…No, its not the name of the latest Broadway show, but just about everyone in North America seemed to experience it in one way or another. Storms in Oklahoma and Texas anchored another strong line North to the Canadian border as this weather system progressed to the East. Additional storms formed pockets across North America so listening was a challenge. Based on comments and observations, propagation was good which may have resulted in the QRN mess we experienced as lightning noise easily propagated from coast to coast.
Geomagnetic conditions were quiet with a variable Bz and elevated solar wind velocities averaging 430 km/s, down from over 500 km/s as the session progressed.
Larry, W7IUV / WH2XGP, reported very high noise levels in Washington state where little was heard.
Doug, K4LY / WH2XZO, picked a fine night to return after completing his transmit antenna work. He had 435 decodes from eleven unique stations. Doug indicates that more improvements are coming this Summer.
Phil, VE3CIQ, is still running 3 dB less power and seeing good results in spite of +10 dB noise levels. Phil also has done some work with his station, adding some components that will no doubt be detailed and reported in the coming days. He provided the following map of his session activity:
Wolf, DF2PY, reported static crashes in Germany during his CW session on 472.5 kHz at 1920z.
Ken, SWL/EN61, in Indiana reported four unique WSPR decodes through a tough night of noise.
John, VK2XGJ, reported good band conditions in New South Wales. Northern and Southern Hemispheres seem to be trading off good band conditions over the last few session.
Neil, W0YSE/7 / WG2XSV, was also fighting high noise levels last night and reports an interesting anomaly where obvious WSPR signals were not decoded. This behavior is one example where QRSS might offer superior performance to WSPR. Neil provided these comments:
Regional and continental WSPR breakdowns follow:
There were no reports from the trans-Pacific, trans-Atlantic or trans-African paths. UA0SNV was present during the session but no reports have been filed at this time.
Eden, ZF1EJ, reported my stations and WH2XZO on Doug’s maiden voyage after antenna maintenance.
Laurence, KL7L / WE2XPQ, operated through a session that never really got dark and exchanged reports on the reliable path to southwestern Canada and northwestern US.
Merv, K9FD/KH6 / WH2XCR, was not active during this session so there are no reports for Hawaii.
Jim, W5EST, presents a multi-part discussion entitled, “PART 1: VOLTAGE AND POWER ACROSS A SPECTRUM IN THE 630/2200M BANDS”:
“The May 17 blog discussed rms, average, peak, and peak-to-peak current or voltage measures. The way some of them relate to each other and to power can depend on whether the waveform is a sine wave (single frequency) or not. What happens when there are multiple frequencies?
Fortunately, if a 630m transmitter conveys very little power in harmonics to the antenna system, there is no practical reason to determine the combined power of harmonics and main signal. Harmonic suppression instead is measured in relative terms–dB down from the main signal—and the more suppression the better. You don’t want a 2nd or 3rd harmonic of 630m that would land in the broadcast band!
Turning now to multiple frequencies in-band on 630/2200m, modes like CW, QRSS, WSPR and several other modes stay on the same frequency for tens or hundreds of milliseconds, or more, at a time before shifting to some nearby frequency for a similarly long time. You can speak of current and power values even when the signal assumes one frequency after another like WSPR does in its 6 Hz bandwidth.
If there’s a duty cycle D where the RF signal is on and off, then the average power P compared to the on-power is P = D x PON . Generally on MF/LF, however, one simply states PON itself (in dBm or watts) and duty cycle D itself, as in WSPR transmit percentage TxPct.
Okay, so we essentially regard multiple transmit frequencies less than a few Hertz apart as if they were a single frequency, right? Generally, yes, on the 630/2200m transmit side.
It’s the receive side I want to emphasize today and tomorrow. NOISE! Band noise is spread all across the dial. So are lightning and static crashes. Multiple noise sources from the sky, ground wave noise from the local region, neighborhood and home noise. Random noises (of various noise colors), and deterministic noises. Noises from AC power lines, cars, motorized appliances, TVs, computers. Noise plus signal itself. Regarding noise, how do we talk about voltage, current and power?
Keep in mind that the concepts of rms, average, peak, and peak-to-peak pertain to the waveform–the graph of voltage or current versus time. When talking about noise, I’ll focus on rms and power because, no matter what the waveform of signal and/or noise, rms governs these formulas:
P = Irms2 R = Vrms2/R
P(dBm) = 10 log10 (P/.001 watt) = 10 log10 (P watts) + 30 dBm
Suppose you have a 1 microvolt rms signal S at the 50Ω input to your RX. How many dBm?
P(1uVrms ) dBm = 10 log10 [(Vrms2/R)/.001 watt) = 10 log10 [(10-6 voltsrms)2/50Ω)/ 10-3 watt)
P(1 uVrms) dBm = -107dBm.
P(10 uVrms) dBm = -87dBm.
P(0.1 uVrms) dBm = -127dBm, etc.
How do we calculate for 1 microvolt rms of noise N? Same way! Waveform doesn’t matter.
But wait—what happens if you filter the noise? Usually, noise is such a squirrelly mess that it’s spread evenly over a range of frequencies in the receiver bandpass. So we conventionally, arbitrarily, speak of the noise power in a 2.5 KHz reception bandwidth. Signal-to-noise ratio is the difference in dB of the signal power minus the noise power.
SNR(dB) = S(dBm) – N(dBm)
WSPR SNR is SNR referenced to noise power in a 2.5 KHz bandwidth. That’s 2500 Hz bandwidth, which is far wider than the entire 200 Hz WSPR2 band on 630m or 2200m.
The WSPR decoder gives a lot of negative dB reports because it measures noise power as if received across that 2500Hz bandwidth. If you have a receiver that can filter IF noise and/or audio noise to less than that 2.5 KHz, 2500 Hz, then the WSPR decoder simply assumes the reduced noise still covers 2500 Hz and increases or artificially strengthens the reported SNR accordingly.
Negative WSPR SNR does not mean, though, that the signal is necessarily buried in the noise on an RX waterfall or spectrum display. Suppose the software can resolve the spectrum into 10 Hz segments or bins, one of which contains your desired signal. A -20dB SNR signal, the way that the WSPR decoder reports it, may nevertheless visually poke up about 4dB above the noise level on the display. That’s because 4dB ~= -20dB + 10 log10(2500Hz/10Hz). There’s one-250th the noise power in the 10Hz bin than there is in a 2500Hz bandwidth, but signal power is just as much in the bin as in a 2500Hz bandwidth.
Let’s talk about how noise combines with other noise in a further blog post. In the meantime, your questions and experiences are invited. Do write us!
Additions, corrections, clarifications, etc? Send me a message on the Contact page or directly to KB5NJD gmail dot (com)!