The following antenna measurements with reference to a dipole at the same height were done in August 2003:

Initially I only wanted to do a comparison between two spiderbeams, one of them using a different feed method suggested by a user.
These measurements were done first, finding out that there is no difference between the two feed systems.

Then, because I had never been on such an ideal test terrain before (see description below), I decided to stay longer and also run some more tests of the spiderbeam against a reference dipole, installed rotatable at the same height (10m).

Description of the test setup

As can be seen from the picture, I found a very nice field which is open to all directions, only a few trees here and there but they dont matter at all. The terrain is incredibly, totally flat in all directions for at least 30 kilometers (typical north German terrain characteristic), making it very well suitable for such test measurements.

For the first test,
I set up 2 identical spider beams, each on a 10m aluminum pole as you can see from the picture. The antennas were spaced 60m apart from each other, I had the station in the middle (where the car is) and could switch between the 2 antennas instantly. The feed lines to each antenna were identical, the coax cable was brand new.

For the second test,
I took down all the directors and reflectors of one spiderbeam and kept only the driven element (multiband-dipole).

Test results
I made lots of contacts with stations in DL, EU and DX, never telling the QSO partners which antenna is which one, only switching between "antenna 1" and "antenna 2" several times. Both antennas were always rotated to radiate exactly in the direction of the QSO partner.

The results on 10m and 15m were:
10% of the people said - spider beam is better by 3 S-units
60% of the people said - spider beam is better by 2 S-units
20% of the people said - spider beam is better by 1 S-unit
10% of the people said - no difference

The results on 20m were:
30% of the people said - spider beam is better by 2 S-units
50% of the people said - spider beam is better by 1 S-unit
20% of the people said - no difference

What do I conclude from this test?

The differences between the results on 15/10m and 20m are mainly because of the installation height.
The higher the antenna is installed (in terms of wavelength) the more visible becomes the gain of the beam vs. the dipole.
This is a well know fact and can also be viewed easily in simulation.

The gain difference between the spiderbeam and a dipole is at least one S-unit, most of the times 2 S-units.
Theoretically, one S-unit is 6dB. Most S-Meters are not very accurate, but 90% of the people always saw a signal increase with the spider beam, no one saw it the other way round.
The theoretical gain of a 3-element yagi like the spiderbeam over a dipole is around 6dB.
So, the gain measurements made on that weekend match very well with the calculated values.

So all in all I conclude (once again...) that the antenna works exactly as specified... :-)
There is no magic involved either, it just shows the typical behaviour of a 3ele yagi vs. a dipole, installed rotatable at the same height...

(BTW it also shows again that a rotatable dipole installed at reasonable height is not such a bad antenna...) :-)

----------( end of test report )----------

additional user reports

OK, for those of you asking for more tests or measurements...
Below please find a few excerpts from mails that arrived here during the last weeks,
i.e. user reports from guys who received their spiderbeam kit a while ago and have the antenna assembled and up in the air by now...

-------------------------- user reports following: --------------------------

--- 7Q7BP (G3MRC) ---
...Just to let you know that on Tuesday afternoon we finally managed to get the Spider Beam up on top of the 45 foot tower, and it seems to be working really, really well, so I am very pleased. [...]
I set the resonance with an antenna analyser when it was just a couple of metres from the ground to a point very low down in the CW section of each band, and then checked again when it was at 45 feet, and as expected, the resonant frequency had moved between 100 and 150 Khz higher, which is fine for me as I mostly only do CW and Data modes. [...]

--- VK9XW / VK9CD (DM5TI, DL2RMC) ---
... have assembled a spiderbeam for our planned DXpedition to VK9X/C (coming up in October 2003), put it up on an 8m telescopic mast, carefully measured and tested its performance, and also used the antenna in several contests.
Our comparison antenna was a 2 ele triband trap beam, put up on a 12m mast. The results in favor of the spiderbeam exceeded our expectations by far. Stations barely readable on the trap beam were easy to read (and work!) on the spiderbeam. We are very content with our tests so far [...] and view this antenna as an optimum solution for DXpeditions, regarding weight and transportation length.

--- UA9CLB / UA9CDC ---
...pleased to be able to report good news. [...] were testing your antenna during WPX CW in Turkey when doing Multi-OP (YM3ZF) together with Sergey, TA2ZF. Came back home very unhappy with everything but antenna. Managed to compare spider with the commercial 5 el tribander that Sergey has and spider outperformed that log-periodic commercial by at least 2 s units. [...]
...just worked in WAE SOAB LP from UA9 with spider beam and added 2 el 40 meters. I seem to have managed to set new LP record from here without going to 5B4... :-) [...] Great to be able to operate from the country site (very quiet), heard most everyone who called. (see 3830 reflector for details).

--- DM3ML ---
... just found out that I worked CT9D among the 19 QSOs during the last hour of the IOTA contest on 20m. I worked you with my new spiderbeam. I only heard you after finding the spot in the cluster and turning the antenna in your direction, then got through with the first call! Thats the way a DXer likes his antenna. In one sentence: the spiderbeam really works good for me and I am positively surprised! [...]
The assembly of the antenna took place during several summer afternoons. Everything fit together perfectly and enough Kevlar rope and DX-wire was supplied. Working with the epoxy resin was easier than suspected. I put the beam on the telescopic tower at my weekend garden shack, which I can extend to 9m height. All the SWR values were on the spot immediately. On 20m and 15m the minimum values occured at 14250 and 21300, on 10m at 28050. Next time I will shift these frequencies a little bit. I was even more pleased with my tests regarding front-to-side and front-to-back ratio (4 S-units or 3 and more). I have used numerous directive antennas before, but have never before experienced such an outstanding directivity on all bands (incl. 20m!). The best example was a VP2M station. I could not hear him with the antenna pointing NE, he came above the noise after rotating to SW, finally arrived at a solid 56 report when the antenna pointed exactly in his direction (W), and (even more important) came back on my first call, using 300 W. The signals on the spiderbeam are generally 2 S-units louder than on my R7 vertical and the noise is much lower.
I have quite a high success rate when calling for rare ones [...] With ZK1JD coming back on first call and FO5RA answering my CQs, the antenna really must be working, congrats & TNX!