Everything posted by Mod_Tom
-
Building a 40/20 wire dipole from scratch: wire, feedpoint, choke, and how far to trust the model
On modeling, both programs are useful, but check them against each other before trusting either. EZNEC is based on NEC, so wire junctions, segment length, and ground choices matter. A frequent error is mismatched segment lengths at a junction, or too few segments on a short wire. MMANA-GAL has its own calculation engine and a handy geometry editor. Its ground handling is simpler, so differences in feedpoint impedance between the two aren't surprising. If the two agree on the general shape of the pattern and on how resonance moves when you change height, you can trust the trend, even if the exact numbers differ. If you add traps or loading coils later, model them as loads or as inductor/capacitor elements and keep the real component Q in mind. A 40/20 trap dipole is shorter and works on two bands without a tuner, but the traps add loss and narrow bandwidth. For a first build, a plain 40 m dipole used on 20 m through a tuner on ladder line is simpler to build and to debug, though its 20 m pattern will have more lobes. If you move on to a vertical, the radials need more attention than the radiator does. Elevated resonant radials, usually two to four, need less wire than a buried or ground-laid field, and sloping them downward raises the feed impedance toward 50 ohms. Ground-mounted verticals without a good radial field lose a lot of power as ground loss, so budget the time for that.
-
Getting real results at 5 watts: rigs, antennas, calling frequencies and LiFePO4 power for portable QRP
On antennas, the three common portable choices each have trade-offs. An end-fed half-wave with a 49:1 transformer is quick to deploy and works on its fundamental and harmonics, but efficiency depends heavily on the feedline and the counterpoise. A short counterpoise wire or a few feet of coax acting as one can change the SWR and the pattern. A linked dipole with a 1:1 choke at the feedpoint is a little more work to set up but more predictable, because each band is resonant and you only change the links. A simple inverted-V on a light mast or tree is a good compromise, and a flat-top dipole at a good height will usually beat a lower, bent one. For QRP, loss is the enemy. Use thin low-loss coax, keep the run short, and avoid a lossy tuner in the line if a resonant antenna will do. A choke at the feedpoint or the rig end reduces common-mode current, which keeps RF off the keyer paddles and the rig, and can quiet the receive noise floor. On batteries, LiFePO4 is popular because the voltage stays fairly flat across most of the discharge curve. A 4-cell pack is nominally 12.8 V and charges to about 14.4 to 14.6 V. Check your rig's maximum input voltage before connecting a fully charged pack, because some rigs have a limit not far above that. A fuse close to the battery, polarity-protected connectors, and a proper LiFePO4 charger are worth the small cost. A small pack from a few amp-hours is plenty for a few hours of 5 W SSB or many hours of CW, since receive current in these rigs is low.
-
Hamfest game plan: what to bring, what to test, and how to get more out of your local club
For anyone newer to this, I'd put the club tables and forums at the top of the list, not the bargain bins. A radio is only useful if you know what to do with it, and clubs are good for that. Ask about Elmer programs, whether the club does licensing classes or has a VE team, and how they use their repeater. Many clubs are glad to explain what they do and to invite you to a meeting without any commitment. If you're thinking about a first HF setup, the forums can help more than another piece of used gear would. A talk on simple wire antennas, grounding and RF safety, or logging software can save you from buying the wrong thing. And if a club is building something, like a portable go-kit, a Field Day station, or a repeater upgrade, ask how you can help. It's one of the quickest ways to learn practical skills and meet people who've already solved the problems you're about to run into.
-
Keeping RF out of ESP32 and Pico shack controllers: CAT, PTT, antenna switch and rotator interfaces
Treat the cables as antennas first and the code second. Most of these failures come from RF getting onto the USB cable, sensor lead, rotator cable or relay control wiring and then being rectified inside a chip. A few ferrite cores on a lead rarely fix that on their own, but they help when the type matches the band. Common-mode chokes made from several turns of the cable through a suitable mix (31 is a popular choice for HF) usually do better than a single clip-on bead. The same goes for sensor and rotator wiring. Put a small series resistor and a bypass capacitor right at the microcontroller pin, and keep the capacitor lead short. For a rotator position pot, an RC filter plus averaging many ADC samples in software cleans up a lot of jitter. If you use the ESP32's built-in ADC, remember that it's not especially linear and that ADC2 can conflict with Wi-Fi. An external ADC like an ADS1115 over I2C is often easier for position feedback. Also separate the supplies where you can. Relay coils, rotator motors and the logic should not all share one thin wire and a single regulator. Put flyback diodes across relay coils, and drive the coils through a ULN2003-type driver or a MOSFET, not straight from a GPIO pin.
-
Callsign groups, sequential vs. vanity, and what actually happens when you want a different call
The most common vanity mistake is listing only one or two calls. Because you can submit a whole ranked list, it's worth building a long one that mixes best-case calls with realistic ones. The system works down your list and grants the first call that's actually available and that your class permits. Putting a Group A call first when you hold a General isn't harmful, but it won't be granted, so don't count on it to filter anything. Also check each candidate's status in ULS before you apply. A call that shows as expired or cancelled may still be inside its two-year waiting window. And if you're hoping for a deceased relative's call, the rules for who can request it during that window are specific. Read the FCC's language rather than assuming a family member automatically qualifies. One more thing that trips people: the new call appears on your ULS record, but your old call keeps working until the change is granted. After that you ID with the new one. There's no overlap period where either is fine.
-
Starting Morse code from zero: Koch vs. Farnsworth, which trainers to use, and how do you get past the plateaus?
I'd add structure and community on top of that, because the tools only get you so far. CW Academy, run by the CWops group, has free online courses with volunteer advisors and small classes. A beginner course starts from nothing and focuses on copying by ear, and having a person set homework and check your progress helps with the problem most self-taught learners hit, which is stopping. Check the current signup periods and requirements on their site, because classes run in sessions and I can't confirm the dates for you. On the plateau question, there are a few common ones. The first usually shows up when the character set is nearly complete and you start copying real words. Letters that were fine in isolation fall apart because the next character arrives before you've finished processing the last one. The cure is to copy behind: let a character or two go by before you write it down, so you're recognizing the sound as a pattern and not racing it. Copying behind is the bridge to headcopy. To start headcopy, try it on words you can predict. Take plain text practice, or a callsign list, and don't write anything down. Just follow along. You'll lose the thread often at first, which is fine. Mix both modes: paper for a few minutes to check yourself, then head only for a while. Eventually whole words become the unit you hear, and that's the point where CW starts to feel like language. Flashcards and phone apps are fine as a supplement, but treat a card as a prompt for sound, not as an image. A card showing "K" with the audio playing is more useful than a card showing "-.-". An app is good for the five minutes you'd otherwise waste, but it shouldn't be your main practice, because the main practice needs real timing and not much visual feedback. For confidence, the biggest step is to separate learning from performing. You can be solid at 15 wpm in practice and still freeze on air, because the sending has to be right and the other station isn't predictable. A few things help. Listen to real CW on the bands before you try to call anyone, so you get used to fading, noise and different fists. Write out your standard exchange (callsign, signal report, name, location) and rehearse it until it's automatic. Start by answering a CQ from someone sending at a speed below your comfort limit. Many operators will slow down if you ask politely. And remember that nobody expects perfection from a new operator: asking for a repeat with a short "?" or "AGN" is normal practice. If you plan to key it yourself, practice sending too, ideally with a paddle or straight key, and record yourself to hear whether the spacing is even. Poor spacing makes you harder to copy than a slightly slow speed does.
-
What makes a rig a keeper: boat anchor charm vs. modern SDR convenience?
One thing I'd add on the boat anchor side is that condition matters more than model. Two radios of the same type can be completely different propositions depending on whether they've been stored dry, modified, or worked on by someone who left a mess inside. Cut-up chassis, drilled panels and hacked-in accessories can lower both the usability and the value, though some people don't mind if it works. For classifieds, I'd lean toward sellers who show the underside of the chassis and mention what has already been replaced. Phrases like "works great" without any detail don't tell you much. It's also fair to ask whether the seller has checked all bands, the transmit output, the metering, and the calibrator or any built-in crystal oscillators, if the radio has them. As for what I'd skip: anything that looks like it has been stored somewhere damp, with rust, corrosion or a musty smell. Cosmetic issues are easy to live with, but water damage can turn a project into a money pit. A clean radio with a few faults is almost always a better start than a pretty one with hidden problems.
-
What makes a rig a keeper? Vintage tube gear vs. modern SDR in the shack
For me the dividing line is whether the radio still teaches you something. Modern SDRs are great tools, but a lot of the work is hidden behind menus and DSP. A tube rig makes you see the whole chain: preselector, loading, plate tuning, AGC behavior. If you want to understand why a signal sounds the way it does, that's useful. On the restoration question, I'd treat the dim bulb or variac as a safety measure rather than a cure. It limits current if there's a shorted capacitor or a failed rectifier, but it won't fix old electrolytics. If the gear has been sitting for decades, inspect it first. Look for bulging or leaking caps, cracked insulation on the line cord, and a missing or incorrect fuse. Then replace the obvious suspects before applying full power. It's slower, but it's cheaper than replacing a power transformer. For shipping heavy gear, double-box it and pull the tubes and wrap them separately. Receivers with big knobs and glass dials suffer most from transit, so say clearly in the listing whether you'll ship at all or want local pickup. Buyers of boat anchors tend to respect that.
-
Choosing a handheld: Baofeng vs FT-65R vs ID-52A vs TH-D75, plus what to sort out before an emergency
I'd add a point on the D-STAR side. The ID-52A and TH-D75 are strong radios, but compare them on how you'd use them. The TH-D75's APRS and GPS features and its extra band appeal if you travel or like data modes. The ID-52A is the more straightforward choice if you mainly want a D-STAR dual-bander. Either way, check firmware status and the current software or app support on the manufacturer's site before buying, since those details change over time. On emergency use, practice matters more than the model. Have the whole group program the same channel plan, label the channels clearly, and test the radios on both a repeater and simplex. Also try a simple drill with the stock antenna against your upgraded one, so you learn how much range you really gain and don't have to guess later.
-
Building a ham go-kit in tiers: what goes in the pocket, the grab-and-go bag, the car and the home station?
Regarding power, I'd keep the system simple and fused. A LiFePO4 battery, a fuse close to the battery, a Powerpole distribution block, and a charge controller that matches the panel and battery is easier to troubleshoot than a pile of adapters. Measure the actual draw with an inline meter on your own radios so you aren't guessing at runtime. Also remember panels lose output when partly shaded, so a small foldable panel in a cluttered yard may deliver much less than the label suggests. For antennas, I'd carry one that works for your most likely distance. For local VHF/UHF, a roll-up J-pole hung from a tree or a mast can make a big difference over an HT's rubber duck. For HF, an NVIS-style dipole for 40 or 80 m is a common regional choice, but it takes space, rope, and a way to get it up, so practice. Keep a spare length of coax and a small SWR or antenna analyzer if you can, since a bad connector is a common failure. And for family radios, a quick caution: if someone in the household isn't licensed, they can use FRS or MURS, but they shouldn't be transmitting on ham frequencies on their own. Keep that in the plan so no one grabs the wrong radio in a stressful moment.
-
How do you separate real ham radio news from hype? Building a routine for gear, band conditions, and FCC/ARRL updates
One habit I'd recommend is treating anything pre-release as "unconfirmed" until there is an order button or a manual you can download. Manuals are underrated for this. A PDF will tell you more than a promo page: actual menu structure, filter options, supported modes, power output by band, and connector types. If a rig claims great digital mode support, the manual usually shows whether it has a proper USB audio and CAT interface or just a data jack and a hope. For antennas, I'd look for a stated efficiency or at least a described design, such as trap, loaded, end-fed, or fan dipole. Marketing for compact multiband antennas often skips the tradeoffs. A short radiator on 80 or 160 meters can work, but the bandwidth will be narrow and the efficiency lower. That is physics, not a flaw in one brand. On firmware, I'd wait a short while after a release if the radio is your only HF rig. Read the forums and user groups for reports of problems with the new version. Then export your settings first, if the radio allows it.
-
Bench checklist before keying a first homebrew QRP transmitter (Tuna Tin 2, BITX, uBITX and similar)
Adding to the filter step: if you are using a kit with relay-switched low-pass filters, it is worth checking each band's filter on its own with a signal generator or a nanoVNA if you have one. A nanoVNA or even a tracking generator makes it easy to see the cutoff, and to verify the toroid winding counts and capacitor values were right. A filter built with a wrong capacitor value, or a toroid with one turn too many, can sit at the wrong frequency and look perfectly fine on a power meter. Also, for low-power work the classic SWR bridge (the Tandem match or a simple Wheatstone-style bridge) has a minimum drive level, because the detector diodes need enough RF to conduct. At 200 or 300 mW the reading can look worse or better than reality. If the bridge was built for 5 W and above, don't trust it at milliwatt levels. An antenna analyzer, or a nanoVNA connected to the antenna with the transmitter off, will give a cleaner picture of the match. On soldering: use flux, keep the iron clean, and heat the pad and the lead together. For toroid leads, scrape or burn the enamel and tin the wire before inserting. A lot of 'dead' builds come from one bad joint on a toroid or an electrolytic polarity error, so a slow second visual pass pays off.
-
Building a contest year: which ham radio contests to target, what the exchanges are, and how to avoid log submission mistakes
For a first contest year, I'd treat the exchange as the main thing to practise. ARRL Sweepstakes is the best example: serial, precedence, call, check and section sound like a lot until you've written it down a few times. Set up a macro for each field in your logger and run a few practice QSOs against recorded audio or a friendly local net before the real weekend. CQ WPX is easier to learn because the exchange is only RST plus a serial number. Prefix multipliers also teach you how to scan the band for new stations. CQ WW and ARRL DX are good for learning zones and states, but DX stations in the pileup can be fast, so be ready to copy a call correctly on the first or second try. One logging tip: set your station's power, category, and operator class in the contest setup before the first QSO. Fixing that after the fact is the sort of thing that causes a log to land in the wrong category.
-
Nets and special event stations: how do you find the right one, check in cleanly, and chase the certificate?
A practical way to pick a net is to decide what you want from it before you listen. If it's conversation, look for ragchew or club nets. If it's practice for emergency work, ARES or similar group nets will teach you procedure and tactical calls. If you want to learn message handling, a traffic net is the best fit, but expect more structure and a learning curve with radiogram format. Once you've picked a candidate, listen for two or three sessions without transmitting. Note the net's opening script, how net control asks for check-ins, whether there's a list or a roundtable, and how long the session runs. A net that sounds fast or formal isn't unfriendly. It's usually just efficient, and most net controls will help a new station if you say it's your first time. One caution: if a net's information comes from an old website or directory, confirm it with the sponsoring club or group before relying on it.
-
Sequential vs. vanity call signs: how the groups, ULS, and the two-year rule actually fit together
A couple of practical additions on the club and special event side, since they get mixed up with personal licenses. A club call is licensed to the club, with a trustee as the responsible licensee. If the trustee changes, the ULS record has to be updated, and it's worth keeping the trustee's contact information current because that's where FCC correspondence goes. Before a club chases a vanity call, check the trustee eligibility rules for that group, because club requests aren't necessarily treated the same as an individual's. For special event calls, the thing to keep straight is that they aren't a new license. The 1x1 call is a temporary designation that operators use for the event, and the usual Part 97 requirements still apply: identify at the required intervals, stay within your own privileges, and keep your log. Since coordinators schedule these in blocks, apply early for a popular date, and don't assume a specific 1x1 will be free. On the lookup side, there's a quirk when people talk about prefixes and suffixes. A prefix in the ITU sense is everything before the numeral, so KA in KA1ABC. Some operators use /P, /M, or similar suffixes for portable or mobile. The FCC doesn't require them for domestic operation, and they're optional, though operating abroad under a reciprocal arrangement can carry its own identification requirements. Check the relevant country's rules for that.
-
New to ham radio? Share your first QSO, first radio, and what you'd tell the next newcomer
Good thread idea. For anyone who's licensed but nervous about the first repeater call, here's a simple approach. Find a local repeater listing (your area's repeater coordinator or a club site is usually more accurate than a random app), program it into the radio, and just listen for a while. You'll pick up how local people check in, whether there's a regular net, and how much dead air is normal. When you're ready, you don't need to say 'CQ' on a repeater. Something like 'This is [your callsign], new ham, listening' is enough. If someone answers, it's fine to say you're new and ask for a signal report. Most operators are happy to help. Nets are also easy entry points, since the net control operator has a structure and will call on you when it's your turn. One practical thing: do a radio check on simplex with another local ham, if you can, before relying on a handheld in a emergency. It teaches you how far the rig actually reaches without repeater help.
-
Building a remote station: IC-7610/RS-BA1 vs SCU-LAN10 vs Flex vs RigPi, and how to handle latency, security and the rules
On latency, it helps to separate what each mode actually needs. SSB tolerates a fairly large round-trip delay because the conversation is half-duplex in practice. The annoying part is usually jitter, not the average delay. A jittery connection forces a larger audio buffer, and then the delay creeps up. Wired internet at both ends beats Wi-Fi or a congested mobile link almost every time, and codec and buffer settings in the remote software often matter as much as raw ping time. CW is the harder one. Streaming your audio sidetone back from the remote radio means you hear the delayed version of your own keying, which is disorienting. Better designs send keying as timed events to a keyer at the radio end, or let you monitor a local sidetone while the remote radio does the real keying. Full break-in is realistic only with a very low and stable round-trip time, so many operators accept semi-break-in for remote CW. For FT8 and other timed digital modes, run the decoder and the software at the remote site, either on a PC beside the radio or on something like a RigPi-style controller, and remote into that machine's desktop. Then the timing-sensitive part (clock sync, audio to the rig) stays local and only the screen traffic crosses the internet. If you stream audio from the remote end into a PC at home instead, the added delay and clock drift can cost you decodes. Keep NTP or another time source healthy at whichever end runs the mode. One architecture point on the radio options. With the IC-7610 and RS-BA1 or the SCU-LAN10, the vendor software is the glue, and you depend on it for audio, CAT and PTT behavior. With a Flex, the radio itself handles network streaming and SmartLink handles the connection setup. With a Pi-based controller, you have more flexibility and more to maintain. None of them is automatically best. Match the choice to how much maintenance you want to do at a site you may not be able to visit.
-
Getting the most out of 5 watts: rigs, antennas, power and where to find QRP activity
On batteries, I'd keep the point about LiFePO4 voltage in mind. A 4-cell pack at full charge can sit near 14.4 V or higher, and some rigs have an upper limit not far above that. Read your rig's manual for its maximum supply voltage, and consider a small regulator or a different pack arrangement if you're near the edge. The flat discharge curve is nice, but it also means you get little warning before the pack is nearly empty, so a cheap voltmeter on the pack is worth adding. For kits, I'd suggest a simple order for learning: start with a very simple receiver or a small single-band transceiver kit, then move to something like a multiband CW kit once you're comfortable with toroids and alignment. The Pixie is fun as a first project, but treat it as a learning exercise and don't expect it to work well on a busy band. If you want a rig to use reliably on a hilltop, a purpose-built portable rig with proper filtering is more rewarding.
-
Good on-air habits: CQ calls, breaking in, nets and sharing frequencies
On the shared frequency point, I think a good rule for new operators is "ask, don't assume." Silence on a frequency doesn't always mean it's free. You may be hearing only one side of a QSO, especially on HF with skip, or a net that's between rounds. Ask "is this frequency in use?" and then listen for a good while. A short wait costs you almost nothing, and it can save you from stepping on a QSO that took someone a long time to set up. For interference, it helps to separate "it's annoying" from "it's harmful." Someone calling CQ on top of you by mistake is a conversation, not a complaint. Move or politely tell them. Repeated, intentional jamming is a different situation, and the useful response is documentation rather than debate. Time, frequency, mode and a short audio recording, if your local rules allow one, give a coordinator or regulator something to work with. Also, check your own gear before blaming someone else. Poor receive performance, front-end overload from a nearby transmitter and bad grounding or cabling can all look like someone else's problem. A different antenna position, a filter or a ferrite can sometimes fix it without anyone ever having to talk. For first-timers, a practical tip: write out your first CQ or net check-in on a card, with your callsign in phonetics. Reading from a card is nothing to be embarrassed about. It stops your mind going blank when the radio answers.
-
Cheap SDR chain for HF and VHF: where dynamic range, noise and antenna choice actually limit you
On gain: don't tune it by how pretty the waterfall looks. A decent test is to disconnect the antenna and put a terminator on the input, then note the noise floor in the bandwidth you actually use. Reconnect the antenna. On a quiet VHF band with a reasonable antenna, the floor should rise by a few dB. If it doesn't move, you are probably running with too little gain to benefit from the antenna and the receiver's own noise dominates. If the floor rises a lot and strange carriers appear and move when you change gain, that is the classic sign of overload or intermodulation. Real signals stay put when you change gain, while intermod products often shift or change level out of proportion. The 8-bit ADC in an RTL-SDR gives a theoretical dynamic range of about 50 dB, and real devices get less. That is why a strong FM broadcast signal can wreck reception at 100-200 MHz even when you are not listening to FM. A broadcast-band notch filter ahead of the dongle often helps more than any software setting, because the strong signal never reaches the ADC or the tuner's mixer stage. For the 1090 MHz ADS-B band, a filtered LNA placed close to the antenna usually does more than a longer cable run to a better dongle. Bandwidth matters too. Thermal noise is about -174 dBm/Hz plus 10 log10(bandwidth), so a 2.4 MHz capture sits roughly 64 dB above that figure before the receiver's own noise figure is added. Narrowing the FFT resolution makes weak carriers stand out on the display, but it does not improve what the front end sees. Those are separate things.
-
Getting the most from 5 watts: rigs, antennas and power choices for portable QRP
On the kit side, I would separate learning builds from field radios. The Pixie is a good way to understand a simple transceiver, with a crystal oscillator, a mixer and an audio amplifier doing double duty. It is not a good benchmark for portable performance because the receiver is wide and the output is small. Something in the Hendricks style, or a TR-35 build, gives you a more selective receiver, better keying and a workable duty cycle for real contacts. Operating habits help as much as hardware. Listen before calling, use the QRP calling frequencies as a meeting point rather than camping on them, and keep exchanges short for contests such as QRP-ARCI events. Sending your call clearly at a speed the other station can copy is worth a few dB. A tuned antenna and a good CW or SSB technique will win over extra power most of the time.
-
Newly licensed? Share your first radio, first QSO, and the advice you wish you'd had
On mentors, don't wait for an Elmer to find you. Most clubs have a few members who like helping new people, but they can't tell what you need unless you ask. Go to a meeting, introduce yourself, and bring a specific question, like "How would you put up an antenna for 40 meters on a small lot?" or "Can someone show me how to set up my radio for a local net?" Specific questions get better answers than "Where do I start?" Club meetings can feel a bit cliquey at first, but many clubs also run nets, VE exam sessions, field days, and informal coffee meetups. Those are usually easier places to start conversations than the formal part of the meeting. If your club has a loaner antenna analyzer, a tower trailer, or a portable station, ask about them. Seeing someone else's setup is often the quickest way to understand what you actually want. As for first QSOs, don't worry if yours is short, nervous, or just a signal report and a callsign exchange. A good first contact is one that works, and you can go from there.
-
Digital mode setup baseline: time sync, audio levels, CAT, and what decode thresholds actually tell you
One check I'd add to the list is to look at the waterfall before blaming the decoder. If the noise floor looks uneven, with a strong slope or stripes, you may have receiver AGC or a noise blanker interacting with the signal. A noise blanker on a rig can distort narrow digital tones, so it is worth switching it off to see if decodes improve. The same goes for aggressive DSP noise reduction. Many rigs apply it before the audio reaches the USB codec, and it can smear weak tones that the decoder would otherwise recover. On time sync, the DT column is also a good indicator of the other station's clock. If many stations on a band show a similar offset, the issue may be yours. If only one shows a large offset, it is probably theirs. When replying to a station with a large DT, the decoder may still handle it, but that does not mean the other side is hearing you correctly.
-
Go-kit sanity check: what's in your grab-and-go, and what have you actually tested?
One suggestion for the testing question is to run a full drill with only what's in the kit. Use a weekend, shut off the main power to your radio corner, and run everything from the battery bank for a few hours. Log the actual amp-hours used, not the number on the label. Receive current, transmit current, and any extras like a display or a cooling fan all add up. That gives you a real runtime estimate to plan around. Also check the unglamorous parts. Do the Powerpole connectors match across all your cables? Is the fuse the right size? Can you find the right adapter in the dark? A lot of kits fail because a single pigtail or SMA adapter is missing, not because the radio is bad. For repeaters being down, programming a few simplex channels and agreeing on a calling frequency with your local group is worth doing before you need it. A repeater needs power and a link, and either can go out. Simplex only needs two radios in range of each other.
-
Remote station options compared: Icom LAN, Yaesu SCU-LAN10, Flex SmartLink, RigPi, and where latency, security and rules bite
On latency and modes, it helps to separate delay from jitter and to decide where the timing-critical work happens. FT8 and other timed digital modes should have decoding and encoding at the remote end, with that machine's clock disciplined by NTP or similar, instead of streaming audio across the internet and hoping the timing holds. The same logic applies to CW: if you send keying events in real time over a variable connection, element timing can wobble. A local keyer at the radio, fed text or buffered timing information, usually sounds cleaner than raw key-down/key-up over the WAN. Full break-in is the hardest case and may not be pleasant on a long path. For phone, compressed audio codecs keep the bandwidth modest, but the buffer size is a trade-off. A small buffer reduces delay and risks dropouts; a larger one is smoother but makes conversation feel stilted. Wired Ethernet at both ends and an uncongested uplink matter more than which brand of box you use. Mobile hotspots and busy Wi-Fi are where most bad experiences start. On rules, the details depend on your country, so read your regulator's text. In the US the relevant material is the telecommand provisions of Part 97, around §97.213. The general ideas are that there must be a control operator, the station must be protected against unauthorized operation, and there must be a means to end transmission if the control link fails. Operating privileges follow the licence of the control operator for that station and band. Contests and awards often have their own rules about remote operation and station location, so check them separately before assuming a remote contact counts. Also keep in mind that encrypting the control link is a different matter from what goes out on the air, where the emission must still comply with the rules on obscured meaning.