so ive been messing with RemoteHams for a few months now and overall its been pretty solid for just monitoring but lately ive been trying to actually run QSOs through it and the connection just dies randomly. not every time but enough that its getting annoying. usually happens around 10-15 minutes into a contact, doesnt matter if im on 40 or 20m.

my home station is running an IC-7300 connected through the RRC-1258 boxes and the internet link side seems fine, speedtest shows around 40mbps down at the remote end and about 25 at my operating location. latency is maybe 35ms on average which i thought would be plenty good enough. audio codec is set to GSM i think, been meaning to try opus but havent gotten around to messing with the settings much.

what i cant figure out is whether this is a RemoteHams server issue, something on my router at the remote site, or maybe the RRC configuration. the keepalive settings on the RRC units are still basically at defaults because i followed the setup guide and it said defaults were fine for most setups. maybe thats the problem? anyone run into this and figure out what was actually causing it

yeah the keepalive defaults on the RRC boxes are kind of a mess honestly. i had almost the exact same thing happen and it turned out my router at the remote end was killing the UDP sessions after like 30 seconds of what it thought was idle traffic. the RRC uses UDP for the audio stream and some routers treat that differently than TCP and just drop the NAT table entry without warning.

what fixed it for me was going into the RRC web interface and dropping the keepalive interval way down, i think i set it to like 20 seconds, and also making sure the SIP keepalive was enabled. also worth checking if your remote router has a UDP timeout setting you can bump up. some of the cheaper home routers are terrible about this. took me a weekend of poking around to figure that out, the documentation doesnt really make it obvious at all.

also try switching to Opus if you havent, GSM is fine but Opus handles packet loss a lot better so even if you do get some hiccups it tends to recover cleaner instead of just dying.

could also be something boring like the ISP at the remote site doing traffic shaping or throttling UDP after a few minutes. had a buddy with a similar setup and his ISP was the culprit, totally random disconnects that looked like a device issue but wasnt. worth throwing wireshark on it if you can get access to a machine at the remote end just to see if the packets are actually making it out or dying somewhere upstream.