Five Watts and a Wire

QRP is a Q code before it is anything else — it means "reduce power", and as a question it asks whether you should. At some point it stopped being an instruction and became an identity. The working convention is five watts or less output on CW, ten watts peak on voice, and under one watt gets its own name, QRPp. The obvious question is why anyone would deliberately throw away thirteen decibels, and the honest answer starts with admitting that thirteen decibels is exactly what you are throwing away.

The arithmetic, without the romance

Dropping from 100 watts to 5 watts is a factor of twenty, and ten times the base-ten logarithm of twenty is 13 decibels. The S-meter convention on HF puts one S-unit at 6 dB, so that is a little over two S-units. A station hearing you at S7 with a hundred watts hears you at about S5 with five. That is the entire penalty, stated plainly, and it is smaller than most people expect and larger than any amount of enthusiasm can close.

3 dB

Twice the power. Half an S-unit. Genuinely inaudible on the other end.

6 dB

Four times the power. One S-unit. The smallest change worth arguing about.

10 dB

Ten times the power. Around one and two-thirds S-units.

13 dB

Twenty times. Five watts against a hundred. Two S-units and a bit.

25 dB

Three hundred times. Five watts against a legal-limit kilowatt and a half — about four S-units.

Read that table the other way round and it becomes useful. To gain a single S-unit you must quadruple your power. Going from 5 watts to 20 buys you one S-unit. Going from 5 to 100 buys you two. Going all the way to the legal limit — a linear amplifier, a three-phase supply's worth of heat, and a small fortune — buys four. Transmitter power is the most expensive decibel in radio, and it is the one that people buy first.

Where the cheap decibels live

Almost anywhere else. Raising a horizontal dipole from an eighth of a wavelength above ground to a half wavelength can be worth four to six decibels in the low-angle directions that matter for long paths, and sometimes considerably more, because a low wire is not weak so much as pointed at the sky. A three-element Yagi gives something like five decibels over a dipole in its favoured direction. Together that is the difference between five watts and a hundred, paid for in rope and aluminium instead of in a power supply.

And there is an asymmetry that settles the argument. Antenna gain works in both directions. Every decibel you add to the antenna you also get on receive, and you cannot work a station you cannot hear. An amplifier does nothing whatever for your receiver. Spend the first money in the air.

What CW gives you that voice cannot

This is where the five-watt case actually gets made, and it is made with bandwidth.

A single-sideband voice signal needs about 2.4 kHz of receiver bandwidth to be intelligible. A CW signal at conversational speed needs well under a hundred hertz, and a practical receiver filters it at 250 or 500. Noise power scales with bandwidth, so narrowing from 2400 Hz to 250 Hz throws away roughly ten decibels of noise while keeping all of the signal. Ten log of 2400 over 250 is 9.8 dB, and it costs nothing but a filter.

Now put the two numbers together. Five watts of CW is 13 dB below a hundred watts of voice on raw power and about 10 dB ahead on detection bandwidth. The net is something like 3 dB — half an S-unit — and that is before counting the fact that a human ear tracking a steady tone works at a lower signal-to-noise ratio than a human ear reconstructing speech, which is worth several more.

There is a further point that gets left out of these comparisons. A hundred watts of SSB is a hundred watts at the peak of a syllable; the average power in a voice signal is a small fraction of that. Five watts of CW is five watts for the whole duration of every single element you send. The peak-power comparison flatters voice considerably.

So the QRP claim is not that five watts is as good as a hundred. It is that five watts of CW is roughly in the same conversation as a hundred watts of voice, and it gets there with a radio you can put in a coat pocket and a battery you can carry up a hill.

The culture, briefly

There are calling frequencies, and knowing them saves a lot of aimless tuning: 3.560, 7.030, 10.106, 14.060, 18.096, 21.060, 24.906 and 28.060 MHz are the ones in general international use, though North American practice has historically also used 7.040. Call near them, not exactly on them.

There is a tradition of measuring contacts in miles per watt, and an award at a thousand. Five watts across the Atlantic works out around seven hundred, which tells you that most of the impressive numbers come from turning the power down rather than from going further. There is a deep kit-building strand — the low-power end of the hobby is where people still design and solder their own transmitters — and a large overlap with portable operating, because low power and light weight are the same engineering problem viewed from two sides.

Some operators append /QRP to their callsign. It is common in some circles and quietly disliked in others, it is not required anywhere, and it makes your call longer in precisely the conditions where a short call helps. Make your own decision, but know that it costs you something.

The frustration, honestly

Most of a low-power session is spent calling. You will send CQ for twenty minutes into a band that sounds open while a station two S-units above you works somebody new every ninety seconds. That is not bad luck. That is the thirteen decibels, working exactly as arithmetic says it should.

Pileups are worse. When a rare station has forty people calling, the loudest three get through and you are not one of them. The tactics that work are patience and timing rather than volume: wait until the crowd thins, call at the very end of his listening period rather than the start, and if he is working split, listen to where he is actually picking people up rather than to where everyone is transmitting.

Fading will take contacts away from you at the worst moment, because your margin is small enough that a normal QSB cycle can drop you below the noise mid-exchange. You will lose contacts at the report stage and there is nothing to be done about it.

And the trap that catches most people: low power demands a better antenna than high power, not a worse one. Five watts into a compromise wire indoors is not QRP operating, it is a demonstration that the band is closed. People buy a small radio to save money, hang the cheapest antenna they can, conclude that CW does not work, and put both in a cupboard. The radio was never the problem. Read the antenna dispatch before you buy anything at all.

The other honest piece of advice is not to start here. Learn to make contacts at whatever power you have, get fluent at the exchange, and then turn the knob down. Starting at five watts with a new skill, an unproven antenna and no idea which band should be open stacks four unknowns into one silence, and you will not be able to tell which one is failing.

What you get in return is worth saying plainly. Operating at five watts makes you competent at everything else, because nothing is covered up. You learn the bands properly because you have to be on the right one. You learn to listen, because calling blind is expensive. You find the decibels hiding in your feedline and your antenna height because you cannot afford to leave them there. Operators who came up on low power are noticeably better at the whole craft, and the equipment is only part of the reason.