Restating a sub-terahertz band edge on the THz scale
Anyone working above 100 GHz ends up living in two notations at once. Waveguide catalogues, 3GPP study items and channel-sounding papers quote everything in gigahertz — D-band is 110–170 GHz, the IEEE 802.15.3d channel plan runs 252–325 GHz. Photonics groups, spectroscopy suppliers and most 6G roadmap slides switch to terahertz the moment the number passes three digits. Drop one value into the field below and the same carrier appears in both languages, so a band edge copied off a waveguide table lands on a THz axis without a stray factor of ten.
Why the two scales exist at all
The gap hardware left behind
Windows, not one continuous band
Where the label flips
Working a sub-terahertz band list into THz
Type the band edge exactly as printed
Waveguide and allocation tables rarely give round figures. Enter 118.75 for the oxygen line or 325.15 for the upper water-vapour peak; a comma decimal separator works as well as a dot, and any spaces you paste in are ignored.
Read the THz figure as you type
The second field tracks every keystroke, so stepping through a whole band plan means editing one number at a time rather than reloading. Below 1 GHz the output drops under 1e-6 and the display moves to scientific notation on its own.
Copy the bare figure into your sweep script
Each field has its own copy button, and Ctrl+C inside a field does the same thing. What lands on the clipboard is the plain number with no unit and no spaces, which is what a VNA setup line or a plotting axis wants.
Reverse it when a paper quotes THz
The swap button (↔) turns the page round so a 0.34 THz window becomes 340 GHz for your waveguide notes. Typing straight into the right-hand field has the same effect, and either dropdown can be searched for a different frequency unit entirely.
Millimetre-wave and sub-terahertz bands in both notations
The list below mixes the waveguide bands that test equipment is sold in, the two atmospheric lines that dominate this stretch of spectrum, and the windows that 6G channel measurements keep coming back to. Every THz figure is the GHz value divided by one thousand.
| Band or spectral feature | Gigahertz | Terahertz | Why it matters above 100 GHz |
|---|---|---|---|
| W-band (WR-10) | 75–110 GHz | 0.075–0.11 THz | Mature waveguide hardware; the practical floor of sub-terahertz work |
| Oxygen absorption line | 118.75 GHz | 0.11875 THz | A peak to design around, not through |
| D-band (WR-6) | 110–170 GHz | 0.11–0.17 THz | Leading candidate for 6G backhaul and fronthaul links |
| 140 GHz sounding window | 130–150 GHz | 0.13–0.15 THz | Heavily measured for urban microcell channel models |
| Water-vapour line | 183.31 GHz | 0.18331 THz | Strong enough to split the D and G bands into separate windows |
| G-band (WR-5) | 140–220 GHz | 0.14–0.22 THz | Where multiplier-chain sources start to get expensive |
| IEEE 802.15.3d plan | 252–325 GHz | 0.252–0.325 THz | First standardised channel set aimed at 100 Gbit/s wireless |
| Water-vapour line | 325.15 GHz | 0.32515 THz | Marks the top edge of the 252–325 GHz plan |
| 340 GHz window | 330–350 GHz | 0.33–0.35 THz | The first window most people describe in THz rather than GHz |
Absorption peaks kept to their real decimals
118.75 and 325.15 GHz become 0.11875 and 0.32515 THz with up to eight decimal places shown, so a line that a rounded figure would have hidden stays visible on the THz side.
Band edges converted as a pair
Run the lower edge, copy it, then run the upper edge; a 110–170 GHz waveguide range comes out as 0.11–0.17 THz without either endpoint drifting a decade.
Any frequency unit from either dropdown
Both selectors are searchable across the whole frequency list, so a source datasheet quoted in MHz can be pushed straight onto the same THz axis as the band plan.
Sub-terahertz spectrum questions
Where does the terahertz gap begin, and why is it called a gap?
Most literature puts it at 0.1–10 THz, which is 100–10 000 GHz on the other scale. The label is historical: electronic oscillators built on electron transport ran out of steam not far past 300 GHz, and semiconductor lasers only became practical in the infrared, so for a long time neither community had a convenient source or detector for the stretch in between. Some authors narrow the term to 0.3–3 THz, the part that stayed hardest longest.
Which atmospheric absorption peaks fall between 100 and 400 GHz?
Oxygen contributes a line at 118.75 GHz (0.11875 THz), sitting just above the broad 60 GHz complex. Water vapour then supplies the dominant features: 183.31 GHz (0.18331 THz), 325.15 GHz (0.32515 THz) and 380.2 GHz (0.3802 THz). The windows people actually build links in — around 140, 220 and 340 GHz — are simply the quieter gaps left between those peaks, and their depth changes with humidity.
Why do 6G papers still say 300 GHz instead of 0.3 THz?
Because the surrounding paperwork is in gigahertz. Regulatory identifications, waveguide band names, VNA extender specifications and channel bandwidths are all tabulated there, and a channel described as 275–296 GHz reads better than 0.275–0.296 THz. The switch to terahertz tends to happen when a figure spans decades or when the discussion moves to photonic sources, where the optical side of the lab already thinks in THz.
How much harder does the link budget get as the carrier climbs?
Spreading loss between isotropic antennas grows with the square of frequency, so every doubling costs about 6 dB and a factor of ten costs 20 dB. Over 100 m that is roughly 102 dB at 30 GHz, 112 dB at 100 GHz and 122 dB at 300 GHz — before any molecular absorption is added. The saving grace is that a fixed physical aperture gets electrically larger at higher frequency, which is why sub-terahertz demonstrations lean so hard on high-gain, narrow-beam antennas.
What actually produces a signal above 100 GHz in the lab?
Two routes dominate. Electronic sources take a microwave synthesizer and run it through a Schottky-diode multiplier chain — 24.5 GHz multiplied by six gives 147 GHz, 37.5 GHz multiplied by eight gives 300 GHz — at the cost of conversion loss and multiplied phase noise. Photonic sources instead beat two lasers on a fast photodiode; near 1550 nm, a 300 GHz difference frequency corresponds to an optical spacing of only about 2.4 nm. The photonic path gives wide tuning, the electronic path gives output power.
No comments yet. Be the first to comment!