If you weren’t under a thunderstorm last night you were probably impacted by its noise. Even so the band was quite active with a large number of WSPR reports being exchanged during the evening and overnight in North America and around the world. The geomagnetic field was stable overnight and the solar wind velocity has decreased from the previous session but persists above 400 km/s.
Roelof, PA0RDT, reports elevated noise during the session but Joe, VO1NA, was still audible in CW:
John, WA3ETD / WG2XKA, continues to see strong transcontinental propagation:
Ken, K5DNL / WG2XXM, reports that conditions were not great, decoding seven unique stations and being decoded by 35, including WE2XPQ and WH2XCR.
YV1DDH was monitoring WSPR from Venezuela. While we have had YV stations listening in the past, it has been a few years. Hopefully he will continue to listen as the path to South America should exist, particularly at the short distance from North America.
Regional and continental WSPR breakdowns follow:
There were no reports from the trans-African path. UA0SNV was present from Asiatic Russia but no reports were found in the WSPRnet database.
Two-way trans-Atlantic reports continue, including VE1HF’s successful return to the band after a hiatus:
Eden, ZF1EJ, and Roger, ZF1RC, successfully provided reports to North America in spite of the wall of noise in the central US:
In Alaska, Laurence, KL7L / WE2XPQ, is transmitting again and experienced a pretty good session, including reports for WG2XXM in Oklahoma. Also note a late report for VK3ELV in the previous session that was not uploaded. Those details are listed below in the report for Australia.
In Hawaii, Merv, K9FD/KH6 / WH2XCR, reports a weather front moved through his area, seemingly corresponding to good conditions on both 160-meters and 630-meters. The path to JA has returned and two-way reports persist with Australia. Merv also has reports for WG2XKA in Vermont, a very long way away from Hawaii.
In Australia, Phil, VK3ELV, was reported by Laurence, KL7L / WE2XPQ, in Alaska during the previous session but the upload failed.
Additional reports for Phil and Roger, VK4YB, include reports from WH2XCR and Phil received additional reports from Japanese stations:
Jim, W5EST, presents Part 4 in his series of recent discussions entitled, “ESTIMATE ATU SHUNT-COIL L-NET LOSSES 630/2200M WITH ANTENNA ANALYZER”:
“Summary: An antenna analyzer at ATU output can isolate RLOSS of a shunt-coil L-network by resonating the ATU coil(s) and measuring RLOSS. Then disconnect the shunt coil from ground and connect antenna analyzer to the coil end. Measure total RLOSS + RSYSTEM, using antenna analyzer across the combination of just-disconnected shunt coil to loading coil through ATU output to the antenna system, relative to ground. Subtract RLOSS from RLOSS + RSYSTEM to get RSYSTEM. See accompanying illustration.
A shunt-coil ATU with no series loading coil may use less inductance in its shunt coil than a series loading coil would need. This may matter on 630m and especially matter on 2200m. Less inductance trades off with probable need of a vacuum capacitor to implement shunt-coil ATU input capacitance. ATU losses appear comparable either way.
Remarkably, two different types of L-network can match the same MF/LF antenna. The “Hands-On Radio” column in QST (April, 2016, pp 61-62) backgrounds the topic and tells how to use a Smith chart to understand and design L-networks.
Let’s build on that article to learn more about power losses in L-networks for an example of LF/MF antenna system impedance 25 – j3000 ohms. First, background: 25 – j3000 ohm impedance is a right-side (east) Smith chart point very slightly south of east near the circular outer boundary of the chart. Plot that impedance point on p. 61 of the QST article on each of the two Smith charts for types “B” and “C” L-networks and use the techniques it tells to help get some intuition for what works.
A Smith chart helps understanding but its busy east-end may be difficult to use for graphical estimation of numerical values, especially for LF/MF TX antennas. Then a network calculator can help you, like Wetherell’s calculator cited by that QST article. https://home.sandiego.edu/~ekim/e194rfs01/jwmatcher/matcher2.html .
To use Wetherell’s calculator, set Source resistance/reactance to 50 and 0. Set Load resistance/reactance to the antenna system values your antenna analyzer measurements show, such as 25 and -3000 in an example here. Set frequency to 475700 or 137500 depending on choice of band, 630m or 2200m. Leave “Desired Q” input blank. To read the results from its display boxes, remember 1mH = 1,000,000 nH. 1uH = 1000 nH. Ignore the calculated Q values unless you understand what they mean.
Look at the calculated network types that interest you. They are likely to be simpler ones that calculate to non-negative inductance (nH) and capacitance numbers (pF). The network you choose needs to be economically practical to implement at MF/LF transmit power TPO levels.
You’ve designed and implemented your ATU. Measure RLOSS and the total RLOSS + RSYSTEM using the method of today’s illustration.
If you have a T-network with mostly series loading coil (March 22 WG2XSV ATU this blog), then use the formulas from yesterday’s blog. See if PATU/TPO = RLOSS/(RLOSS + RSystem) > 5%. Since the coil is mostly series loading, that estimation method should be sufficient.
If you have no series loading and all shunt loading instead, then how does one estimate ATU power dissipation? Now RLOSS is no longer in series. The above resistance ratio formula doesn’t apply. You could think P = I2R, but ATU RF input current Iin and RF output current Iout might differ in phase and complicate the net current I in the shunt inductor to use in P = I2R.
Suppose with an antenna analyzer you measured RLOSS = 0.8Ω using the method in the illustration, and measured RLOSS + RSystem = 25.8Ω. For the example, I’ve reduced RLOSS from 1.5Ω in the series loading coil design (yesterday’s blog) by the proportion 584uH of shunt inductance here to a 1000uH, j3000Ω loading coil that series-canceled antenna reactance.
This time, use Wetherell’s calculator on 25.8-j3000 Ω to match 50Ω coax. Results: Shunt inductance is 584 uH and input capacitor 80 pF at 475.7KHz:
XL = 2πf L = 2.985×106x 584 x 10-6 = 1743Ω inductive reactance, 584 uH shunt.
XC = 1/(2πfC) = 1/ (2.985×106x 80 x 10-12) = 4188Ω capacitive reactance of 80 pF.
XANT = 3000Ω capacitive antenna reactance, as given for the example.
The big MF/LF reactances control the current division of input current Iin into a current I in the lossy shunt inductance and into the output current Iout in the ohmic antenna/ground system loss. Take the ratio PRL/PRS of power losses I2RLOSS/(Iout2RSYSTEM) to get:
PRL/PRS = (XANT/XL)2(RLOSS/RSYSTEM) = (3000/1743)2(0.8/25) = 0.0948.
Now TPO = PRL+PRS because the TPO gets entirely used up in the sum of ant/gnd system losses PRS plus ATU losses PRL. (Ignoring input capacitor loss.)
What fraction of the TPO gets dissipated by ATU loss in a shunt-coil design?
PATU/TPO = PRL/( PRL+PRS) = 0.0948/(0.0948+1) = 0.0866 or about 8.7% ATU loss %.
The ATU input sees higher equivalent series loss resistance RATU compared to RLOSS viewed at ATU output. The 0.8Ω = RLOSS of the shunt coil gets effectively multiplied by about 5 in this example:
RATU = 50Ω x PATU/TPO = 50 x 0.0866 = 4.33Ω as viewed from ATU input. Not 0.8Ω!
Output current Iout in this shunt-coil design is found by considering input current division in parallel impedances:
Iout/ Iin = XL/(XL-XANT) = 1743/(1743-3000) = -1.387 (minus, opposite phase).
Iin = sqrt(TPO/50Ω) = sqrt(100w/50Ω) = 1.41A rms.
Iout = Iin XL/(XL-XANT) = 1.41A x (-1.387) = -1.96 A rms, a bit less than lossless ATU’s -2.0A.
Ishunt = 1.41 + 1.96 = 3.37A rms.
Voltage drop VC = IinXC across 80 pF input capacitor C suggests what larger voltage rating will be needed:
VC = IinXC = 1.41A x 4188Ω = 5.9KV rms = 8.33KV peak.
A vacuum variable capacitor can implement 80 pF at say 10-20KV voltage rating. But it’s pricey unless you have an alternative source like a junkbox or a hamfest bargain. The shunt coil would be subject to high antenna voltage, same as a series loading coil would be.
A series loading coil can largely voltage-isolate other component(s) that lie nearer the ATU input in a series-coil ATU design. However, the shunt-coil 630m ATU I’ve calculated probably depends on you affordably obtaining a vacuum variable capacitor.
However, the shunt coil can involve smaller inductance than a series loading coil to do the same match. If winding a high inductance coil is more inconvenient the bigger it gets, but you have plenty of transmit power TPO and can affordably get a vacuum variable capacitor, then a shunt-coil ATU may be of interest to you.
Since 2200m demands considerably higher inductance than 630m, the less-imposing inductance of a shunt coil ATU may be attractive on 2200m. To match the same antenna system at 137500 on the calculator use RSYSTEM=25, XANT=-10380. Calculator accepts up to 4 digits so use X = -9990. Shunt L= 6.78 mH, input cap C = 82 pF. By contrast, a series loading coil ATU calculates series L = 11.6 mH, input shunt cap C = 23.1K pF.
I’ve not tried to compare which of shunt-coil ATU or series-coil ATU might have more critical tuning. Breezes and wind move your antenna/tophat, so ATU circuit sensitivity can matter. Feel free to let us know if this is a careabout for you.
What can your experiences tell us? How do they compare with the many numbers estimated here? Do share your seasoned wisdom with this blog!”
Additions, corrections, clarifications, etc? Send me a message on the Contact page or directly to KB5NJD <at> gmail dot (com)!