Guest Post: Extreme QRP–Testing the AX1 with WSPR and 20mW

Many thanks to Keith (KY4KK) who shares the following report:

Extreme QRP – Testing the AX1 with WSPR and 20mW

by Keith (KY4KK)

Thomas, thanks for all of your activation videos related to the Elecraft AX1 antenna.  I ordered one the day you announced the package deal, and it arrived in less than a week.  I’ve activated a few parks with it already (20m SSB).  Like you and many others, I’m impressed.

I was very interested in Thomas Barris’ (DM1TBE) March 12 QRPer post using WSPR to test his POTA antennas in Germany.  Then I saw Bob’s (K7ZB) post about his ZachTek Flea with 300 milliwatts in CW mode.  To me, extreme QRP represents some of the magic of HAM radio.  I’d like to share one of my most recent WSPR experiments related to the AX1.

About a year ago, a friend (NG4S) loaned me his pair of WSPR transmitters and suggested that I explore building and comparing antennas. I’ve been hooked on antennas of all kinds and WSPR since then.

I began doing WSPR tests on the AX1 the day after it arrived.  With two transmitters set to the same frequency and power output, you can do direct comparisons between two antennas under identical propagation conditions.

I’ve already done a couple of comparisons between the AX1 and other commercial antennas.  But I think the test I just completed might be of particular interest because it pits the AX1 against an antenna I’ve seen you use many times – a 28.5’ end fed with a 28.5’ counterpoise.  I used 24 AWG silicone insulated wire. The end of the radiator was placed on a 19’5” telescoping fishing pole.  This is my preferred POTA mast when I can’t use a tall tree.

I spent some time trying to control other variables so that the only significant difference during the test would be the antennas themselves.

For example, the SOTA Beam WSPRLite Classic transmitters don’t have an ATU.  So, I had to make the antennas resonant on the 20-meter WSPR frequency of 14.097 MHz.  For the AX1, Thomas’ videos helped a lot.  I used a clip-on capacitance hat and adjusted the counterpoise to 15’ 2”.  This gave me an SWR of 1.17:1.  For the end fed, I tried the two UNUN’s I had available and settled on the 49:1, which got me the closest (2.2:1).  I then used a manual tuner to achieve an SWR of 1.29:1.

I also wanted to deal with the difference in power output between the two transmitters.  Although they’re identical, and both set to 20 milliwatts, there is no way to ensure both are actually producing that output level.  Based on tests by NG4S, one of the transmitters runs at 19 milliwatts.  The other actually outputs 27 milliwatts.  So, my plan was to run the test for 48 hours. At the end of 24 hours, I would switch the transmitters (and callsigns) so that both antennas would benefit (relatively equally) from one of the transmitters being stronger.

At the end of Day 1, I reviewed the data from the two transmitters on  The end fed averaged a 5.7 dB gain over the AX1 based on reports from receiving stations that spotted both transmitters in the same 10-minute block (simultaneous spots).

On Day 1, the stronger transmitter was on the end fed.  The maps below are from

AX1 – Day 1

End Fed – Day 1

I was pretty impressed that the AX1 got into Europe and Africa on only .019 Watt!  I always have good luck with end feds, so was not too surprised to see this one perform well.

At the start of Day 2, I switched the transmitters and changed the callsigns so that the stronger beacon would now be assigned to the AX1.

At the end of Day 2, the simultaneous spots data showed that the end fed still had a 3.4 dB gain over the AX1. This excludes all spots from the first 24 hours.

AX1 – Day 2

End Fed – Day 2

These maps represent only a portion of the data collected and compiled by the WSPR network.  For example, the graphs below show an hourly comparison of the average distance of the receiving stations for each transmitter for each day [click images to enlarge].

The red line is the AX1 and the end fed is blue.  Note that on Day 2, the AX1 was using the stronger transmitter.  I think it’s interesting to see how each antenna and transmitter combination handled the period over night when 20-meters is less reliable on the east coast.

I’m not nearly qualified to draw any scientific conclusions from this experiment, but I will offer a few humble observations:

  • The 40% difference in output power between the two transmitters didn’t seem to result in a similarly significant variation in the performance of either antenna.  Day 1 and Day 2 contact coverage was almost identical.  While the graphs and maps tell a general story, the proof exists in the raw data.  A deep dive showed that the variation in the received signal strength between the two transmitters was LESS THAN 1dB.  Interestingly, NG4S assured me that the 40% variation in output would be immaterial.  But he also knew I wouldn’t really believe it till I proved it for myself. From now on, I will run comparisons for 1 day and flip a coin to determine which antenna gets which transmitter.
  • Not all WSPR receiving stations run all the time, and this can skew results if you’re looking at averages.  I analyzed the raw data for each antenna, comparing the received signal strength from only stations that picked them up both days.  From this, I once again found only about 1dB variation in received signal strength from Day 1 to Day 2.
  • For the 2-day period, the end fed had 1487 spots compared to 616 for the AX1.  This is obviously statistically significant, but I’m not sure how to interpret it.  Again, looking at the raw data from stations that received signals from both antennas on both days, it looks like the end fed had a 3dB gain over the AX1.
  • Domestically, where the vast majority of my POTA contacts are made, the antennas had roughly the same reach.  The end fed clearly had an advantage for DX.
  • I didn’t pay much attention to the direction of the end fed slope or the direction of either counterpoise.  I just put everything in the least likely position to be tripped over for two days.  Both configurations appeared to be fairly omnidirectional.
  • I really like the AX1.  It’s a well-engineered purpose-built antenna, and I plan to use it very frequently when in populated areas.  As mentioned on QRPer, the AX1’s super power is its speed of deployment.  It was perfect for my recent activation at Santee National Wildlife Refuge (K-0521), where the ranger invited me to set up on his deck overlooking the lake.  He even let me use his rocking chair!  My POTA partners were off in the grass with the bugs.
  • I picked the 28.5’ x 28.5’ end fed for this test knowing that it was an apples to oranges comparison to the AX1. When I’m in a wildlife management area with 100’ pines and plenty of time, my long wires will continue to be my preference.  They may take longer to set up and tear down, but I get multiple bands, and that makes hunting P2P contacts a lot easier.

But Wait – One last thing:

The day after I completed this test, my friends and I activated Woods Bay State Park (K-2915).  I set up the AX1 and 28.5’ end fed in the same configuration as before.  I connected them to an antenna switch feeding into my KX2 set to 10 Watts SSB.  I switched antennas every 10 contacts for a little over an hour and stopped at 40.  Actually, I had 1 duplicate and 1 bad callsign due to my sloppy penmanship – both on the AX1.  This was in South Carolina early in the morning, so half of the country wasn’t awake yet.


End Fed

The results are effectively the same, and prove that with an antenna switch, you don’t need WSPR transmitters and careful management of variables to do meaningful field comparisons of antennas.  The WSPR data just makes it more fun for me.

I’ve been a HAM for about 18 months.  I’m fortunate to have two very experienced HAM friends who activate parks with me nearly every week.  We’re all antenna experimenters and get a lot of enjoyment in reveling in our successes and laughing about our epic failures (mostly mine).  We’re all QRPers too, and always look forward to your daily posts.  Thomas, thanks for all you do to promote the hobby.

Cheers to all,


12 thoughts on “Guest Post: Extreme QRP–Testing the AX1 with WSPR and 20mW”

  1. Thank you so much for not only performing these WSPR and POTA experiments, but for also sharing them with us, Keith! Fascinating stuff.

    It’s no small task to perform antenna comparisons–few of us have the facilities to compare antennas in a perfectly sterile and lab-grade environment–but you’ve done a brilliant job showing us the real-word performance of both antennas.

    WSPR is such an amazing tool for doing this sort of comp.

    Looking at the results, I must say that 20 mw of WSPR probably equates to a good day running normal QRP power (5 watts-ish). I can’t think of a time I’ve worked Africa with the AX1, but I’ve worked all parts of Europe, and into the western US and Canada from North Carolina.

    Again, thanks so much for sharing this and we look forward to any future posts you might decide to share!

    Cheers & 72,

    1. Thanks Thomas. Running the test was fun for me.

      I was thinking about running some tests using two AX1 antennas. The WSPR transmitters are pretty sensitive to high SWR, so I had to work hard to make the antennas resonant. But using an antenna switch on my KX2, I could compare a configuration with the supplied counterpoise against one that I tune to resonance.

      I suspect that I’m working too hard to get the SWR down, but it would be good to know for sure.


  2. Great report Keith! Its amazing such a small antenna yields such great results. I love my KX2, and own an AX-1. I tend to opt for a Random Wire, or simple Speakerwire Doublet.

    The AX-1 has proven results, evidenced by Thomas and your field results. I hope to try the AX-1 soon, in hopes of similar results. Thanks again for the report.

    1. Thanks, Rand. I did find that the supplied counterpoise for the AX1 was too short to be resonant (< 2:1) anywhere on 20m for me. I use a camping clothes line spool with about 37' of 24 AWG stranded wire rolled up. I have marks on the wire for where 20m SSB and CW become 1.2:1 or better. That was roughly 14-15.5 feet. Same for 40m but I don't remember the length. My long wires get the most activity for me too. They just work.

  3. Last Saturday, 3/25, we had a club QRP event at the Ernie Waver Youth Park in Brooksville, FL. We had 5 stations setup. I was using my IC705 and 20m End Fed Wire antenna. One of our ops, John, KK4ITX did WSPR 200mW operation using 2 antennas, one a small mag loop and a homebrew vertical based on the AX1. On initial WSRP transmissions with the vertical got 18 WSPR hits. Later got 30. Then switched to the small mag loop and got over 180 WSPR hits. All of these were in less than 5 minute transmissions. John commented he was much more impressed with the small mag loop. 73, ron, n9ee

    1. Ron,

      I built a small mag loop and ran some WSPR tests on it a few months ago. It did VERY well on 20 meters. One of the advantages we’ve found with mag loops on activations is that three of us can operate on 20 meters at the same time by nulling each other out.

      I haven’t done a head to head comparison of the AX1 and mag loop, but I think it would be interesting.


  4. Wow! This is a great test, and confirms my observations as a chaser that the AX-1 – while not necessarily a dummy load, performs A LOT worse than a wire antenna.

    Factoring in the power levels of the different transmitters, going from 19mA to 27mA is a 1.5dB gain. Then, normalizing the results for the additional power gain on :

    – Day 1: 5.7dB – 1.5dB to account for additional signal for the end fed = 4.2 dB advantage of using an end fed (normalized)
    – Day 2: 3.4dB + 1.5dB to account for additional signal for the AX-1 = 4.9 dB advantage of using an end fed (normalized)

    That’s roughly a 4.6dB advantage for the end fed – almost a full S-unit of difference, such that you may easily end up under the noise floor for a lot of stations where you would have been copyable before, as evidenced by the significant reduction of stations that were able to detect the AX-1 – the anecdotal equivalent of throwing your wire antenna on the ground and operating like that. For context, that’s nearly the difference between running a typical 5W QRP radio and a higher power rig like a 20W G90 – from my experience the results between those two are significant and noticeable both as an activator and chaser.

    The map shown in Keith’s analysis was very helpful and showed how a potentially lower angle of radiation might affect the overall activation results.

    This would all be consistent with the results I experience as a chaser (hearing stations running a wire vs. the AX-1). Given the setup time and simplicity advantage of the AX-1 over an end-fed wire, I feel portable vertical antennas have an important place in an activator’s arsenal, as long the major drawbacks are understood.

    1. Thanks Mike!
      I’m getting lots of great ideas for new tests. If I can get a wire thrown on the ground to tune up well enough, it’s in the queue for testing. Maybe I can finally reach some of those closer stations. Better yet, if I add an amplifier and combine my very poor CW skills, I might finally chase those moles out of my lawn.

      BTW – I envy your ability to do frequent SOTA activations. Kind of tough in SC.

      Best Regards,

  5. Most agree that antennas are the most important component in QRP work. It’s really encouraging to see that AX1, “dummy load” working so well. Thanks Keith for your diligent comparisons, AND for the very impressive 20mw results.

    Thanks to you too Thomas for the regular stream of interesting posts.
    Bob N4REE

    1. Hi Bob,

      I hesitated for a long time about getting the AX1. Lots of negative comments about it being an expensive dummy load. One day we had finished activating a park and realized we were near a wildlife management area. We would try a 2nd activation. I asked NG4S if I could try his AX1 on the back of his truck because I was running out of steam to get my wire back up in the air. Including set up and activation, I got my 10 contacts in about 15 minutes. That was the day Thomas mentioned the package deal from Elecraft and I ordered it as soon as I got home.


      Keith KY4KK

  6. Keith
    Thank you for sharing your wspr and pota tests. I’m another fascinated antenna experimenter. Not always using qrp but I’m really impressed with what and how you compared the ax1 to the end fed. A very inspirational article.
    Gav VK4GAV

    1. Good morning Gav,

      One of the many things I like about QRP is that when coming up with an antenna idea, you don’t have to worry too much about the consequences of high wattage (heat, arcing…).

      We activated K-9205 yesterday and my end fed completely failed me (I suspect that I damaged the UNUN during packing). I could have taken out the AX1, but W4JM had already gotten 12 contacts and offered to let me use my radio on his antenna, a vertical he made out of PVC pipe and two Slinkys. He’s used it for our last 4 activations, and frequently gets into Europe on 30, 20 & 17.

      Thanks to him, I finished my activation on 20 meters.

      Today, I’m pulling my Slinkys out of the toy box to build one of my own.


      Keith KY4KK

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.