so i've been meaning to get into LEO satellite work for a while now and last weekend i finally sat down and actually did it instead of just reading about it. been licensed for about 6 years but mostly hf so this was pretty new territory for me.

managed to track AO-73 with gpredict and got my ft-991a set up for split operation, uplink on 70cm and downlink on 2m. the doppler correction was the thing that really got me — i knew intellectually that i'd need to compensate but actually doing it manually while trying to hold a qso is a whole other thing. i kept losing my own downlink signal because i was chasing it the wrong way for the first couple passes.

got a few contacts in there though which was cool. my question is about the linear transponder behavior — i noticed that when the bird gets crowded with stations the signal levels all kind of drop together, is that the ALC behavior of the transponder compressing when total power goes up? ive read a few things about this and some sources say you should basically run as low power as you can get away with but im not totally sure how to judge what 'enough' is. running about 5w on uplink with a yagi and it seemed okayish but hard to tell.

also gpredict is doing most of the doppler correction automatically for me but it feels like im still drifting a bit on the downlink. is that normal or am i missing something in the config?

yeah what you're seeing with the signal compression is exactly right — linear transponders have a fixed total power budget and all the stations using it are basically sharing that amplifier. so when the bird gets busy every signal gets a smaller slice, yours included. the usual rule of thumb i've heard is to find your own downlink and then back off your uplink power until you can just barely hear yourself, then add maybe a tiny bit back. if you're running so much power that you can hear yourself really loudly you're probably stomping on other people.

5w to a yagi sounds reasonable depending on your antenna gain and elevation angle honestly. high elevation passes give you more margin to work with.

on the gpredict drift thing — are you using it to control the radio directly via hamlib or just reading the frequencies off the screen and tuning manually? if its manual correction youre going to be behind the curve especially near AOS and LOS when doppler rate of change is highest. getting hamlib set up to drive the rig makes a huge difference, took me a whole afternoon to sort out the config but worth it.

nice work getting those contacts, AO-73 was one of the first ones i worked too. one thing that tripped me up early on — make sure you're accounting for the transponder inversion. AO-73 uses an inverting transponder so if you tune your uplink higher your downlink goes lower, which makes the doppler correction feel really counterintuitive when you're first learning it. i was going the wrong direction for like three passes before someone on the mailing list pointed it out to me. gpredict should handle this automatically if you have the transponder configured correctly in the satellites file but worth double checking that the inverting flag is actually set.