Foundations of Amateur Radio
The other day fellow amateur Randall VK6WR raised an interesting question. Using his HP 8920A RF Communications Test Set, which you might recall from our adventures in measuring radio harmonic power in 2023, that report is on my Github repository, but I digress, Randall wondered if the signal strength he was seeing on several radios were the same and discovered that in fact they were not.
It made Randall ask who set the standard and following on from that, what does this look like in the real world?
In 2014, episode 149 of the series "What use is an f-call?", I published an article titled "The simple S-unit". In it I referred to a standard for S-units defined in 1981. Unfortunately, I didn't provide any references, so, armed with more than a decade extra experience, Randall encouraged me to investigate.
Twenty seconds into my search, I discovered IARU Region 1 Technical Recommendation R.1, which has four statements related to the topic at hand.
Under the title "STANDARDISATION OF S-METER READINGS" it states that:
1. One S-unit corresponds to a signal level difference of 6 dB,
2. On the bands below 30 MHz a meter deviation of S-9 corresponds to an available power of -73 dBm from a continuous wave signal generator connected to the receiver input terminals,
3. On the bands above 144 MHz this available power shall be -93 dBm,
4. The metering system shall be based on quasi-peak detection with an attack time of 10 msec +/- 2 msec and a decay time constant of at least 500 msec.
So. Job done, right?
Yeah, nah, not so much.
The web page I quoted from is linked from the Wikipedia S-meter entry and was archived in 2005 and at the time existed on a Swedish domain in the home directory of Kjell SM7GVF. The page has two additional interesting things, the words "Brighton 1981" and "Torremolinos 1990", both of which refer to IARU conferences.
The reports for these meetings are online. In searching for any reference to the definition of the S-unit, the 1990 report shows that resolution "83-1" had the status of "Action completed", whatever that means.
The 1981 conference document has all manner of interesting references, including "Log Forms and Summaries for International Contest Use", "Meteor Scatter qso procedure" and the definition of the standard way to determine Morse Code speeds using the word "PARIS" followed by a 7 bit word space, to name three.
The one we're interested in is called "BM/134 - S-Meter Standards", appearing on page 33 and 34 of the 1981 report. It's a photocopy, so you can see the text from other pages superimposed. I'm making this observation because this is essentially a standards document, intended to be adhered to by industry and the amateur community.
It gets better, or rather .. worse.
The text that is referenced by Wikipedia uses numbers for the four elements, where BM/134 uses letters. The third item in BM/134 says that it applies for "bands above 30 MHz", but the document I just quoted appears to be unique in saying that it applies to "bands above 144 MHz".
The fourth item, dealing with the way that the meter responds has been altered on BM/134. The text "+/- 2 ms and a decay time" are in a different font and at an angle. Worth noting that the change includes "ms" twice, rather than "msec" as the unit for milliseconds used elsewhere.
Searching for a phrase within the standard, I discovered the Region 1 HF Manager Handbook v7.01, which appears to include the S-meter standard in chapter 11.1.2, but closer inspection reveals that the fourth item is missing, the one about quasi-peak detection. This is significant because the S-meter standard is based on a CW signal, not an SSB signal, which fluctuates.
There's no reference as to where or when this was removed or by whom. These changes are repeated in subsequent versions of the HF Managers Handbook.
There's other differences too, instead of using millivolt and microvolt as shown in the original BM/134 standard table, all units have been converted to millivolt for no discernible reason. The new table, including typo, is also copied everywhere.
While we're at it, the original standard contains the letters "V", "E", "R", "O", "N" at the top. They don't show in the HF Managers Handbook either. This is curious, since last time I checked, those letters signify an organisation that at least some here will recognise, the "Vereniging voor Experimenteel Radio Onderzoek in Nederland", known to the the people who don't speak fluent Dutch, as the peak body for amateur radio in the Netherlands, VERON.
Searching its website does not reveal their contribution to this standards document, which I have to say, is par for the course, much of our amateur radio history is poorly documented or archived, if at all, something which I've spent plenty of my time attempting to remedy over more than a decade, one article at a time.
Moving on.
The phrase I mentioned earlier bears reading out in its entirety.
From BM/134:
"We hope that the current recommendation will be followed by all equipment manufacturers, so that in a not too distant future one will know how to interpret the strength report of the other station."
It goes on to say:
"Societies should advise as much as possible their members about equipment manufacturers adhering to this recommendation and shall try to avoid publication of receiver designs which do not in principle use the recommended standards."
Which brings me to you.
What have you done lately about this? It's only been 45 years. Perhaps it's time to implement this?
I can tell you that preliminary results show that the S1 level sensitivity associated with the radios that Randall showed me are reporting S1 when really they should be reporting about S4.
You might wonder why this is the case?
Calibration appears to be the underlying cause.
While S9 itself is variable in accuracy, as-in some radios are more accurately close to -73 dBm, others are consistently 5 dB shy of that.
When an S4 signal is reported as S1, then the 6 dB step size is not correctly implemented.
Similarly, when an S9+20 is reported by a signal that's only 16 dB stronger than S9, there's more fudging going on.
The differences between a signal with and without preamp are also worth noting as being inconsistent.
In other words, the level and steps associated with S-meter units are all over the place, which is interesting, since the authors of the standard already alluded to this when 45 years ago they wrote:
"Simple means for calibration of at least the 6dB level ratio should be published."
While we're digging for causes. Why is there not a standards library associated with the IARU, where documents like BM/134 exist and with it their current level of application?
Before you tell me, money, I'd point out that in 1981, IARU Region 1 had 263,945.88 dollars, or francs, guilders, pounds or glass beads in the bank, it's unclear which, since there's no units stated. At the 1990 conference IARU Region 2 had a reported net worth of $150,000 with $40,000 annual income and Region 3 reported that "its finances are also sound".
How do I know? That's in those conference documents too.
Curiously, the 1981 report states that "when presenting the 1981 and subsequent accounts the Treasurer would also prepare a brief report which would explain the reasons for items of such expenditure for which the purpose was not obvious."
I wonder what happened to those explanations?
So, what is the current status of BM/134 and what are the legal implications of me publishing it on my Github page?
While we're at it. I realise that I can "fix" the Wikipedia page, but where do I point it at?
I'm Onno VK6FLAB
