The power figure that is not in the instructions

1Six things make a radio device legal to sell in Great Britain

If a device deliberately transmits radio and it is placed on the market in Great Britain, the Radio Equipment Regulations 2017 (S.I. 2017/1206) require six things. None of them are obscure. Every radio product you own has been through this.

1. A conformity marking on the equipment or its data plate, and on the packaging. Regulation 44.
2. A declaration of conformity, or a simplified one with a web address for the full text, accompanying every single unit. Regulations 13(3) and 43.
3. Technical documentation drawn up and kept for ten years. Regulations 9 and 11.
4. Instructions stating the frequency band and the maximum radio-frequency power transmitted. Regulation 13(2).
5. The manufacturer’s name and a postal address. Regulation 12.
6. The importer’s name and postal address, where there is one. Regulation 23. The regulations define an importer as a business established in the United Kingdom, so a seller shipping in from outside it is not one.

Five of those have to be with the product or on it. The sixth, the technical documentation, has to exist, be kept for ten years, and be produced if an enforcing authority asks for it. I am working through all six. This article is about the fourth one, and specifically the second half of it.

2Whose job is it to put the figure there?

Before anyone says that is the American manufacturer’s problem, it is not only. The duty to supply the figure is the manufacturer’s wherever they are, and selling direct from outside the United Kingdom does not shed it. Contracting to sell a finished product to a buyer in Great Britain is placing it on the Great Britain market.

And whoever sells it here carries a duty of their own. It is not a vague one, and it is not left to inference. It names the regulation this article is about.

3Six documents, and every page of their website

The current operator’s manual, thirty two pages. The previous revision of that manual. The manufacturer’s brochure. The AML Pro data sheet. The product guide the GB and Ireland seller hands to customers. The twelve page company catalogue. And the manufacturer’s own website, all four hundred and eighty one pages of it.

What they say instead

This is the page in the current manual that explains what the device does. The whole claim is on it, and so is the word they use in place of a number.

The sales brochure describes the transmitter the same way.

It is not reserved for the radio either. In the same brochure the word does duty for the battery compartment and for the USB socket.

So I counted. Five documents from the manufacturer and its GB and Ireland seller, one hundred and fourteen pages between them, searched for the word and for the number.

They print it for their other locators

It is not as though they are unfamiliar with the idea. Turn to page five of the same twelve page guide and their cable locators get a number in the opening line. The PL-VF3 has “a 3W power output”. The PL-VF10 is “equipped with a powerful 10W transmitter and optional 1/2 and 1/4 power outputs”.

Two pages earlier the AML Pro gets fourteen bullet points. Housing weight, sensitivity settings, headphones, battery, USB port, back-lit display, GPS, levelling meter, warranty. Not one of them is a power. If SSI have published it somewhere I have not looked, show me and I will add it to this article the same day.

4Go and look in your case

That is everywhere I can reach from the outside. Regulation 13 is about information that accompanies the equipment, so what matters now is what came in your box, not what is on a website. The figure to look for is a maximum RF power, in milliwatts, in dBm or as an e.i.r.p. Worth checking the rest while you are in there: a declaration of conformity, a conformity marking on the unit and on the case, the manufacturer’s postal address, and the name and address of whoever imported it.

5Then I asked a distributor, as a buyer

In August 2026 I wrote to one of their distributors as what I am, a leak detection company looking at the device for plastic pipe. Among four questions:

The reply was prompt and helpful, with a formal quotation and a datasheet attached. It answered the frequency, 2.45 GHz. It did not answer the power, and it did not mention that it had not. A second datasheet followed, produced by the distributor themselves, listing frequency, battery, dimensions, weight, operating temperature and IP rating. No power.

6Except on one page, in one language

Then somebody sent me a copy of the operator’s manual as it was issued in 2014. One hundred pages, revision 01.15.2014, SSI’s own artwork and their web address on the cover. It is a German-market copy: the back page has been replaced with a German distributor’s name and address. Inside it, the same manual runs in seven languages, one after another, each with its own technical specifications page.

The English pages are the ones marked with the American and British flags. Their specification table ends at the frequency. Turn to the German pages and the same table carries one more line.

I checked all seven. French, Italian, Spanish and both Portuguese sections give the frequency and no power, exactly as the English does. The word mW appears nowhere else in the hundred pages.

Somebody added that line afterwards

Now the part that cuts the other way. That line was not set by whoever typeset the manual. Every other label in that table is bold. Leistung is not, and it is set in a different typeface from every other word on the page.

That typeface appears on exactly two pages out of a hundred. One is this one. The other is the back page, which is where the German distributor has put their own name and address over SSI’s.

It is also the only line in a hundred pages whose text layer is broken, so a search of the file for “Leistung”, “power” or “mW” returns nothing at all. You only see it by looking at the page.

So there are two numbers. Here is what each one is worth

I do not know which of those is the true maximum, and neither does anyone else outside that company. Which is what happens to a figure the law says should have been printed in the first place.

7The number falls out of their own certification file

Set the paperwork aside for a moment. Somebody once put this device on a bench and measured what came out of it. The AML holds one FCC authorisation, granted in March 2012, and the Intertek report behind it is public.

What the laboratory measured

That report does not give a transmit power in watts either. What it gives is the measurement one is derived from. At 2450 MHz, the strongest emission recorded was 111.4 dBµV/m at three metres. Converting that to radiated power is one line of arithmetic any engineer can check. Power equals field strength times distance, squared, divided by thirty. Put the numbers in and you get:

That is a peak radiated figure, worked out from their highest measured field strength using the standard conversion. It is not a transmitter output measurement, because the certification never made one, and it comes from the sample they submitted in 2011, which is the certification they still claim for the product they sell today.

You do not have to take my word for any of it. The FCC record is public, and the grant and every exhibit behind it sit at fccid.io/A47-AML. Go and check the line yourself.

It is my arithmetic, not their declaration

Regulation 13(2) does not say the figure should be obtainable by anyone determined enough to find a fourteen year old American test sheet and work out which line matters. It says the manufacturer must supply it with the equipment.

8Forty one milliwatts, held in mid air

Hold the thing the way the manual tells you to. Handle parallel to the ground at waist level, at least eighteen inches out in front of you, so the blade hangs in the air with about half a metre of nothing underneath it.

Now the ground. Soil loss at this frequency varies enormously with how wet the ground is, so here is the case I am using and where it comes from: moist mineral soil at about twenty per cent water by volume, from the standard dielectric dataset for wet soils (Hallikainen and others, 1985). In that soil, at 2.45 GHz, the signal is down to a third of its strength after about three centimetres. Not three metres. Three centimetres, the width of two fingers.

Drier ground is kinder and wetter ground is worse. I have used a middling case, and it is a condition their own marketing names.

That works out at roughly three hundred decibels of loss for every metre the signal travels. And it has to travel every metre twice, because a signal that goes down and stays down is no use to anybody.

So how much has the device got to spend on the whole trip? About a hundred decibels.

Now be as generous as you like with that hundred decibels. Spend every last one of them on the ground alone. Ignore the air gap under the blade. Give it every benefit going.

Fifteen concrete walls for every metre of ground

Wi-Fi runs at almost exactly this frequency, and you already know what one wall does to it. It is why half the country has a repeater on the landing. A concrete wall costs a 2.4 GHz signal somewhere around twenty decibels. A single metre of moist soil costs about three hundred.

But it will still find the shallow ones, surely

It is a fair question, and I am not going to answer it with arithmetic I do not have. The sums above show that a claim cannot be met. They say nothing at all about what this device does a foot down, and I am not going to pretend otherwise.

But getting a signal back up is only the first hurdle, and it is the easy one. Whatever is down there still has to reflect differently from everything around it, and the device still has to tell that apart from the rock, the root, the void and the patch of reinstatement beside it. Nothing on the record shows it doing that at any depth. The one peer-reviewed test was unblinded, one operator, on ground where somebody already knew where the pipes were.

And shallow is the hardest place to prove anything, not the easiest. A foot down you are usually working along a visible scar, between a cover and a meter, over a strip somebody else cut and filled. Those are the cues that tell an experienced operator where to expect a result before the device is even switched on. The full investigation takes that apart.

A bigger transmitter would not help

The ground is what stops it, not the size of the transmitter, and no power you could fit in a handheld at this frequency would change the answer. Lower frequencies buy penetration and give up resolution, which is why utility radar meant to reach any real depth runs well below 2.45 GHz, and why it is run against the surface rather than held in the air. You keep it on the surface because the moment you lift it, the loudest thing the receiver hears is its own transmitter a few inches away, and the second loudest is the ground directly underneath. The laboratory report for this device records that it “transmits and receives simultaneously”.

9Why a number this small travels badly

I cannot tell you why the figure sits on one page in one language and nowhere else. Nobody outside that company can, and I am not going to pretend otherwise. What I can tell you is what the figure means once you have it.

Forget decibels for a second. Here is the job, described plainly.

Stand with the thing half a metre above the ground and switch it on. Now wait for a whisper of radio to travel down through the ground, brush past a plastic pipe, and climb all the way back up through it again.

That is the trouble with the number. On its own, a figure in milliwatts means nothing to most buyers. Set it next to what this device is sold as doing, and anybody who works with radio has the one thing they need to start checking the claim for themselves.

Which is exactly what regulation 13(2)(b) asks for. Not the frequency. They publish that everywhere, in every document, in every language. The power. Of the two numbers the law wants in the instructions, the one they print is the one that tells you nothing, and the one that is missing is the one that would let a buyer do the sum.

10So what does this actually mean on site?

Strip the decibels out and here is what is left.

Somebody is standing on a verge, about to tell the people behind them where it is safe to dig. That is the only thing this device is for, and it is the only reason anybody buys one.

One of the numbers they would need to judge that appears once, on a German page, in a typeface nobody else on that page used. The frequency is published in every language. So are the battery, the weight, the temperature range and the ingress rating.