John Johnson

had almost the exact same symptom on mine about a year ago and it turned out to be a cold solder joint on the driver board, not the PA at all. tiny hairline crack that you basically couldn't see without a loupe and good lighting. the joint would open up as the board warmed and then sometimes reseat when it cooled down. i found it by very carefully reflowing suspicious joints around the driver transistor area while watching the output on a dummy load. took me an embarrassingly long time to find it but once i reflowed that one spot it never happened again. worth going over those joints carefully before you dig any deeper into the RF sensing stuff.

so our ARES group has a county-wide exercise coming up and im putting together a portable station for the first time on my own without just riding along with someone elses gear. ive been doing this for a few years but always just showed up and plugged into whatever was already running so actually sourcing and setting up my own kit is new territory for me.
the plan is to run an IC-7300 as the primary HF rig, maybe a 2m/70cm mobile unit for local nets, and i need to keep everything going for potentially 8-10 hours. the exercise site has a parking lot and a small shelter but no shore power, so its either generator or battery and ive been going back and forth on this for a while.
for antennas i was thinking an end-fed halfwave on a 31ft jackite pole for HF and just a mag mount on the car for VHF/UHF. the EFHW with a decent tuner should cover 40 through 10 okay i think. my main hangup is the power side — a small Honda 2200i runs clean enough for the radios but its another thing to haul, needs fuel, makes noise which the served agency isnt thrilled about. on the other hand a 100ah lithium battery is expensive and i dont totally trust my math on how long it would actually last under real operating conditions.
anyone run similar setups for extended exercises? curious how people handle the generator vs battery tradeoff when you actually need the station to stay up reliably

yeah same boat as you a while back and i ended up doing the dipole just because it was easier to put up and take down if something went wrong. noise thing is real, my neighbor has some kind of switching supply that absolutely murders my S meter and the vertical made it way worse for some reason. dipole at 25 feet still gets me into europe on 40 most evenings around sunset so its not like its useless at that height

so i finally did my first POTA activation yesterday at the state forest about 20 minutes from my house, been putting it off forever because i kept thinking i needed more gear or a better antenna setup or whatever. anyway i threw the KX2 in a backpack with my random wire and a 9:1 unun and drove out there, found a picnic table near the parking lot which felt like cheating but whatever
got 10 contacts in about 45 minutes which i know is the minimum but man the last 3 were a struggle, 40m was getting noisy and i couldnt figure out if it was the antenna or just band conditions. packed up right as it started drizzling. logged everything when i got home and uploaded to the POTA website, showed as valid so i guess it counts lol. anyone else have a rough first activation like this or is it usually smoother

I own both and use the 891 exclusively for POTA now. The receiver is surprisingly good - not 7300 level but more than adequate for field work. The real advantage is the integrated CAT/audio interface and that 100W gets through when QRP won't.

the warbling hum thing sounds like it could be your LO getting into the audio path, especially if youre running the VFO and audio stages close together on the same board. direct conversion receivers are notoriously sensitive to that kind of stuff. the fact that it changes when you touch the enclosure makes me think theres a ground loop somewhere or maybe the bypass caps on your audio amp power rails are insufficient. id try stuffing a 100uF right at the chip supply pins if you havent already, and make sure all your grounds are star-grounded back to one point rather than daisy chained.
on the BFO injection question — for a standard SBL-1 or similar diode ring mixer you generally want around 7 to 10dBm at the LO port, going higher doesnt help much and can actually start to degrade things depending on the diodes. 10dBm should be fine. if sensitivity feels low id check your RF preamp gain first, or just try adding a simple one or two transistor amp stage ahead of the mixer. sometimes that makes a huge difference.

yeah the 3.3v to 5v thing bites everyone at least once. those cheap relay boards usually have an optocoupler that wants a proper 5v signal to trigger reliably — 3.3v is kind of in the grey zone depending on the specific board you have. some of them actually work fine on 3.3v, some dont, total lottery.
easiest fix ive found is a BSS138 based level shifter, you can get a little sparkfun board or the adafruit one for a few bucks and it cleans everything right up. alternatively if you want to avoid the extra board, look for relay modules that specifically say "active low" with optocoupler isolation and are labeled 3.3v compatible — they do exist. sainsmart has some that work fine direct from the pi gpio without any drama.
the chatter is probably just the marginal trigger voltage causing the coil to not fully latch. once you get solid logic levels that should go away. i ran a similar setup for about two years driving a 6-position antenna switch and it was completely reliable after sorting the level issue.