Measure harmonics

All levels are referred to the transmitter output: the calculated chain attenuation is in the tinySA's ext_gain, the frequency-dependent remainder is corrected for each measurement point. So what you see on screen is the level your transmitter puts out, not the level at the instrument.

A peak only counts as a harmonic if all three conditions are met:

1. Level: at least 12 dB above the median of the span. Not 6 dB: with around 200 measurement points, the maximum of pure noise sits 7 to 9 dB above the median all by itself.
2. Frequency: the offset from n·f0, divided by the order, stays below 12 kHz. The normalisation matters: a small error in the fundamental frequency grows with the order.
3. The level must not follow the noise floor. Hence the two runs 10 dB apart. The spans are centred on the nominal n·f0, and the tinySA slightly raises the level at the set centre, so a noise peak tends to land there and meets the frequency condition by chance.

If nothing is detected, the result is an upper bound (median + 12 dB), not "not present". If an excess is within the measurement uncertainty, it says "borderline", not "failed".

Which limit applies

The AFuV (German amateur radio regulations) gives no number. § 16 says unwanted emissions must be "limited to the lowest possible level" and refers to the guideline values of the Bundesnetzagentur (BNetzA, German regulator). The numbers are in the ITU Radio Regulations (in German: Vollzugsordnung für den Funkdienst), Appendix 3, table I:

43 + 10 log(P) is an attenuation below P watts and therefore a fixed absolute value, whatever the transmit power: 10^−4.3 W = 50.1 µW = −13.0 dBm. The relative limit only applies where it is less stringent, so above roughly 500 W above 30 MHz and above roughly 5 W PEP below. An example: 750 W PEP on 20 m may produce 7.5 mW, the same 750 W on 2 m only 75 µW.

The step is at 30 MHz, so the limit is determined per frequency and not once for the band: on 10 m the fundamental is below it and every harmonic above. The assessment is absolute and not in dBc, because that is how the limit is worded and because above the tinySA's mode boundary at 800 MHz the ratio is less certain than the absolute value.

The rules specify reference bandwidths: 10 kHz from 150 kHz to 30 MHz, 100 kHz up to 1 GHz, 1 MHz above. Here the measurement uses 10 kHz. For a harmonic, which is a narrow line, this gives the same power; a broadband spurious emission would be underestimated.

And what does ETSI say?

As a radio amateur, you are bound by the ITU Radio Regulations. Before it could be sold, your radio had to clear a second hurdle: ETSI EN 301 783, the harmonised standard for commercial amateur radio equipment. Its table 6 requires at the antenna connector:

For an unmodulated carrier that is the same number as under the ITU Radio Regulations: PEP and mean power are equal for a carrier, and 43 + 10 log(X) below X gives 50.1 µW again for any X. So both lines in the chart almost always lie on top of each other, and the report says so. They separate at exactly one point: at 30.000 MHz the ITU RR already counts as "above" and uses 70 dBc, ETSI still counts as "below" and uses 50 dBc.

The row for mobile SSB equipment is the least stringent of the three and is not applied here: this page cannot know whether you have such a radio, and when in doubt the stricter value applies.

Business radio device: then EN 300 086 applies

Not every handheld is an amateur radio. If it was approved as a business radio device, EN 301 783 is not the standard it was tested against, but ETSI EN 300 086, table 5a. And that one is much stricter, because it is worded in absolute terms and ignores the transmit power entirely:

That is 23 dB less than the ITU Radio Regulations allow. That's why each device records what it is approved as; you can change it for this one run in step 1. The report only ever shows one of the two ETSI standards, the matching one. Assessing against a standard nobody tested to says nothing about the device.

The equipment type is information, not a test. Enter an amateur radio as a business radio device and you get a "failed" with no basis. The other way round, a real excess goes unnoticed. The ITU Radio Regulations apply in every case and are always assessed too.

ETSI has a value of its own where the ITU Radio Regulations are silent: table 7 limits what a switched-on device emits while not transmitting: −57 dBm up to 1 GHz, −47 dBm above. That is exactly the state the pass without transmitting measures, and the table "Already present without transmission" classifies each line accordingly. It does not count towards the verdict: the same cable also picks up the broadcast transmitter next door, and you can't tell the two apart without disconnecting the device under test.

Limitations

• The attenuator chain is not calibrated against a standard. Comparing the two runs checks the linearity, not the absolute value.
• Only harmonics are measured. Non-harmonic spurious emissions (PLL spurs, mixing products) would only show up in a full sweep while keyed, and on this instrument that takes two minutes.
• There is no measurement above the specified range of the attenuators. Such harmonics are shown as "not measurable", not as "passed".
• Below about 5 MHz the tinySA shows leakage from its own oscillator, which the chain correction turns into a phantom value. That's why the overview sweep starts at 10 MHz.
• A result does not replace a measurement in a lab with a calibrated chain. It shows whether something is obviously wrong.
• A saved report keeps the verdict that applied at the time of measurement. If the rules or this page's calculation change, an old report is not reassessed.