mS, and gradually increase it in 50mS steps until you are just being heard, then add another 50mS to cater for receivers with a slow squelch. That is the basic idea anyway.
