NVIS Propagation: How to Work 100-500 Miles on HF

There's a gap in HF communications that nobody warns you about. Groundwave gets you maybe 30-50 miles on the lower HF bands. Traditional skywave skip can bounce signals 500, 1000, even 5000+ miles. But what about that donut hole in between? The 100 to 500 mile range where your signal goes over nearby stations' heads but doesn't travel far enough to come back down at DX distances?

That's where NVIS comes in. Near Vertical Incidence Skywave. It's not glamorous, it won't get you a DXCC award, and most contest operators couldn't care less about it. But for regional communications, emergency nets, statewide ragchewing, and served agency work, NVIS is absolutely the right tool.

What Is NVIS, Exactly?

NVIS works by shooting your signal almost straight up, where it hits the ionosphere's F layer and comes back down in a roughly circular pattern around your station. Think of it like bouncing a ball off the ceiling rather than throwing it across the room.

Conventional skywave propagation uses low radiation angles (5-20 degrees above horizon) to achieve maximum distance. NVIS deliberately uses high radiation angles (70-90 degrees) to achieve the opposite: reliable short-to-medium range coverage with no skip zone.

The signal goes up, refracts off the F2 layer (sometimes F1 during daytime), and returns to earth within about a 300-mile radius. Some operators report consistent NVIS coverage out to 400-500 miles depending on conditions, antenna design, and frequency.

What makes NVIS special compared to VHF/UHF for regional work:

• Doesn't need line-of-sight
• Covers mountainous terrain effortlessly
• No repeater infrastructure required
• Works when everything else is down
• Single antenna covers the entire service area

The Physics (Kept Brief)

I'm going to keep this practical, not textbook.

For NVIS to work, two things need to be true simultaneously. First, your operating frequency must be below the critical frequency (foF2) of the ionosphere overhead. The critical frequency is the highest frequency that will reflect back when transmitted straight up. Second, your antenna needs to radiate the majority of its energy at high angles.

The critical frequency varies with solar activity, time of day, season, and latitude. During daytime with moderate solar activity, foF2 might be 7-10 MHz. At night, it drops, sometimes below 4 MHz. During solar maximum years, daytime foF2 can push above 12 MHz.

Here's the practical upshot: frequencies between 2 MHz and 8 MHz are the NVIS sweet spot for most of the continental US. During the day, 40 meters (7 MHz) often works well. At night, 80 meters (3.5 MHz) and 75 meters (3.8 MHz) are more reliable for NVIS.

Nope. You can't reliably do NVIS on 20 meters. I see people claim this online, and while it might technically work during exceptional solar conditions, it's not dependable. Stick to 40m and below for consistent NVIS.

Frequency Selection: When to Use What

This is where most people overthink it. Here's what actually works in practice across the lower 48 states:

Daytime (sunrise to ~2 hours before sunset):

• 40 meters (7.0-7.3 MHz): Primary NVIS band during the day. Reliable, good bandwidth available.
• 60 meters (5.3 MHz): Excellent NVIS band. Limited to specific channels and 100W ERP, but propagation is very consistent.

Nighttime and twilight:

• 80 meters (3.5-4.0 MHz): The workhorse night NVIS band. Noise floor is higher, but signals are strong.
• 160 meters (1.8 MHz): Works for NVIS but atmospheric noise is brutal, especially in summer. Mostly useful in winter months.

Transition periods (dawn/dusk):

• This is when NVIS gets unreliable on any single band. The ionosphere is changing rapidly. Having the ability to shift between 40m and 80m during these windows is valuable.

I ran a week-long test last spring on 40 meters, making scheduled contacts with a station 280 miles away at 0700, 1200, 1700, and 2100 local time. Daytime contacts were solid every single day. The 2100 contact failed three out of seven days as the F layer degraded after sunset. Switched to 80 meters for evening skeds after that and never missed again.

The NVIS Antenna: Low Is the Point

Here's the thing that confuses people coming from DX operating: for NVIS, you want your antenna LOW. This is counterintuitive if you've spent years hearing "get it up as high as you can."

For maximum high-angle radiation on 40 meters, your horizontal antenna should be between 0.1λ and 0.25λ above ground. On 40m (λ ≈ 40 meters), that means:

• Ideal height: 4 to 10 meters (roughly 12 to 33 feet)
• Acceptable: as low as 3 meters (10 feet) off the ground

At these heights, a horizontal dipole produces a radiation pattern that's mostly straight up. Put that same dipole at 15 meters (half wavelength), and your pattern shifts toward the horizon. Great for DX. Terrible for NVIS.

Simple NVIS Dipole

The cheapest, most effective NVIS antenna is a half-wave dipole strung between 10-15 feet off the ground. Seriously. That's it.

For 40 meters:

• Total wire length: approximately 66 feet (use the 468/f formula, then trim)
• Height: 10-15 feet. Trees, fence posts, PVC mast, whatever holds it up.
• Feedline: 50-ohm coax. RG-8X is fine for 100W.
• Orientation: doesn't matter much for NVIS since the pattern is nearly omnidirectional at high angles

For 80 meters:

• Total wire length: approximately 130 feet
• Height: 8-15 feet (lower height relative to wavelength means even a low dipole works well here)

I built my 40m NVIS dipole using $8 worth of wire from the hardware store, a BNC-to-binding-post adapter as the center insulator (don't judge me, it worked for two years), and two trees about 70 feet apart. Fed it with 30 feet of RG-8X. My first contact was a station 190 miles north, S9+10 both directions, running 50 watts. Not ideal.

Wait, that's a lie. It was completely ideal for what NVIS is supposed to do.

The NVIS Dipole with Ground Reflector

Want to improve your NVIS signal by 2-3 dB? Add a ground reflector beneath the dipole. This is just a wire (or wires) laid on the ground directly under your antenna, cut to the same length or slightly longer than the dipole.

The reflector increases upward radiation and reduces the energy going into the ground. On a QRZ forum thread from early last year, W4xxx (not sharing the full call, it was a contentious thread) posted modeled results showing 3.1 dB improvement with a single reflector wire versus bare earth. My own A/B testing was less dramatic, maybe 1-2 S-units difference, but consistently there.

How to do it:

• Lay a wire on the ground (or a few inches above) directly beneath your dipole
• Cut it to 5% longer than your radiating element
• No connection to anything. Just a parasitic reflector.
• Multiple parallel reflector wires (spaced 1-2 feet apart) can improve performance further

Inverted V for NVIS

An inverted V with a 90-120 degree included angle at the apex, hung from a single center support at 15-20 feet, works well for NVIS. The drooping ends concentrate radiation at high angles. Bonus: you only need one tall support instead of two.

The trade-off is slightly higher SWR than a flat dipole (the angle changes the feedpoint impedance), but most modern rigs handle it fine, and you can always use a tuner.

Power Considerations

NVIS doesn't need high power. The path loss is relatively low since you're essentially bouncing off the ionosphere right overhead. The dominant factor limiting your signal quality is usually noise at the receive end, not path loss.

Practical power levels:

• 25-50 watts: Works for NVIS voice contacts within 200 miles in quiet band conditions
• 100 watts: Reliable for most NVIS work, including noisy evenings on 80m
• More than 100 watts: Rarely needed for NVIS unless you're fighting extreme noise or need to maintain a net through poor conditions

I run 75 watts on my NVIS setup for weekly 40m nets and have never had a readability complaint from stations within 350 miles. Save the amplifier money for a better antenna.

NVIS for Emergency Communications

This is where NVIS really earns its keep. ARES groups, RACES organizations, and state emergency management agencies rely on NVIS for intrastate communications when everything else fails.

Why NVIS dominates for EMCOMM:

• Covers an entire state from one station
• Works through terrain that blocks VHF
• No infrastructure dependency (repeaters can fail in disasters)
• Rapid deployment: a wire dipole takes 15 minutes to set up
• Reliable in disasters because the ionosphere doesn't care about hurricanes

The SHARES (SHAred RESources) HF network, used by federal agencies, operates NVIS frequencies for exactly these reasons. Many state ARES/RACES plans include designated NVIS frequencies on 80m and 40m for intrastate coordination.

If you're involved in emergency preparedness, having a tested NVIS setup with a go-bag antenna should be near the top of your priority list. I keep a 40m NVIS dipole pre-cut and wound on a kite winder in my go-bag. Deployment time with two helpers: about eight minutes from car trunk to on-the-air.

Common Mistakes

Antenna too high: If you hang your 40m dipole at 60 feet, you're doing great for DX and poorly for NVIS. For dual-purpose use, you need two antennas or a compromise height around 25 feet (which is mediocre at both but functional).

Wrong frequency at wrong time: Using 40 meters after dark for NVIS will frustrate you. The critical frequency drops at night. Switch to 80m.

Expecting 20 meters to work: It won't. Not reliably. The critical frequency rarely exceeds 12-14 MHz even during solar maximum, and at night it can drop below 5 MHz.

Ignoring ground quality: NVIS performance is affected by ground conductivity. Wet, rich soil helps. Rocky, sandy, or frozen ground reduces ground reflector effectiveness. A few radials or a reflector wire mitigates this somewhat.

Not checking foF2 before operating: Real-time ionospheric data is available from NOAA's Space Weather Prediction Center and sites like prop.kc2g.com. Check the critical frequency overhead before choosing your operating band. Takes 30 seconds.

Modeling and Real-World Performance

I modeled my 40m NVIS dipole in EZNEC (antenna modeling software, free version available) at 12 feet height over average ground. The model showed:

• Maximum gain: 6.2 dBi straight up
• 3 dB beamwidth: about 100 degrees (huge cone upward)
• Minimum gain at horizon: -8 dBi (which is exactly what you want, suppressing low-angle radiation)

Real-world results have matched this well. Stations within 50 miles are loud. Stations at 200-400 miles are consistently workable. Stations beyond 600 miles are rarely heard. The skip zone works in reverse for NVIS.

Dual-Band NVIS: 40m/80m Fan Dipole

If you want one antenna that covers both primary NVIS bands, build a fan dipole. Two dipoles, one cut for 40m and one for 80m, fed from the same feedpoint with a common coax run.

Construction:

• Cut a standard dipole for 80m (~130 feet total)
• Cut a second dipole for 40m (~66 feet total)
• Connect both at the center feedpoint, fanning the wires out at 6-12 inch spacing
• Hang the whole thing at 12-15 feet

The interaction between the two elements will require some trimming. Start long and cut. An antenna analyzer (NanoVNA works great for this, about $50 on Amazon) speeds up the tuning process enormously.

I'll get hate for this, but honestly? A resonant fan dipole at 12 feet outperforms a $400 commercial multi-band vertical for NVIS work every single time. Verticals radiate at low angles by design. They're geometrically wrong for this job. Don't let anyone tell you that you need to buy an expensive antenna for NVIS. You need wire and two supports.

Go String Up Some Wire

NVIS isn't complicated. Low horizontal antenna, frequency below the critical frequency overhead, and you've got reliable regional HF coverage that works through mountains, across states, and without any infrastructure. It's the backup plan that works when cell towers are down, repeaters are off the air, and you need to talk to someone 200 miles away.

Grab some wire. Find two supports about 10 feet high. Get on 40 meters during the day or 80 meters at night. Make contacts. It's that straightforward.