Other SOTA sites: SOTAwatch | SOTA Home | Database | Video | Photos | Shop | Mapping | FAQs | Facebook | Contact SOTA

23cm antenna?


I think this shows that antenna design has moved on massively in the last 40 years. I’d put the improvement down to mainly the fact that we have had home computers capable of taking much of the hard work out of the maths over most of that period. Even a Spectrum 48 is capable of running massive amounts of iterative calculations to improve designs.

Now we have unbelievable compute power for pennies and lots of simulation software to play with. The fact that so much of the NEC type code has been written, released, improved that it’s no surprise that we can knock up antenna designs that outstrip the old stuff.

Should my mice need a ladder, I know where to send them. Most of them however, need major surgery to undo the damage they suffer when my “Little Nipper” lets fly!

:scream: :mouse: :scream:


I’ve just calculated the average signal strengths from the two session logs. Bearing in mind that these are contest exchanges, and at best subjective. There is a spread from S1 to S9, though. Much less of the routine “S9” often heard in larger contests-

Average reports of my signal: PCB - S6.38 Yagi - S7.33

Average reports sent by me: PCB - S6.4 Yagi - S7.0

PCB session - 34 QSOs. Yagi session - 36 QSOs

Very few marginal contacts in either session.


OK, I did the decent thing and finished building the 15 ele yagi, based on dimensions from K7MEM online calculator.

A quick field trial from the Long Mynd suggests that it has similar gain to the old J Beam 15 over 15, and a nice clean polar diagram. Better still, it weighs 280 grams with the feeder, compared to 880 grams. The feeder is about 1.5m RG223 (double screened, silver plated terminated with SMA plug) which I grabbed at the Norbrec rally. The balun is a short length of semi rigid, from the same source.

I cut 17 elements in readiness, but the piece of pipe I pulled from a skip was only long enough for 15. At just over a metre, its quite practical to carry, though.

It took me about 100 Hours of day dreaming about clever jigs etc, but once I set to, it took about 2 hours work with steel rule, scribe, hacksaw, vice and drill :o)

I just need to acquire a smaller tripod, and I’ll be attempting a few more challenging activations.


Looks good Adrian, but a bit big for me coming from a 13cms perspective :grin:

And did you do the RF absorption test on the plastic tube? At 23cms shouldn’t matter much and if it is performing as well as your J Beams - then it’s-a-working! I had problems with my 13cms 21el design in that the plastic tube (sample) I had intended using for the boom was heating up in the microwave. Abandoned this for a fibreglass tube section of one of my old SOTA poles. Did the microwave test and all fine with this.

Interestingly, later I was testing the glue I intended to use to secure the folded dipole. It appeared to heat up in the microwave. I suspected the cardboard I had used for glue sample. Tested the cardboard on it’s own. Sure enough it heated up. Tried the glue again on a piece of fibreglass. All ok!

Shortly I’ll put some pics on this Reflector as to the SOTA make-up of the beam (SOTA friendly (no need for a tripod) - same idea as with the SG Labs 4 el pcb yagi with the hole to accommodate a SOTA pole as you can see in this TOPIC). And a more detailed explanation on GM13 to follow.




Hi Jack, I did consider that, but the skip only had one piece of tube on offer, so it became an academic question! :wink:

It is marked as UPVC, though, which I’m fairly sure is not good - I remember the PVC insulation charring on the output link of my 70MHz 4CX250B linear…

However, as only the low impedance (high current low voltage) parts of the elements are close to the boom, I thought that the dielectric losses might not be too significant. I may be wrong!

Anyway, its a starting point, and the experience of building it was sufficiently enjoyable for me to contemplate a Mark II. The SWR is not perfect, and placing a finger near the end of the driven element improves it, so I think it needs to be a fraction longer. In construction, I fed the aluminium rod through the boom, and then formed the folded dipole. It is a very neat arrangement, but doesn’t allow for easy modification.

The tolerance for the reflector and directors was quoted as ± 0.7mm, and I found this quite easy to achieve. They are all within less than half of that, so with care I would consider trying a 13cm version if I get that bug!



Was that the final or the driver? :grin:

I saw a spectacular 70MHz PA once that had a pair of those big Russian triodes in it. It comfortably achieved the UK legal limit. Very comfortably!


Yesterday, I ventured back onto the north end of the Long Mynd, well outside the AZ, but with clear takeoff towards the Stoke repeater / beacon, to carry out a more objective antenna comparison.

The first surprise was the level of QSB, which made accurate comparison difficult, as it takes time to swap antennas. At the start of the test, the signal was often low, around S2, peaking to S9 for several seconds at a time. After an hour, this pattern had almost reversed, the signal being often around S9, sometimes sinking right down.
I’m guessing this is likely to be multi path phasing, with the Wrekin as a possible reflection suspect. I have a lot to learn about this band, though!

Anyway, the setup was an FT290 connected to the SG Labs transverter via a switched attenuator 0 to 15dB in 1 dB steps. I connected an analogue multimeter across the S meter, to give an easy to read level reference point.
Method - connect PCB antenna, note meter reading. Swap antenna, adjust attenuator to give the same reading. Repeat.
(I did move the antennas relative to each other, and to the car, in case there was interaction. None discernible)

I repeated the comparison a number of times over the course of an hour, noting in each case the QSB peaks. Not ideal, but by the end I was reasonably happy with the result. I may experiment with different locations, to see if I can achieve a steady signal.

So, the main comparison was between my home brew 15 ele, and the SG Labs supplied PCB antenna. I haven’t found a spec for the latter, but it is 2 elements, both fed. So, I’m assuming similar performance to the HB9CV at around 4dBd.

Result: My 15 ele appears to have 8dB gain over this, so approx 12dBd.

The theoretical gain (from the K7MEM calculator) is 13.9dBd.

A couple of dB off. Is that measurement error? Deficiency in my construction? Absorption in the PVC boom? Difference between theoretical gain and what can be achieved in practice? Feeder loss? - though 1m of RG223 should be a lot less than 1dB.

Don’t know yet, but I’m happy with it for now!

What of the old J Beam 15 over 15? Well, I didn’t have a way of mounting that vertically polarised, so I couldn’t spend the same time watching the QSB on the meter. I did hold it in my hand at the same height as the others etc for a while, and it seemed very similar to the HB Yagi.

All good fun :smile:

Test setup:

All good fun :smile:


The 290 does look like a proper radio with a display you can read without glasses and big buttons for gloved hands. Best of all, it will have only gently sipped current from your battery during the tests where an 817 would have sucked current like it was going out of fashion.

I think hitting with in 1.9dB of the theory is not bad for a 1st attempt homebrew antenna at that frequency. Especially as there could be many other sources for the difference.


The FT290 is rather heavy though; 1.3 kg compared to 0.92 kg for an FT817.


Estimating antenna gain from distant signals with QSB is a difficult proposition. Your beam seems to work well. Try it on distant activators and you will be pleased with the result, I suspect.

73 Andrew VK1DA VK2UH


Yes, that took me by surprise, I wasn’t expecting significant QSB. The path to the beacon was approx 50 miles, line of sight. There is a big steep-sided lump of rock (the Wrekin G/WB-010) a few degrees off the path, though, and I wonder if there was some incident reflection.

I’m hoping to be active from a summit again during the August UKAC (activity contest) and we’ll see how it works in practice.



Hi Adrian,

I often find QSB or even flutter on longer distances on 1296. Haven’t had enough experience of higher frequencies to know whether it’s common there. i think atmospheric conditions affect propagation greatly, even for line of sight paths and possibly especially when relatively close to the ground where convection currents/updrafts are flowing. In summer it may well be more pronounced than with more stable cool air at other times.

One of the fascinating things about the higher frequencies!

73 Andrew VK1DA VK2UH