Why do I keep getting bad JT9 signal reports?

 

I have been active on JT9 mode recently. I think this mode has great potential for becoming one of the leading HF digital modes, even for DX-pedition purposes (some tweaks needed though).

I am running digital modes exclusively with my great little Flexradio Flex 1500 and I have a small QRP linear amplifier boosting the signal to 20-30 watts.

Based on reports on-the-air the typical TX power in JT9 mode seems to be between 5 and 30 watts. So I am perhaps running it in the upper end of the scale. Having said that, my normal report is 10-15dB lower than what I send back. This seems to be quite consistent on various bands. So what the heck is happening here?

JT9 has dB in S/N

JT9 is great in tough band conditions and when the conditions are good you can run qrpp and still get great results. The dB in WSJT-X user interface is Signal to Noise ratio (S/N), not the signal strength of the signal you are receiving. Of course, a strong signal usually also means better readability, thus larger S/N dB figure in the user interface.

Instead of me trying to explain the S/N please refer to Wikipedia:

S/N ratio explained

Quad has less noise than Yagi!

I wrote last year in this blog about my Quad project. I have now more experience from this antenna alongside with my yagi. The Quad antenna system has definite edge over the yagi in signal readability! Less static coming from antenna equals better S/N.

I still have the 450 ohm ladder line routed from my ham shack to the tower. The project to move the tuner to the tower is still work in progress and my target is this spring.

Tuner as pre-selector

Due to antenna construction the radiator has to be tuned. My DIY tuner is based on great hiQ S-match design by PA0FRI. This tuner has quite narrow band-pass filter functionality built in.

S-match information

My S-match prototype in the picture (silver coil roller inductor, vacuum capacitor, Amidon toroids, red wire is silicone insulated).

The tuner acts as a narrow pre-selector in similar fashion than how magnetic loop antennas work.

This type of antenna matching effectively eliminates noise going to the rig.

Fully digital signal processing chain

Have you ever heard complaints about audio ground loops or RF in audio when you are transmitting? I have suffered perhaps from every issue possible regarding RF and audio problems. Not any more! When I switched to Flex 1500 all those nasty RF and receive noise issues were simply gone. Once the Flexradio has transformed the received signal to I and Q and digitized those the received audio never leaves the computer. The processing from Flex to WSJT-X is done fully digital eliminating all those nasty RF issues. I highly recommend this type of fully digital setup for anyone seriously involved with digital modes.

Location, location, location!

Last but not the least. Perhaps the most overlooked issue is the location of the QTH. As I have written in my QRZ.com profile. “My QTH is one of the highest points in near vicinity and my antennas are 120m above sea level. Nice clear takeoff angle to all the directions with reasonable noise level.”

One of the exceptions regarding the reports was G7APP. I sent –01 and received +01. I checked the location of G7APP from QRZ.com and google maps. No wonder I got such a great report! Great countryside location!

Final thoughts

I have stopped worrying about the JT9 reports. The report is the indication of one’s working conditions, not necessarily an indication how strong the transmitted signal has been. The software on both ends is the same, the S/N dB calculation is done with exactly same algorithm.

If you constantly receive better reports than what you send, please consider what I wrote above.

Hope to work you with JT9!

Cheers, Toni oh3t

Fun With Flex 1500 and JT-65

CQ CQ

In the passband you can see JT65-HF traffic and a little bit below is very active PSK traffic.

If you have never tried JT65-HF I definitely recommend trying. It is fun mode to work and very effective during tought band conditions. And even better, you never know who answers when you call CQ. With just 5 watts you can work the whole world!

And finally from the pskreporter we can check the propagation. And here we can see that my JT65 transmission is heard in VK and YB land as well.

My quad stack was beaming to NA.

Break Is Over

Finally some activity also in this blog.

Antennas are still the same

No improvements to antenna setup during summer :(

I tore a ligament in my right hand during antenna construction just before summer while building new radiator elements for 12 and 10 meter bands. I guess it is needless to say that my summer antenna construction season was basically over before it even started.

Maybe next year with better luck.

Natural High

Natural High for Antennas

What is the best way to launch antennas to trees?

I have been using a fishing rod and a golf ball for quite some time to sling the antenna wires over trees. That setup has been working reasonably well, but it has been lacking accuracy and a bit more height is always preferred.

Here is the my antenna launcher. I found reasonably cheap used long bow. The bow is 44 pounds strong. I drilled a hole to a golf ball and glued the arrow with strong epoxy. Then I shortened an old baitcasting rod and used hose clamps to connect it to the bow.

Foxhunting With a Twist

Over the past several weeks I have been participating to QRP Foxhunting on Sundays and Mondays. The idea is to find low power QRP stations and “shoot them” by making a QSO. This is a friendly event among QRP enthusiasts in EU. The master of ceremonies is Jos, ON6WJ and he coordinates the EUQRPFOX mailing list and the website. Every now and then foxes and hunters are given special themes such as straight keys, vintage rigs, and most recently stealth antennas.

If you are interested about this activity, please join us by participating the weekly hunts and by subscribe to euqrpfox reflector by following the link below.
Join reflector (mailinglist).
You can find more information from EUQRPFOX website, link below.
EU QRP Foxhunt website
The following is the description of my hunt during the past weekend.

Foxhunting with 92cm magnetic loop

Mission: 60 minutes, foxes on the loose, band 20m, QRP 5 watts, stealth antenna.

My home QTH is full of antennas and squeezing yet another antenna seems to be a very difficult task. To avoid antenna coupling I decided to go mobile and run the KX3 from dedicated battery. I charged the battery overnight so it was ready to go on Sunday morning. I did some map searching about possible high places in near vicinity of my home QTH, but could not figure anything that would have the access route on the right side of the hill and have a short drive by car. And I definitely did not want the hill between myself and all the foxes. Instead for opting a high place I decided to drive to a nearby lake. There is a parking lot next to the lake and cross country skiing tracks leaving to all directions so the place should provide good takeoff to EU. The parking lot is on the north side of the lake, so that should favor the southbound and south west antenna directions even further. Google maps. Lake Ruotasjärvi. View towards south. I checked everything in my home QTH before finally packing the antenna to the van. When I got to the lake I quickly took the antenna from the van, connected the coax to the loop, and hoisted it up to 4 meters. I tied the rod to the mirror of the Vito for keeping it upright. The capacitor provides reasonably accurate and very fast tuning with the tuning rod. Within few seconds I was able to put loop into SWR 1.0, measured with KX3 and 1 watt. Then it was time to go back inside the car and put the earphones on and find the foxes. My first reaction was “Oh sh*t!” There was massive contest splatter all over the bands. But hey, foxhunting is not for fainthearted! I started scanning the bands and by luck I found Zoli (HA2PP) hiding a bit above 14,056. Unfortunately he was in and out from the QSB plus terribly in QRM from contest stations. So, I made a mark for this frequency and continued scanning. What made the scanning process especially cumbersome was the very narrow bandwidth of the loop. The practical < 2 SWR was less than 7kHz so I had to retune the antenna every now and then while going up and down the 20m band. Unfortunately, I could not find anyone else so I decided to wait and see if band conditions would improve and I could work the fox station. And then just like a miracle a strong signal came calling CQ EUQRPFOX just a bit below Zoli. My antenna was already tuned for the frequency so I had a little room to maneuver. With help of KX3 I quickly zero-beated to the fox who was calling CQ. I gave my call twice, and yes, Andy (SP9NLI) came back. I think he was as surprised as I was. He was real S5, and coming loud and clear. We quickly exchanged the reports and antenna information. That QSO really made my day! The remaining time I continued waiting for Zoli. There was strong S9++ contest station exactly on the same QRG, I tried to get his attention in between the CQ robot calling test. However, despite my official fox hunting gloves, I could not give a successful shot at this fox. After the foxhunt, I went to 17m, retuned the antenna and called CQ for few minutes. A couple of stations came back. At this time I was getting a bit cold already despite the beautiful sunshine coming inside the van so I decided to call it a day. Finally I snapped some pictures to my archives. Beautiful weather and very nice field trip on Sunday!

Best 72s and Happy Hunting!

Quad Twins

After a lot of work and struggle, with the kind help from local HAM community I managed to raise the quad #1 to its final height. Once the #1 was up I started the construction of the #2 and I finalized the construction on Tuesday. I snapped a picture of the tower on Tuesday afternoon. The weather was grey and temperature was –7C.

Critical Measures

Stacked two identical HRS Quads. Original idea from DK7ZB. Thanks!
5 bands: 10-20.
Feed system, 450 ohm ladder line from shack. Tuner S-match by PA0FRI.
Tower rotating 360 degrees.
Upper quad center height 20m, lower quad center height 9m, stacking distance 11m.

Eznec Model Figures

Eznec calculated performance figures for the stack.
20m
 
17m
 
15m
 
12m
 
10m
 
Reasonably good for DX, I think… but always room for some improvement.

Quad vs. Yagi, part 3

 

On the final day of XT2TT worked them QRP while testing the Quad. Also 9M6 was very loud S9+10 while my quad was pointing to east. Pileup was over 1k wide. I gave my call twice “oh3t/qrp oh3t” and he came back immediately with oh3 oh3 5nn. Very rewarding to test the antenna with good results. I have to admit that I used my old yagi for working Clipperton team TX5K, since I am still a bit hesitant to put kw through my tuner-work-in-progress.

Antenna Bandwidth

Antenna bandwidth is critical aspect of design. I am suffering from snow and icy conditions at least 4 months each year. Resonant antennas can not be designed to be very narrow, since the snow and ice will alter the resonance point and swr significantly.

The yagi design that I have been using as the reference for comparison with the quad can be seen below. I used eznec to calculate the swr and optimized the feed impedance to correspond whet it would look like with the matching stub or for example gamma feed (pick your favorite).

As can be seen from the plot the feed impedance for this 3 el yagi is reasonably low. I used 15 ohm as the feed impedance for drawing this plot.

The reasonable feed swr can be considered to be below 2 for the antenna. The low point for swr < 2 is 14.14 MHz (the upper figure) and the high point for < 2 swr is 14.36MHz. So the < 2 swr range for this antenna is 200kHz. This is fine for example 40m band in EU and plenty for 17m and 24m WARCs but for 20m this is unacceptable.

The point here is not to start optimizing the 3 el yagi but to demonstrate that narrow yagi designs that are optimized for gain and f/b may lead into trouble. Personally I am not a big fan of traditional yagi designs, but prefer the modern OWA designs because of the winter conditions.

Quad and Bandwidth

Traditional quad antennas are equally narrow. One of the main reasons for opting to go with the HRSQ design is the tough winter conditions. The HRSQ is tuner fed, and antenna can be tuned regardless of the snow, ice, and rain. However, the pattern will change as I have already demonstrated in previous analysis.

But lets go extreme. If the resonance point is shifted down and tuner is used to put the antenna into resonance and swr 1.0 the pattern will change.

Plot with 13.5MHz will revert the pattern, the gain is few dB below the nominal, f/b is destroyed, but I can still keep on working and make contacts with full power.

This plot is drawn with 14.5MHz. The pattern has similar characteristics to feeding way below the nominal frequency. The gain is again below the design, but contacts can be made with full power.

Clearly this design due to its nature of feed with tuner and force feeding the antenna has its benefits in tough winter conditions.

HRSQ is clear winner in bandwidth.

Turning Radius

One of the critical design criteria for my tower was to be able to stack the antennas. Stacking has the impact of increased gain and better control over the launch angle. Stacking has also the impact of really pushing down the launch angle below 5 degrees with good gain.

In order to stack antennas one has to avoid the guy wires. High gain 15 and 20m yagis require long booms and in case of my tower setup I do not have the real estate to spread out the guy wires to accommodate long boom yagi stack. Further, I studied the different monoband and multiband yagi options and always run into problem with the guy wires. So I had to optimize the turning radius.

The turning radius for these antennas were calculated for the modeled yagi, HRSQ in diamond and “standard” quad configurations.

Antenna	Turning radius, meters
Yagi	4,22
HRSQ Diamond	4,27
HRSQ square “standard”	2,5

So the monoband 20m yagi has a similar turning radius than the HRSQ in diamond configuration, but 1,7m longer than HRSQ in square configuration. So we have a clear winner.

Summary and Final Words

These three articles were covering comparison between quad and 3 el yagi.

Based on the simulation results I would conclude that very few quad myths actually hold true. Quad is not equal to 3 element yagi. Quad is no more fault tolerant to construction errors than yagi. Yagi is clearly better in all performance related metrics compared to quad.

Yagi is the very clear winner in this comparison!

Quad wins in turning radius, and especially HRSQ wins in antenna bandwidth, but that comes with the extra burden of high power tuner in the feed system.

So, how reliable is this comparison? Based on my experience from the models which I have done and the results from field tests and on-air performance comparisons I was referring earlier, the results are very close to what you will get from these antennas.

Of course, quad and loop antenna enthusiasts may say that it is not fair to compare monoband yagi with multiband quad and declare the yagi as clear winner.

Yes I agree, it is not fair. Monoband yagi is superior antenna, but the quad design is very easy to multiband. The point was to demonstrate that quad is very close to yagi, but not superior. The benefits of quad design comes in the next acouple of sentences. In fact all the results and plots in the articles were from monoband yagi and multiband quad. But the results for quad are what they are, even in the multiband setup. So basically, I would say it is reasonably fair to build a quad that is close to 3 el monoband yagi, and get 4 bands for “free” without compromise in performance! It is very fair!

Thanks for reading!