The small SWL beam (1...30 MHz)

Use this forum to discuss any radio-related topic outside of asking for advice (check out our Questions and Advice forum for those topics). As with all of our forums, treat others with kindness and respect.
Post Reply
User avatar
Andrew (grayhat)
Posts: 280
Joined: Sat May 28, 2022 5:56 am
Location: JN63pn

Re: The small SWL beam (1...30 MHz)

Post by Andrew (grayhat) »

forgot, if you have time, load the NEC model and run a sweep between 0.1 and 33.01, set the step at 0.2 and check the gain curve, then run another sweep between 80 and 130 and check the gain curve, finally run a pattern check at 90 MHz and look at the 3D lobes
13dka
Posts: 134
Joined: Sat May 28, 2022 4:27 am
Location: On or near a dike

Re: The small SWL beam (1...30 MHz)

Post by 13dka »

I"m out at the dike, forgot the power bank to feed the pre-amp. It appears like this thing really needs one, other than that, it"s a strange animal for sure, not quite as calculated and it's too soon to tell anything definitive but it seems to show some interesting traits. Will report tomorrow. 13dka out. :)
User avatar
Andrew (grayhat)
Posts: 280
Joined: Sat May 28, 2022 5:56 am
Location: JN63pn

Re: The small SWL beam (1...30 MHz)

Post by Andrew (grayhat) »

13dka wrote: Sun Jul 17, 2022 10:07 pm I"m out at the dike, forgot the power bank to feed the pre-amp. It appears like this thing really needs one, other than that, it"s a strange animal for sure, not quite as calculated and it's too soon to tell anything definitive but it seems to show some interesting traits. Will report tomorrow. 13dka out. :)
now I'm REALLY curious :) !

... to know if it's a BAZINGA or not :D
13dka
Posts: 134
Joined: Sat May 28, 2022 4:27 am
Location: On or near a dike

Re: The small SWL beam (1...30 MHz)

Post by 13dka »

Tl;dr: We will know more in a few days, sorry. :)

dirloop_dike_1.jpg
dirloop_dike_1.jpg (13.01 KiB) Viewed 2050 times

Long (!) version: Of course you know that one can't fully assess an antenna in one night, unless it's radically different from the antenna it's being compared with. I compared it with the LoG because I'm very familiar with its performance (and I needed the pole for the SDL) and the SDL is supposed to perform not radically different from that in terms of DX aptitude. The results of last night are quite mixed and need some interpretation. Let's start with the bad stuff:

1. Level

The missing preamp was making this comparison particularly difficult because the signals decreasing after sunset dropped to levels near (and below) the receiver noise floor on the SDL. Here's an example from the POTA waterhole around 14,300 kHz:





This may be different in high noise environments where a preamp doesn't do any good but in low noise conditions like at "the dike" it's obviously mandatory. Of course, the conditions were pretty weird too - how can sunspot numbers/SFI in the 150s, A=5, K=1 be worse than SFI=67 with no sunspots 2 years ago? I cross-checked with Kiwi receivers and called a friend who confirmed the mediocre condx, probably the aftermath of the repeated R1-R3 blackout conditions in the past days.

2. Directionality

This is really odd: I rotated the antenna like a champion (of which I have almost nothing "on tape" because that requires getting out of the car and squinting on the S-meter while turning the pole) and it generally worked on LW/MW and above in the evening. In the morning, on the other side of the greyline, I rotated my butt off and yielded zero results on the higher bands, as in the nulls were gone. 20dB F/B ratio is like 3 S-units and therefore hard to notice but the nulls are noticeable for sure, as demonstrated (OK difficult to see) in the next video, shot in the evening. Balun box pointing to the right = North, pointing to the left = South, station is CN8YZ in Morocco on 15m:




But these nulls seemed to have disappeared in the morning. Maybe it's just 20m where they don't show up so much in general. A different theory is that the antenna is very sensitive regarding the ground it stands on, and the tide is changing the close environment within the relevant 10 Lambda zone a lot.

Now to the encouraging things: First off, the SDL is indeed even worse than the LoG on LW, but it's actually better on MW! Here's an example showing more the unidirectional advantage than the better levels in the beam direction but this is quite nice - Radio Caroline playing the blues vs. Radio Murski Val (?) from Slovenia attacking Europe with accordeons on 648kHz, that's Southwest vs Southeast, the antenna turned Northwest:



dirloop_dike_2.jpg
dirloop_dike_2.jpg (15.24 KiB) Viewed 2050 times

3. Takeoff angle and SNR

The most puzzling and hard to assess question is the actual takeoff angle of the antenna. To answer that question for the entire range will take at least until fall, when transatlantic signals will return on MW. The simulation says that some more height may be an advantage for that on MW, which is something to keep in mind. We both know that the big advantage of directionality in reception is SNR, not level. The following examples represent this to some degree, I wish the level differences wouldn't be so big tho.

This is a Turkish station on vacation in Norway, causing a pileup among just of all Japanese stations:





Now at this point I didn't have the reference antenna yet, but I was puzzled by the unusual (to me) amount of Japanese stations I heard there, so I connected to a local Kiwi SDR for comparison. That SDR is not very good and it heard only the loudest guy from Japan, and a lot more of the Spanish guy calling. The SDL was turned NW and the Japanese guys were really quite loud on it, despite the general lack of level. So I hastily dropped the LoG onto the volleyball square, connected my new super bargain switch, obviously grabbed a defective coax cable, swapped that out and then the Japanese stations were all gone of course. Luckily the time was just right for the Tokyo VOLMET (JIA) on 13 MHz:





Of course the azimuth may have been just wrong for the LoG at this point (with its nulls theoretically pointing N/S) but the SNR difference is pretty dramatic, isn't it? Next in line was the weak evening signal from Radio New Zealand to confirm that:





Again, the SNL err... SDL had a noticeable SNR advantage. This advantage however reversed in the morning, when the direction of the signal likely changed to the long path (well, from here there's only 4000km difference between long and short path to NZ) and no rotation of the SDL could fix that:





Here's a number of videos shot in the morning with some more "true" DX, showing again how great the LoG is, and A.) that the SDL's output level sucks without a preamp....

ZL on 40m (!):


VK:


Uganda, VK and Idaho:



... and B.) that the SDL received all of these signals too. Also heard this morning were New Caledonia (new one!) and of course Hawaii and California, the latter with big signals. And this is the maybe least obvious but the most important takeaway from this night. Let me explain: I have spend all-nighters at the dike with all sorts of antenna contraptions including several small loops - the YouLoop with and without preamp, the ML-200 with small rigid and bigger flexible loops... With this experience I can tell with much certainty that (unpopular opinion) buying a pricey active loop for DX on shortwave is a massive waste of money. They are great <5MHz, above 10 MHz their steep elevation angles make the bands actually "close" after midnight and I have an old video somewhere showing how the puny whip on a Tecsun beats the expensive ML-200 on 15MHz WWV before sunrise. On the SDL, the signals dropped occasionally below the receiver noise floor due to the low passive output level but they were still there - all of them.

So my preliminary summary of the performance is a careful "promising". That a 2 square foot loop can do what the 20 square foot LoG can do and then some is certainly asked much but I'll see Tuesday or Wednesday night (when it's too hot to be at home anyway). I have yet to try it much more and practice putting it up: I know it's a symmetrical antenna with galvanic separation so this shouldn't matter much but I tried flipping it around indoors and got the impression that there is a tiny difference, so I tried to make sure I have the "RF" side up and the other one down by marking it. But there were tourists with their kids nearby and I was eager to get it up as quickly as possible...

dirloop_dike_3.jpg
dirloop_dike_3.jpg (4.57 KiB) Viewed 2050 times
:mrgreen:
Hank Michalenka
Posts: 10
Joined: Tue Jun 07, 2022 5:28 pm

Re: The small SWL beam (1...30 MHz)

Post by Hank Michalenka »

what would you be using for a feeder line? I didn't see it mentioned anywhere...
13dka
Posts: 134
Joined: Sat May 28, 2022 4:27 am
Location: On or near a dike

Re: The small SWL beam (1...30 MHz)

Post by 13dka »

Well if this is getting an LNA and used outsides only, any old 50Ohm coax would do, cheap 75 Ohm sat cable likely too. For passive lossy antennas like the LoG I tend to want really low loss coax, and if that coax will run through the noise halo of a house (or any noise sources for that matter) I'd want the shielding to be as dense as possible, that's why I spend the money for H-155 to be on the safe side (5mm cable is quite convenient too, and the cheap clip-on ferrites off Amazon fit on that).
User avatar
Andrew (grayhat)
Posts: 280
Joined: Sat May 28, 2022 5:56 am
Location: JN63pn

Re: The small SWL beam (1...30 MHz)

Post by Andrew (grayhat) »

hi, first of all let me say that I wasn't expecting a full test, just some initial impressions, and then, no need to be sorry, YOU BUILT IT and took the effort of testing it, so you have all my gratitude, THANKS THANKS THANKS

that said. let me add some notes, refer to the images related to the antenna patterns and gain at the bottom of the first page of this thread

1. level (gain)

given the low gain of such a small antenna, a preamp is most probably a need, and I guess it won't cause SNR issues, also, if you look at the gain graph, going below 1MHz the gain drops quite dramatically, that should explain why LW performance, w/o a presmp, was so poor

2 and 3. directionality and takeoff

raising the antenna above 3m lowers the lobe at lower frequencies at the expense of the lobe on higher ones (at/above 14 MHz), also, if you look at the pattern graphs, you'll see that, while the lower angle is... pretty low, the higher one, at least at lower frequencies is between 5 and 10 degrees from vertical, this may explain the differences in directivity at different times, my guess is that some NVIS signals came through even if the antenna was aimed at a different heading, yet I think that the low angle is confirmed by some of the DX stations you received and by the ability to null out signals

now some additional notes, first of all a test with an LNA should reveal more infos, second, it may be possible that, depending from the nature of the terrain, the optimal height is different from 3m which was simulated on average ground, then, I wonder if changing the chokes placement may help a bit but that's just s wild guess
User avatar
Andrew (grayhat)
Posts: 280
Joined: Sat May 28, 2022 5:56 am
Location: JN63pn

Re: The small SWL beam (1...30 MHz)

Post by Andrew (grayhat) »

13dka wrote: Mon Jul 18, 2022 4:43 pm Well if this is getting an LNA and used outsides only, any old 50Ohm coax would do, cheap 75 Ohm sat cable likely too. For passive lossy antennas like the LoG I tend to want really low loss coax, and if that coax will run through the noise halo of a house (or any noise sources for that matter) I'd want the shielding to be as dense as possible, that's why I spend the money for H-155 to be on the safe side (5mm cable is quite convenient too, and the cheap clip-on ferrites off Amazon fit on that).
right, some cheap TV/sat 75Ohm coax should fit w/o problems then, by the way, it also depends from environment :) and then I think it definitely needs a preamp
13dka
Posts: 134
Joined: Sat May 28, 2022 4:27 am
Location: On or near a dike

Re: The small SWL beam (1...30 MHz)

Post by 13dka »

Andrew (grayhat) wrote: Mon Jul 18, 2022 8:06 pm hi, first of all let me say that I wasn't expecting a full test, just some initial impressions, and then, no need to be sorry, YOU BUILT IT and took the effort of testing it, so you have all my gratitude, THANKS THANKS THANKS
I have to thank you for sharing an awesome idea so I could build something that looks like it will be at least a killer concept for SWLs and hams with very restricted space for antennas, and potentially something to massively prefer over any big antenna that can't be moved, and/or mounted high enough. Nothing beats a LoG in portability but the SDL folded together plus a short fiberglass pole can still be brought in (or tied to) a rucksack.
Andrew (grayhat) wrote: Mon Jul 18, 2022 8:06 pm raising the antenna above 3m lowers the lobe at lower frequencies at the expense of the lobe on higher ones (at/above 14 MHz), also, if you look at the pattern graphs, you'll see that, while the lower angle is... pretty low, the higher one, at least at lower frequencies is between 5 and 10 degrees from vertical, this may explain the differences in directivity at different times, my guess is that some NVIS signals came through even if the antenna was aimed at a different heading, yet I think that the low angle is confirmed by some of the DX stations you received and by the ability to null out signals
[...] it may be possible that, depending from the nature of the terrain, the optimal height is different from 3m which
I ran all bands in 4NEC2 with my estimated "highly moist" ground and "seawater" and confirmed 3m is still the optimum height. Well I probably had 3.5m due to the length of the pole segments but testing changes in height was not on the schedule yet. :) However, signals coming in at a high angle would be another quite possible explanation for the evening/morning difference. At the end, the ionosphere decides about the angles and can deviate from simple geometry and outdated earth->sky->earth bouncing models.
Andrew (grayhat) wrote: Mon Jul 18, 2022 8:06 pm a preamp [...] I guess it won't cause SNR issues
Au contraire, mon ami! :) I've repeatedly experienced LNAs improving SNR. First off, modern quality LNAs have very little noise and even if they had not, the added few dBs of noise wouldn't be noticeable in even slightly noisy environments. On the other hand, in a low noise enivronment with a very lossy antenna the signal may be close to the receiver noise floor. To try a very coarse calculation, let's say the signal is 3 dB over the RX noise floor = SNR 3dB. Now add 20 dB of gain (assuming it's night and the sum of atmospheric/galactic and thermal noise is far lower than the receiver noise) and subtract 2 dB of noise coming from the LNA, that's still an SNR of 18dB now (disclaimer: numbers may be totally off here). I've seen this particularly with short whips and the super lossy YouLoop on the dike, but even indoors for some strange reason, on a frequency as high as 17MHz (I have this on video). Also, the placement of the LNA directly at the feedpoint likely beats increasing the level 100ft down a feedline.

Edit: That "calculation" is extremely simplified of course, and really only fits very lossy antennas. I'm not really an electronics buff and most certainly not good at math. :) Furthermore, even if this turns out to be not the one antenna answering all SWL prayers, I'm already sure it could be the MW DXer's upcoming small weapon of choice, due to overall gain/sensitivity vs LNA noise being much less of an issue down there than the lack of directionality, or bidirectional antennas not really cutting it when you sit right between the wanted and the unwanted signals. Instead you get an antenna in roughly the same size and shape as SMLs but with the ability to hit the loudest unwanted station on the head with the null and attenuating all of their neighbors as well, without the need for crossloop contraptions etc. That alone would be pretty big news at any rate!
User avatar
Andrew (grayhat)
Posts: 280
Joined: Sat May 28, 2022 5:56 am
Location: JN63pn

Re: The small SWL beam (1...30 MHz)

Post by Andrew (grayhat) »

as a note, while running the NEC simulations I found (it was expected) that by lowering the resistor to about 200 Ohm or so, the cardioid pattern disappears, while, raising it to 1K or more, the pattern is more pronounced and the lobe becomes lower and narrower, this at the expense of gain (rising the resistor gain lowers) and matching (impedance at feedpoint changes); the choice of the 530 Ohm value was a compromise between gain, pattern and feedline matching
Post Reply