Three times this week I've called CQ on 7.030+/-, and every time someone
has come back to me and patiently accepted my ropey morse code.

While I love walking up mountains, you guys who keep the radio on in the
shack and work SOTA activators deserve a shout.

Thank you.

Why bother using a choc-block on the end of a piece of RG-58 when you can have something this overengineered:

Lets say you want to quickly figure out how many dB there are between 60w and 100w. One way is to approximate this as difference from 50w to 100w, and everyone knows that 2x is 3db. Normally I'd use a calculator if I wanted more precision, but here is a technique that gives about 2 significant figures using a simple table that can be memorised:

2  3.0
 3  4.8
 5  7.0
 7  8.5
11  10.4
13  11.1
17  12.3

This evaluates 10.log10() of the first few prime numbers. The technique is to find the prime factorisation of the number in question (or a number close to it), look up the log values for the prime factors and add them together.

We are making use of the fact that log(AB) = log(A) + log(B)

Let's consider the question of the ratio of 60 to 100.

100 is 20dB. We know that.

60 is 10 * 3 * 2. We know that 10x is 10dB, and from the table 2x is 3.0dB and 3x is 4.8dB.

x60 is thus 10 + 3.0 + 4.8 = 17.8dB

The difference between 100 and 60 is thus 20 - 17.8 = 2.2dB, which is pretty close to the actual value of 2.2184.

For large prime numbers such as 71, pick a number that is close and has a sane factorisation such as 70 = 10 * 7. We know that x10 is 10dB, and the table tells us that x7 is 8.5 db. 10 + 8.5 = 18.5dB, (the actual value is 18.5125).

Easy. Now just learn the table.

I just spent a few enjoyable evenings building a noise bridge. After my fawning over Small Wonder Labs kits, you'll be glad to know that I built this one from scratch Manhattan style.

I didn't notice it at the time, but guess who was the designer of the New England QRP noise bridge? Yep Dave, K1SWL again.

It has been useful already: it seems that the integrated feeder & balun that I use for SOTA trips is broken at the connector. 

I built the bridge without a controllable 'other leg': it will only measure return loss relative to 50 Ohms. The bridge seems to work up to 50MHz: I get S8 noise with the DUT open circuit and S0 with a ~52Ohm terminator.

The noise power is much higher at lower frequencies: it gives about S9+40 on top band, dropping to S9 on 40m.

Did I say that I have S9 noise at my QTH? That unfortunately means that I can't use it to measure where my 40m vertical is resonant because I can't hear the noise for the noise. Doh.

Anyway, at CUWS we've previously had problems using a miniVNA near other transmitters, and hopefully this will be immune to that problem.

I just finished a mobile app for UK band plan:

http://phil-wise.com/bandplan

It makes use of HTML5 application caching, so if you bookmark the page to the home screen on an IOS or android device, it will work even when you are offline.

Please try it out, any feedback is welcome.

I just finshed putting my Small Wonder Labs SW+40 in a case. I've built two kits from Dave Benson now, and both have been excellent. The stock SW+40 doesn't come with a keyer (unlike the Rockmite!) so I needed to put one together. I could have bought one of the many pre-programmed PIC keyers that are floating around, but I'm an open source kind of guy who doesn't like buying things I can't take apart and understand. Instead I spent many pleasurable hours writing one that is very easy to port to new devices, which I'll open source sometime.

One of the use cases for this rig is going to be SOTA activations, and I didn't want a separate battery kicking around, nor a power switch that could be easy knocked on. Since I was writing embedded software anyway, I made the rig auto power on when the paddle is pressed and power off after 20 minutes of no activity. It draws about 100nA on power off, so the embedded 800mAh LiPo should last a long time.    

There was a contest on today, so I got the chance to compare it to my FT-857's receiver. Frankly the SW+40 is nicer. It has a sharper filter and doesn't suffer from AGC pumping like the '857. The transmitter is ~20db but is does come with diode QSK :)

I'm I getting any better at morse? Let's use my RufzXP score as a proxy and plot that over time (the x axis is days):

I'm getting better slowly, aparently. The trend line says I manage about 8 more points per day, which is pretty crappy. I need to find a better way to get train.

Back in the dim and distant past I bought a Rockmite 20m kit radio from Small Wonder Labs. It is a crystal controlled, direct convertion rig with a built in keyer.

It only runs 500mw, and I sat and called CQ with it in the shack for a bit today. Just as I was about to give up I heard a SQ calling, he was loud and on one of the two frequencies that the rockmite can use. I called him back.

M 0 D E G, oh yes! He can hear me! We had a quick exchange of signal reports and names, which was confused a little at my end because the direct convertion architecture of the rockmite means that there is an image response approximately 1.4khz (twice the sidetone frequency) from the transmit frequency. I heard a loud station while I was transmitting (rockmite has diode QSK) and stopping to figure out what was happening wasn't the right move. It all worked out in the end, and the rockmite gets another one in the log.

There is something magical about making contacts on the rockmite that I don't get on the FT-1000 + Quadra that sits next to it. Part of it is that a kw amp, sharp crystal filters and no broadcast band break band break in makes contacts easy, but I think there is a more subtle thing at work: I understand the operation of the rockmite from top to bottom.

I've got S9 noise problems, and I wanted to build a shielded magnetic loop to try and reduce the near field noise sources. Rather than just build the thing and convince myself that it was better, I want to do something approaching a real experiment.

The hypothesis is that the loop will out perform the vertical. I built a quick shielded magnetic loop using 3m of RG58 and a FT37-42 balun. Because the impedence of a small loop is low, I made it with a 1:4 impedence ratio using 10 trifiliar windings and connecting two of them in series.

This was mounted with the center about 2m off the ground. I could hear signals on both the Vert and Loop. All good.

I then ran WSPR and switched between the two antennas every 16 minutes by hand. A quick python script on the ALL_WSPR.txt file gets the data into a form I can use. The next question is how to analyse the data, and it is important to do this because actually looking at the data.

The scheme I picked was to find all the call signs that were copied on both antennas, and plot a scatter graph of all the SNR for each pair of reception reports. If a call sign was heard twice on one antenna and three times on the other, there would be 6 points plotted.

This should result in a nice win/loose graph with the diagonal representing a dead heat.

I tried to fit a line though the resulting graph, and the good news is that the slope is about 1.0, which you would expect. The y-intercept is the critical value, and should give the relative performance number we want...

+0.4db, in favour of the vertical.

Well that is a surprise on two counts. I thought that the vertical would be worse, and this means that I don't have to carry on with plans to build a separate recieve antenna and antenna switching system. It is also very small, hinting that something that we are not measuring is causing the noise problems.

Not On The Air - On The Air (NOTAOTA) aims to encourage activations of things that haven't previously been considering worth activation, and might be just the thing to get involved with while you are waiting for those last IOTAs to be activated.

NOTAOTA combines exclusivity with a deep source of 'new ones'. Once something has been activated once, it can't be activated again, because it wasn't 'Not on the air', so once you've got a NOTAOTA in the log you can be sure there wont be a massive follow up expedition. Once a NOTAOTA is done, it is done.

On the otherhand, NOTAOTA is ideal for hams who can't afford a ship+helicopter: if you can concive it, you are done.

'How can I get started?'

Simple. Follow these rules for activation:

1) Pick a NOTAOTA identifier. It will be of the form <something>-001
2) Make 10 QSOs. You MUST mention NOTAOTA and the identifier in each QSO.
3) Post a comment here to stop someone else accidentally activating the same category again.
4) Done!