Thursday, May 28, 2009

Spectrogram - grab it while you can

Spectrogram is a Windows-based audio spectrum analyzer that uses the computer sound card. Originally freeware, later versions had more features and required a registration fee to be paid. I first came across the program when building my Elecraft K2 back in 1999. By looking at the spectrum of band noise, you could easily set up the crystal filters so that both sides are evenly balanced and the passband where you want it to be.


The company that developed Spectrogram, Visualization Software LLC, is closing down at the end of May, and according to a notice on the download page, the website and downloads will not be available after that. The last version, Spectrogram 16, is now freeware again. So grab it while you can. You never know when you might need it.

Wednesday, May 27, 2009

KTune - remote tuning for the K3

As I mentioned in a post about a week ago, I added support for tuning the K3 using a Griffin Powermate USB knob into the next, unreleased version of KComm. I found this to be a very comfortable operating feature, allowing you to tune the radio while sitting relaxed in your chair instead of leaning forward to spin the dial. With my back trouble, that's a real help.


I thought it would be nice to have this capability when other software - or even no software at all - was loaded. I also thought it would be nice to have the tuning commands respond to global hotkey events, regardless of which application is in the foreground. The programming tool I use, Lazarus - an excellent free clone of Delphi, does not support global hotkeys so the implementation of the feature in KComm only works when the KComm main window is in the foreground.

To overcome these problems I knocked together a little standalone application using a different programming tool. It is called KTune. It sits in the Windows system tray and responds to keystrokes generated by the Griffin PowerMate by sending UP and DN tuning commands to the Elecraft radio.

You need to program the PowerMate to send the keystrokes KTune responds to. If you want to use it at the same time as another logging or datamode program then you will also need to set up a splitter using the Virtual Serial Port Emulator, which very fortunately is freeware.

It will probably be a while before I get around to making a proper web page for the program but if you have enough savvy to get it set up by yourself then you can download an installer for the program and try it.

Tuesday, May 26, 2009

eBay ads

A few ham radio sites and blogs I've looked at recently had ads in the sidebar for ham radio equipment being offered at eBay. I thought this would be rather a nice addition to my own site, but I haven't figured out how to get them. After Googling "eBay ads" I subscribed to the eBay Partner Program, which appeared to be the thing to do. However, the ads I got were not terribly interesting from a ham radio point of view.

After a bit of further investigation it seems that the eBay auction ads I have seen are coming from Google. I have had Google ads on my site for years but have never seen an ad showing ham radio items from eBay displayed there. It's a bit of a mystery.

Postscript: Mystery solved. I finally got the auction ads to appear by telling Google I wanted image ads only, not text and image ads.

Noise cancelling

When I moved to this QTH in 2001 the electrical noise on the HF bands was quite low considering this is a high density residential area. But noise levels have risen over the years.

About two years ago I was disturbed to hear a constant "arcing" type noise at a strength 7 to 8 affecting most of the HF bands 15m and below. I first noticed it when switching on the radio after a few months of inactivity, so I did not know exactly when it started. A neighbour had taken delivery of a new TV during that period, and the interference seemed to be coming from his house. However, when I tried to raise the subject the neighbour denied having anything that could be causing interference, so I was unable to pursue it. I resigned myself to having to live with the interference.

A week or so I ago I observed that what sounded like the same interference had increased in strength to S9+10dB on 20 metres, making the band unusable. Again, I wasn't sure exactly when it happened as my interests had been diverted to the higher bands 10, 6 and 2 metres for a couple of weeks. It seems possible that changes I made to the antenna system at the time - such as adding 10m and 6m elements to my multi-band dipole - may have increased the pick-up of the noise. 40m and 80m were similarly afflicted.


I was unenthusiastic about trying to raise the matter with my neighbour again, since even if I was able to show that his TV was causing the interference, it seemed unlikely that he would agree to buy a new one just to keep me happy. Someone recommended trying an MFJ-1026 noise cancelling unit, and reviews on eHam.net suggested that these can be very effective at eliminating noise from a single point source, so I decided to order one. It was delivered on Friday and I spent much of the holiday weekend trying it out.

The principle of the noise canceller is that you have a second antenna which ideally should receive just the noise you want to eliminate. You then balance the signal strength of the two antennas so that the strength of the noise is equal, and adjust the phase of the two signals so that the noise received on the noise antenna cancels out the noise received on the main antenna.

The MFJ-1026 is sometimes advertised as an "active antenna with noise cancelling", and comes with an internal antenna and preamp. However, this is really pathetic at picking up anything on the lower frequencies. Although it did pick up enough noise that I could cancel it out on 20m, this could only be achieved by reducing the gain of the main antenna to a low level to match the noise level received by the whip. It was completely ineffective on 80m.

The manual advises erecting a second "sense" antenna. However, even a random length of wire proved to be a poor antenna on the lower frequencies, just as it would be for normal receive use. What's needed, if you are going to use a non-resonant antenna, is an ATU to get the maximum signal out of it.

It occurred to me that I could use my Wonder Wand L-Match antenna (similar to a Miracle Whip) as my noise antenna. This picked up a considerably stronger noise signal on 20m, allowing a higher main antenna gain level to be used. It also allowed the MFJ to cancel noise more effectively on 80m.

I have made an MP3 sound clip that demonstrates the difference that the MFJ-1026 noise cancelling unit makes on 20 metres. The first five seconds of the recording are with the noise canceller switched out so you can hear the interference at full strength. Then I switch the noise canceller in, the noise vanishes, and you can hear a conversation that you could barely tell was there, never mind copy, before. I switch the noise canceller out and in again a couple more times during the 30 second clip so you can hear the contrast. Click here to listen to the sound clip.

I should be overjoyed to have eliminated this awful interference from 20m, my favourite band. Actually, I am a bit depressed. It all seems like a messy kludge. I had to rearrange my desk and make up new cables to locate the MFJ-1026 to one side, because it was rather an eyesore sitting above the K3 right in front of me with the huge Wonder Wand telescopic practically touching the ceiling.

Changing bands is no longer just a button press. I have to tune the Wonder Wand and then adjust the MFJ's level and phase controls as the settings change from band to band, even from one end of the band to another. And the MFJ-1026 is no use with my little QRP rigs, which don't generate a strong enough signal to activate the RF-sensed TX changeover in the noise cancelling unit.

But when all is said and done, the MFJ-1026 does a pretty good job of removing this particular noise source. It's better than going QRT, which may well have been the only alternative.

Monday, May 25, 2009

6 metre bonanza

There was a great 6m opening here in Europe this afternoon. I thought something must be up when I heard loud signals from G stations as near as South Manchester on 20 metres. With skip that short, there had to be sporadic-E about.

I'm sure that stations with beams and good VHF locations did better than me, but I was happy to work Croatia, Italy, Austria, Hungary, the Czech Republic, Switzerland, France and Germany on my attic dipole. One that got away was Cyprus. He was a good signal, but there was a huge pile-up, and I dislike pile-ups.

I was using my Elecraft K3 and the latest version of KComm, my logging software, which can now spot VHF contacts to the DX Cluster in the format used by propagation analysis sites like vhfdx.net. Unfortunately the program seemed to be doing some odd things, like overwriting a station's locator or name field in the log with one from an earlier contact. And it crashed after adding one contact. I don't know when, if ever,the program will be stable enough that I can release a new version for others to use.

Friday, May 22, 2009

Tiny QRP

Everyone knows that QRP means "low power" and that the accepted definition of a QRP operator is one who uses 5 watts or less (10 watts PEP on SSB.) But to me, true QRP doesn't just mean low power, it means small, simple and low cost. Some folks who have worked DXCC with QRO and big antennas decide to do it all over again using QRP for a new challenge - but with the same antennas. To me, that's cheating. A 3 element beam on a 70 foot tower is not QRP, no matter how little power you are running.

A little interest of mine (pun intended) is to try to put together the smallest possible effective QRP HF radio station. With my MFJ Cub, wall wart power supply and a Christmas Key from Milestone Technologies I have a QRP 20m station that is pretty cute and small. However, the Cub only has a headphone output, and sometimes it's nice to be able to listen on a speaker. So I was interested when I saw a speaker for MP3 players advertised on Ebay that looked like it was MFJ Cub-sized.

When the speaker arrived, it was smaller than I imagined from the seller's picture, as you can see. It has a built-in audio amplifier and a lithium rechargeable battery, which must be recharged from a computer USB port. This is a bit of an inconvenience, and I wondered whether I could rig a power source from within the MFJ Cub so it could be powered (or charged) from the Cub's 12V supply, but I'll leave that for now.

The problem I didn't anticipate was that the jack for plugging into the Cub (or MP3 player) is captive, and folds out of the speaker. This made it impossible to plug it into the Cub at all. However, I cut the jack and a short length of cable off an unused pair of ear buds and was able to solder this to the jack (which fortunately is gold plated and very easy to solder to). I was then able to retract the jack back into its slot with the wire attached, and stick a piece of black insulating tape over the slot. The result: a tiny speaker with an attached lead that I can plug into the MFJ Cub or any other of my small QRP rigs that don't have a speaker output.

I think it looks rather cool.

Thursday, May 21, 2009

Sheer incompetence

I'm sorry. This has nothing at all to do with ham radio, and it may mean nothing to readers from outside the UK who have not been following the sorry saga of our Members of Parliament expenses frauds. But some things just have to be said.

The number of times an MP has "explained" a claim for something they didn't have or weren't entitled to claim for as "a mistake" beggars belief. How can claiming for interest payments on a mortgage they didn't have be "a mistake"? If these people are that careless, ought they to be in charge of running the country? (Come to think of it, they were that careless. Look at the banking crisis and the mess the country is in.)

If I accidentally forget to declare some income on my tax return and the Inland Revenue finds out, presumably all I have to do is say "it was a mistake" and they will let me off? Somehow I doubt it.

I don't just think there should be an election, I think that every MP who has been found to have made fraudulent expenses claims - in other words, just about all of them - should face prosecution, just like any member of the electorate who made a fraudulent claim for social security benefits. Which means that none of them would be able to stand for re-election. That would teach them.

Tuesday, May 19, 2009

Odd K3 bug

One of the things I did today was add support for a Griffin PowerMate USB tuning knob to the next version of KComm. Why did I do that? Well, I have had a PowerMate kicking around in a drawer for a couple of years after I used it with my old and now abandoned K2 remote program K2Net. I have persistent back problems at the moment. Reaching forward to turn the dial of the radio doesn't help. And while I've been searching for Sporadic-E signals on 10, 6 and 2 I've been doing a lot of that. So I thought that having the Griffin PowerMate as a tuning knob right at the edge of the desk within easy reach would be easier on my back.

First, I had a fight to get the PowerMate working. It would appear that the version 1 point whatever drivers that came with the PowerMate originally don't work under Windows XP SP 2 or 3. I eventually hit on the idea of seeing if there was newer software on the web and downloaded a version 2 point something. That solved the problem, and I soon was able to tune the K3 through KComm by getting the PowerMate to send special keystrokes that are converted into VFO UP and DN (down) commands.

This was working nicely until I happened to be listening to a CW signal and could not tune away. The K3's CW decoder was turned on, a fact that I did not initially consider relevant. After checking the source code, using the communications trace window to check that the UP and DN commands were being sent to the radio, I spotted that the commands were being ignored whilst the K3 was displaying decoded CW text in the VFO B part of the display. When no CW is received after a few seconds, the display reverts to showing the VFO B frequency and if you are twiddling the PowerMate knob at the time, the K3 suddenly starts changing frequency.

I've informed Wayne at Elecraft, and he is looking into it.

A tropo summer?

According to the BBC website, the UK should prepare for a heatwave. Well, that will be nice, especially as few of us can afford to go abroad to seek the sun thanks to the government's incompetent handling of the economy and a pound that is worth little more than a euro.

Of course, in nanny state Britain, such news must be accompanied by advice on how to teach grandmothers to suck eggs, etc. According to the article, the Department of Health has advised that homeowners can stay cool by painting their houses white and planting shrubs for shade. I can see that going down well in a conservation area. I presume this advice came from the same source as the recommendation to place a paper bag over your head in the event of a nuclear attack.

Another "tip" for surviving the heatwave is to "identify the coolest room in the house." Thanks, I'd never have thought of that. It seems to me a good place to start cutting government spending would be to sack all these so-called advisors.

One possible benefit of this heatwave, should it actually come to pass (and I'll believe it when I see it) is that we could have a summer of fantastic tropospheric propagation on 2 metres.

Does anyone remember the scorching summer of 1976? I was licensed as G8ILO then, and had an IC-202E - a little 3W output battery powered SSB portable transceiver. When I got home from work in London I used to drive to a nearby high spot and make S9+ contacts into continental Europe. I even worked Belgium on the built-in telescopic whip of the IC-202 from inside a pub!

These exceptional conditions hung around for several weeks. Some people got so tired of working Europeans they started calling "CQ G only." That was a fun time. If it happens again, it will be a nice compensation for the lack of sunspots.

DC20B keyer bug fixed

The keyer in my DC20B 20m QRP transceiver now works properly. Chuck Carpenter, W5USJ, read about the problem I and others had found with the keyer of the DCxxB QRP transceiver kits not playing back memories correctly, and sent me a Tiny 13 keyer chip that he had from an older kit to try out. I plugged it in to my DC20B and it now plays back memories perfectly. This would seem to suggest that there is a programming bug in the keyer chips supplied with the newer kits.

The list of problems experienced by purchasers of these kits is not encouraging: long waits for delivery, missing parts, replacements not forthcoming, sidetone much too loud, unpleasant noise burst between characters when sending, the keyer bug just mentioned, broadcast band breakthrough, and the TX frequency of the 20m kits being 1KHz too high. And as the designer has apparently gone backpacking, no "official" help with any of these issues appears to be available. Very disappointing.

Monday, May 18, 2009

Antenna Activities

Over the weekend an RSGB 144MHz contest was on. I thought that this would be an excellent time to find out how my newly constructed Moxon Rectangle performed. For a few hours after the start of the contest I heard nothing at all. But eventually I did hear some fairly weak signals, enough to show that the Moxon has noticeable directivity and a good front to back ratio, exactly as it should have. Unfortunately it also illustrated that VHF under normal conditions offers only line of sight propagation and if there is a mountain range between you and all the other band activity you won't hear anything.

On Sunday, a blizzard of red lines on the VHFDX.net 50MHz real-time map spurred me to go up in the attic and add a 6 metre dipole to my four band attic dipole. This was a nice easy job - I only had to prune an inch from the end of each length of wire to get the resonant point where I wanted it, and I observed no effect whatever to the 80, 40, 20 and 10m dipoles sharing the same balun and feeder. I haven't worked anyone on 6m yet but I heard a beacon in Croatia coming in nice and strong, so it's just a case of waiting for the Sporadic-E clouds to be in the right position. I used to have a delta loop for 6m and made Sporadic-E contacts in the past using my K2 and a Ten-Tec transverter with just 7W output, so I'm hoping the K3 with 100W will do even better.

While I was in the attic I decided to get the Moxon Rectangle out of the shack and up there as well. I used the length of RG-58/U left when I took down the 2m Slim Jim, and since I didn't have any way of fixing it I just rested it pointing south east on an old plastic beer crate normally used as a support for cutting wood and similar tasks in the garden. My wife wants the beer crate back so I have ordered a cheap antenna rotator on eBay which will allow me to turn this little beam from inside the shack. I hope this will not prove to be a waste of money. The 2m contest showed that I cannot expect to hear anything during normal conditions on Two so I'm hoping that tropo and Sporadic-E will still be capable of bringing signals over the mountains and allow me to make some interesting VHF contacts in the coming months.

Sunday, May 17, 2009

QSL Bullies

I was listening to some stations coming in on Sporadic-E on 10m this morning, and heard S51GL answer a contact. He replied with something along the lines of: "We worked before on 20 September 2004 on 20 metres. I sent you a QSL card through the bureau three years ago. If you don't have computer logging perhaps this will jog your memory to send a card as the final courtesy of a QSO." I heard him reply to several stations in this manner, saying "I will not send a QSL for this contact as I still do not have a card for my last contact from you." I called him, and he said something similar to me, which was odd as I do not have a previous contact with S51GL in my log.

Even if I had, he probably would not have had a QSL from me as I do not reply to bureau cards. It's not a deliberate policy of mine, it's just that the chore of going through the log checking each card received from the bureau to see if it's a valid contact and I haven't sent a card already is something I can never muster the enthusiasm to get round to.

If I think a contact is worth a paper QSL I will make one out at the time. Quite often I will even send it direct, at my own expense. But I do not think it is worth QSLing every rubber stamp QSO, every exchange of PSK31 brag files, every point given out in a contest. That's just spam. Nobody needs a card from me for an award - if they do, they should say so. I think a QSL is merited only if the contact is memorable in some way, like an especially enjoyable ragchew, or someone's first-ever contact.

I do log every contact automatically with eQSL.cc, so if anyone really needs a card from me they can log on to the site and have one printed and mailed to them for a couple of dollars. It does perplex me that this hobby, which has embraced the Internet in every way when it comes to getting contacts, from DX clusters to propagation alerters and chat forums to arrange skeds, is so unwilling to use it for confirmation of those contacts. Come on guys, it's the 21st century!

It may have been worthwhile in the early days of radio when every contact was a real event, but sending a QSL to confirm every contact today is totally unnecessary and a waste of time, money and paper.

Saturday, May 16, 2009

Receiving GB3VHF

I have always had rather a soft spot for the 2 metre band. It was the band that got me interested in getting my own ham radio license. I had built radios and other things in my teens and listened on the short wave bands, but I was never particularly interested in being able to transmit.

Then one day, wanting something different to build, I decided to construct a 2 metre super-regenerative receiver from a design in Practical Wireless. Using it, I heard stations fairly local to me having interesting conversations in English. And that was the motivation that led to me coming on the air as G8ILO in 1974.

In those days, 2 metres was not the busy band it is in parts of the country today. There were no ready-built rigs at all. Most of the operation was on AM using home-built crystal controlled transmitters, or modified ex-taxi radios. For receive people used a down-converter and short wave receiver. You would call CQ and then tune a chunk of the band, low to high, listening for any replies.

A lot of the time, there was nothing to hear on two metres at all. Except that at the time, I lived about 30 miles from the Wrotham beacon, GB3VHF. That was something I could always hear. So I got very used to hearing dah-dah-dit dah-dit-dit-dit dit-dit-dit-dah-dah dit-dit-dit-dah dit-dit-dit-dit dit-dit-dah-dit daaaaaah ...

Up here in the back of beyond I'm about as far from the Wrotham beacon as it's possible to get and still be in England, and I'm screened by the Cumbrian Mountains from the rest of the country. The only SSB activity I have heard so far has been someone operating from one of the Lake District mountains that I can almost see from the window. So at the moment I have no idea whether this QTH has any potential for working VHF at all.

Turning the pages of the latest RadCom a couple of days ago I saw on the Data page a screenshot of GB3VHF being received using the WSJT software. Apparently, GB3VHF gives its call and locator once every two minutes in JT65B mode. So I thought I would give it a try.

The picture above shows GB3VHF being received using the Moxon Rectangle I constructed a couple of days ago, which is still clamped to a shelf just above the shack door. I don't hear it every time. I guess I get a decode about once every ten minutes or so. I was going to upload an MP3 of the audio that was recorded during that one minute period, but you can't hear anything at all except noise.

It would be interesting to know if there are any other 2 metre beacons that transmit using JT65B which I could try to copy. And how far afield GB3VHF can be received using this mode. But if all I can receive from here on 2m is a distant beacon buried deep in the noise, it probably isn't worth spending much more time and effort on the band.

Friday, May 15, 2009

E-type propagation

Back in the early 1980s I was starting to get a bit disenchanted with life as a wage slave, and had ideas about making my living as a freelance writer. It took ten more years to realise this ambition. However, my earliest steps in that direction were to write articles for hobby computing and ham radio magazines.

One of the radio articles I was most proud of was an article about sporadic-E propagation that I wrote and had published in the short-lived magazine Amateur Radio in 1984. Now that the sporadic-E season is upon us once again, and now that I possess a scanner with OCR capability, I thought I would scan the article and add it to the website.

Although a few major changes have occurred since the article was written, such as the break-up of the Soviet Union and the use of the Internet to spread news of band openings, the basic facts explained in the article remain true to this day, and I think it is a useful tutorial in this interesting mode of propagation. So I am happy to present for your education and enjoyment, E-type Propagation almost exactly as it was originally written.

Thursday, May 14, 2009

Another year closer to old-fartdom

It's my birthday. I'm now 56. So forgive me if today's post is a journey into old-fartdom.

I'm old enough to remember when children's TV presenters were respectable, well-educated role models who spoke correctly and wore smart dress. The men wore jackets and ties. Then the trendies and the politically correct brigade took over and decided that these role models were too middle class. So these presenters were replaced by younger people in casual clothes, who talked in the vernacular and didn't present anything too challenging. I often wonder if changes such as this - which were also reflected in the teaching profession - are responsible for the lack of aspiration of many young people today?

The June issue of RadCom arrived today. It contains a letter from a gentleman with an M3 callsign who complains that too much of the content is aimed at "engineers and people who have been in the hobby all their lives." He hints that he may not renew his subscription if so much of the content is beyond his understanding.

I'm not an engineer, and although I have been in this hobby now for over 40 years I started off knowing as little as this M3. When I first joined the RSGB RadCom was more like a learned professional journal. There was nearly always one major constructional project ongoing (remember the G2DAF transceiver?) plus several pages of the late lamented Technical Topics looking at different types of circuits and antennas. I didn't understand much of it, but I tried, and sometimes I succeeded and learned something. Most of what I know about radio today I learned from RadCom and other RSGB and ARRL publications. (I've probably forgotten most of it again, but that's another matter entirely.)

Now I'm all for encouraging new people into the hobby, and you'll never hear me complaining about no-code hams. The use of morse code is not compulsory so I see no reason for continuing to require a knowledge of it before being granted an amateur radio license, any more than people should be required to know how to use PSK31 or SSTV. But ham radio is a technical hobby. It isn't CB. It isn't for people who just want to talk to other people by pressing a PTT, without understanding how their voice gets from A to B.

If we have got to the stage where people who have got an amateur radio license complain that the national radio society's magazine contains too much technical material, perhaps the dumbing down of the hobby in order to attract new people has gone too far. Perhaps people are being encouraged to become radio amateurs whose needs would have been satisfied very well by CB.

I hope that the RSGB will not be tempted to dumb down RadCom to the dire level that has been reached by the ARRL's QST.

Wednesday, May 13, 2009

More on the Elecraft K3 2m option

More information has been released about the new Elecraft internal two metre transverter that will be shown at Dayton. Elecraft has just posted a PDF data sheet for the K144XV module. The product won't actually be available for a few months, though.

It looks a pretty small board, but it will probably cost more than you would imagine from the picture, as you get a new fan panel with a BNC jack for the 2m antenna, and a replacement KXV3 module is needed to provide the IF output. Probably you need a new side panel as well, since that is what it is attached to. But they've managed to fit it in along with the KXV sub-receiver module.

It's a neat option, but personally I'd have been more interested in a new range of Elecraft's existing transverters, styled to match the K3 (rather than being in K2 gray/green) and with better heat sinking to withstand continuous duty modes like FM. Perhaps next year.

A lean sunspot cycle

I have just updated the file of smoothed sunspot numbers for VOAProp, having been informed that NOAA had recently updated its predictions. I see that NOAA is currently predicting that the peak of the next sunspot cycle, Cycle 24, will be reached in July 2012 with a SSN of just 96.4. This compares with a peak of 115.5 for Cycle 23 in November 2001, while Cycle 22 in the summer of 1989 managed over 158. Previous solar cycles have maintained higher SSNs for several years than will be reached by the peak of this one.

Looking back in the table, it seems that conditions will not have been so bad during a solar maximum since Cycle 16 which only managed a peak SSN of 78.1. But in April 1928 there were not too many radio amateurs around to complain about it!

Moxon Rectangle for 2m

Ever since getting a 2m transverter to use with my K3, I have had thoughts of putting up a horizontal 2m antenna to try SSB.

From my location here in the far north western corner of the UK, with the Cumbrian Mountains between me and the rest of England (who are all beaming the other way towards Europe in any case) I have never heard any 2m SSB activity. So it is certainly not worth the cost of putting up a commercial beam and rotator outside, even if I could get away with it. What I had in mind was a small addition to my attic antenna farm, that might enable me to make a few contacts during tropo or Sporadic-E openings.

I was considering an HB9CV, but its feed arrangements are a bit complicated and I wondered whether a simple two element Yagi would be a more practical choice. However there are so many permutations of 2 element Yagis that I became confused. Then I happened across the Moxon Rectangle, which happens to have not only its own website, but even a little software utility written in Visual Basic that will print out a design for you.

The attractions of the Moxon Rectangle over any other small two-metre beam is that it is small - narrower than a conventional Yagi with a boom only a foot long - and is a perfect match to 50 ohms, eliminating fiddly matching arrangements. It is also claimed to have a very good front-to-back ratio. However, when the design utility spit out dimensions in fractions of a millimetre, I started to wonder whether this was superfluous accuracy or whether the design is so finicky it would be impossible for someone with my limited DIY skills to replicate accurately enough for it to work.

I used some wood that I had in the garage for a boom, while the elements were made from four 6mm diameter lengths of aluminium tubing that had once formed a crossed dipole antenna for the weather satellite band. Two of these elements had to be joined at the center for the director of the antenna, the other two formed the driven element. To mount them to the wooden boom and provide attachments for the feeder I used 30 Amp electrical "chocolate block" connectors. What I didn't realise until later was that the brass connectors aren't a tight fit inside the nylon insulating housing, which made for an antenna with slightly wobbly elements.

When I first connected a very short length of co-ax to hook the antenna up to the AA-200 antenna analyzer I got a nice sharp dip going down to a perfect 1:1 SWR - at 135MHz! For a moment this caused me to wonder whether the elements had somehow remembered the original frequency they were used on! I then connected a longer length of RG-8X Mini, and got a much broader SWR curve with double-dips either side of the intended center frequency of 144.3MHz. Fiddling about with the gap between the director and driven element didn't seem to make too much difference either to the position of these dips or the SWR.

With this odd and rather broad SWR curve I'm not sure if the antenna is working properly. With so little 2m activity in these parts it's difficult to test it. I can open up GB3AS, the local repeater, and when I wave the antenna around there appears to be some directivity, but whether it's as good as it could be I can't be sure.

For the moment my 2m Moxon Rectangle is attached by a G clamp to a top shelf in my shack, pointing out the window in an east-south-easterly direction, in the hope that a Sporadic-E opening will occur and I can find out if it works or not. Only then will I decide whether it is worth putting it in the loft, if I should get a commercially made small 2m beam, or whether I should give up the whole idea.

Morse Machine

Due to the fact that Tech-Pro.net will be closing its software downloads site, G4ILO's Shack has become the download page for Morse Machine.

Morse Machine is a Windows version of the Morse Code Teaching Machine designed by Ward Cunningham which was first described in an article in the May 1977 issue of QST. It teaches Morse code using the Koch Method. Right from the start, it sends quickly enough that you have to learn to recognize the sound of each letter. The slowest speed it will send at is 20wpm, so after a few weeks of practise you will be able to receive at that speed.

When I first discovered this program I was extremely impressed, and made huge improvements in my Morse receiving speed. However, the Windows version of the program on the original site is not a true Windows program, and it caused 100% CPU usage while it was running, which made my laptop overheat. That is not a problem with this version (which also runs very well on Linux using wine.)

Tuesday, May 12, 2009

Internal 144MHz transverter for K3

I just received a beta version of new K3 firmware from Elecraft. Among the list of changes is:
  • CONFIG:XVn ADR now has a new option: Int trn1. This routes transverter I/O to the K144XV, the K3’s new internal 2-m transverter option.
This new option does not appear to be announced anywhere on Elecraft's site yet. I presume that it will be formally announced at Dayton. You read it here first!

I don't think there is enough 2m activity in this neck of the woods to justify my getting one in place of my old cheapo Spectrum Communications transverter, but I'm fascinated to know where in the K3 it will go, and what antenna socket it uses. Perhaps it fits in place of the KXV3 module and the KRX3 sub-receiver?

VHF propagation alerter

With the Sporadic-E season started, I've been looking for ways to be alerted whenever there is an opening. It's all very well to say "real amateurs tune the bands" but although I'm at home most of the time and well placed to take advantage of band openings, I'm usually busy working or doing something else and don't have time to spend endlessly tuning the receiver.

So I was pleased to discover DX Sherlock, a system operated by VHFDX.net. This is a set of maps for each of the bands, 10m and up, that shows contacts which have been reported to its DX Cluster. These contacts are plotted on the map, and easily show the position of any Sporadic-E clouds, as you can see in the example.

Another nice thing about this site is that you can have email alerts of Sporadic-E, tropo and other propagation events on the bands you are interested in sent to your mailbox. If you are interested in the higher frequency bands, this site is well worth a visit.

Monday, May 11, 2009

Frequently annoying questions

One reason why I sometimes wish I had never bothered releasing certain programs I have written is that it irritates me that people cannot be bothered to read the documentation and then write complaining about an "error" that has been explained in the online FAQ.

The program that causes most of the annoyance is VOAProp, since it is the most widely used of the programs on my site and has the most users. The most frequently annoying question is "why does the program display a SSN of 10 (or whatever) when there are no sunspots." The answer is that, in accordance with documentation for VOACAP, which VOAProp is merely a front-end to, the program uses a smoothed sunspot number, which is a rolling average over six months either side of the month in question. This value has nothing to do with the number of spots on the Sun's disc at this moment. This may not make sense to some people, but that is what VOACAP's developers say should be used, so that is what VOAProp uses.

The predicted values do get updated every month by NOAA, so the value used by VOAProp may not always be exactly the value in the latest data. I long ago lost interest in VOAProp and no longer develop or support it. And I can't be bothered to update the data file every time NOAA updates its predictions. Whether the value used is 10, or 5, or 15, makes little difference to the propagation map generated, so it just isn't worth the effort of updating it on a monthly basis.

One thing VOAProp demonstrates is that HF propagation cannot be predicted except in the broadest terms. As I have written before (probably somewhere in the VOAProp documentation), propagation models such as VOACAP were developed by short wave broadcasters who were interested in which bands would give solid, reliable propagation to a particular area most of the time. They were not developed with amateur radio DXers in mind, who are more interested in what might be workable on the fringes of propagation. So they will tell you where your signal is likely to be strong, but they are not very good at predicting where weak contacts might be possible.

The main purpose of VOAProp, and the one it was actually written for, is to demonstrate to inexperienced hams how propagation changes from day to day, month to month and between periods of high and low solar activity, so that they can learn which bands and what times of day offer the best chance of working certain areas. But the ionosphere will always have some surprises in store, which for me is what makes HF operation interesting.

Friday, May 08, 2009

Logitech S120 speakers NOT RFI proof!

I am presently using a pair of cheap Trust SoundForce 2.0 computer speakers bought from the local Wilkinson store with my K3. I originally bought them to use with the JUMA since Gerd, who built it, hadn't included an internal speaker. However I tried them with the K3 and found they were a substantial improvement over the built-in speaker. With several turns of the input cable wound on a clamp-on ferrite choke they are almost immune to RF breakthrough even at 100 watts. However, they look a bit cheap, I felt that the sound quality could be improved, and I would be happier with no RF breakthrough at all. Unfortunately I couldn't find a way to get inside the speakers to install extra suppression components. Although two Philips screws are visible at the base of each speaker, the case halves also appear to be secured by screws at the top which are behind the grille, and I can't see how to get that off.

I saw some Logitech S120 speakers advertised on eBay and decided to order them, as they looked a lot nicer, and matched the K3 better, and since they came from a "quality" brand I thought their RFI suppression might be better. The speakers arrived this morning, and are already back in their box to take to the local charity shop. The sound quality was better than the Trust speakers, but I was dubious about how they would perform in the presence of RF since even before I transmitted I could hear a faint pulsating buzz - possibly caused by a DECT telephone base station nearby.

The RFI was so bad I got a loud buzz even on a tuning signal of 6W on 15m. When I pressed the mic PTT with 100W selected the resulting burst of noise almost caused a need for a clean pair of underpants! I can see no way to get inside these speakers at all, but even if I could, it would probably be hard to make them RFI proof given that the susceptibility is so bad to start with.

My loss is the Oxfam shop's gain - and yours, as I decided to post this as a warning. Logitech S120 speakers should not be used within 100m of a radio transmitter - at a very minimum. I just hope that none of the neighbours is using them.

PSK31 with the AGC off

I'd been planning to write about this for a couple of days, but originally decided to leave it for later as I've already made two posts today. But a comment in W9OY's blog made me decide to write about it now.

W9OY wrote: "I especially don't like PSK31. The dynamic range of that mode is so bad that if you run more than 30W you blow out everyone's receiver and mess up the throughput." What nonsense!

For years I've worked PSK31 with the AGC on. Unlike many people, who have grown accustomed to the "Digipan" style of operation where you look at a 2KHz or more slice of spectrum on the waterfall and point and click to select your frequency of operation, I've always tuned the station I want to work to the "sweet spot" and used narrow filters to block out any strong nearby stations that could be depressing the gain and making the one I want to work hard to copy. I especially enjoyed the razor-sharp selectivity of the K3 that allows me to narrow the receiver down to allow just one PSK31 signal through, if necessary.

In the last few days, to allow my station to report received PSK31 activity to PSK Propagation Reporter, I've been operating with the K3 filters opened up as far as they would go, which is 4KHz in data mode at the moment. It's obvious from the waterfall that when a strong signal comes on, it depresses the receiver gain affecting the strength of everything in the passband. The solution, of course, is to turn off the AGC, which on the K3 is as easy as pressing one button.

Normally, if you turn off the AGC, you are in danger of damaging your eardrums. But when working PSK31 you don't really need the sound on at all, so it causes no problem. The only thing you need to watch is that you don't overload the computer sound card, or the receiver line out audio stages. Yesterday evening I noticed multiple traces evenly spaced across the waterfall when a very strong signal came on. The solution was to back off the RF gain a bit. Depending on where the overloading occurred, it might be better to reduce the line output level (something that's also easy to do with the K3.) You can then view a full 3 or 4KHz of spectrum and point and click Digipan style, without any problems caused by loud stations. Assuming their signal is clean, of course - but that's a different matter.

Turning off the AGC and managing the receiver gain by hand is a technique used by "old hand" DXers to copy weak CW stations in the presence of close-by strong ones. If it works for one narrow-band mode, why not use it for another?

PSK31 has poor dynamic range? I don't think so!

JUMA goes home

Last night I had an email from Gerd, DF9TS, who had returned from his business trip and wanted to know if I had managed to finish my evaluation of the JUMA TRX2 transceiver, and was I planning to keep it. I told him that I had enjoyed trying it out, I really liked the sound of the receiver, but I found numerous little annoyances with it.

The filtering is not very flexible, with only three filter widths for all modes. And with the lower filter edge fixed in position, you can't choose the centre frequency for CW or data. Although the receiver has been designed to give exceptionally high dynamic range, the AGC can't be turned off, so you still lose a weak signal that's close to a strong one.

No provision was made for data modes in the original design, and although the JUMA community is developing mods to provide this facility, it is still a bit of a kludge, as it was for the Elecraft K2 (which at least had the excuse that it was designed in 1998 before sound card data modes took off in popularity.) There are also questions as to whether the PA and driver components are adequately heat-sinked for high duty cycle operation such as data modes.

Gerd probably sensed that I was a bit luke-warm about the JUMA, and replied that I should feel under no obligation to keep the radio if I didn't really want it. During successful QRP /P operations while on his trip, he had changed his mind about buying an FT897 which had been his reason for wanting to sell the JUMA in ther first place. So I have decided to send it back to him. I will, of course, pay for the cost of sending it to me in the first place.

I am very glad to have had the opportunity to try the JUMA TRX2. It is still a very nice radio, especially if you are an SSB user (or even CW, if you can adapt to having a spot frequency of 400Hz to get the narrowest filter.) But I am spoiled with the K3, plus too many other QRP radios, for it to be worth buying given the shortcomings I mentioned.

Propadex

If you have visited my site home page today you may have noticed the black "Propadex" button in the left hand sidebar just below my WebProp box. I first saw this in the sidebar on W9OY's blog. Since it bears the legend "QSONET.COM" I thought it must show activity levels on the QSONET fake ham radio system. But in fact, it looks as if it might be a very useful propagation indicator for those of us who like to use real radio.



There doesn't seem to be much information about Propadex other than what you can glean from clicking on the image itself, but apparently it is a propagation index derived from real-time ionospheric F0F2 measurements made by the US Air Force Solar Telescope Network in San Vito, Italy.

The index is relative to conditions averaged over a 60-day period for the same time of day. It shows how much better or worse than average the F0F2 (or Maximum Usable Frequency, MUF) is, in units of 10Hz. So if the index is +100 it means that the MUF is 1MHz higher than the average for this time of day. If it is negative, it shows that conditions are worse than average. It doesn't tell you what the MUF actually is, so you still need to know what average conditions are for it to be useful.

The graph to the left of the number shows the previous eight hours of data. So this could give you an idea of how conditions are changing.

It will be interesting to see how well this indicator correlates with perceived band conditions. It would also be interesting to find out where the data comes from, with a view to adding this indicator to WebProp if it proves to be useful. Another project to add to my ham radio To-Do list!

Wednesday, May 06, 2009

Twenty metres is hopping

There may be no sunspots today but twenty metres is really hopping this afternoon. I'm too busy at the computer to try to work anything at the moment but Fldigi's PSK browser has logged and reported the signals of HL2AXS (Korea), BA5AP (China), YB9EA (Indonesia), 9M2OK (Malaysia), 3B8GT (Mauritius), 5R8FL (Madagascar), UN5DA (Kazakhstan) and a couple of stations from Asiatic Russia.

I heard some of these while I was having my afternoon tea break, and I could have had a go trying to work them if I'd had more time.

Tuesday, May 05, 2009

Too many antennas

I finally found the solution to the problem that was causing bad TX IMD on ten meters and which drove me almost to the point of insanity!

I had taken all the radio equipment off my shack desk, and was slowly replacing and reconnecting it, one piece of gear and one cable at a time. Each time I added something, I checked the PSK31 IMD on ten metres. Nothing seemed to make any difference. I reconnected the 2m transverter to the K3 and everything checked OK, with IMD of >-30dB. There was just one last thing to connect - my 2m Slim Jim. As soon as I connected that, the IMD meter started beeping and reading -14dB plus or minus.

Disconnecting the 2m antenna cable didn't fix the problem, though, although it made a slight improvement. Everything had been perfect, and then reconnecting this one cable had wrecked it. I was tearing my hair out at this stage, as you can imagine.

Then I noticed that moving the unconnected antenna cable about drastically affected the IMD. Grasping the cable in my hand improved the IMD substantially. Gradually it dawned on me that this antenna cable must be re-radiating my 10m signal back into the shack in such a way as to affect the K3 transmitter. When I first disconnected it and moved it out of the way, I had quite by accident found a position where its effect was neutralised.

When I made the Slim Jim I had hung it up in the attic only a few feet from the 10m dipole. Since it was cross-polarized and for a completely different frequency I did not expect that either antenna would have any effect on the other. Both seemed to transmit and receive OK and the SWRs were unaffected. But obviously the Slim Jim was picking up the 10m RF and causing trouble. I had never imagined that the 2m antenna could be the cause of my 10m problem and so I wasted most of the holiday weekend chasing red herrings and running up blind alleys trying to solve it.

This morning I went up into the attic, took down the Slim Jim and coiled up its feeder cable. The loss of 2m FM is no big deal for me. I could probably disguise it outside if necessary, but there is so little 2m activity here anyway that I really can't be bothered.

Just for the hell of it, while I was in the attic I also removed the old 10m dipole and added two 8ft. long elements to my 80/40/20 dipole, fanning them away slightly in a very shallow inverted V. This gains it a bit more height and eliminates the need to manually switch between the two dipoles.

You can get great results with attic antennas, but it's easy to get carried away. You can't just keep stuffing more and more antennas into the loft space without finding that they interact in some way. Between my now 4-band dipole and my MFJ magnetic loop I can use about 80KHz of 80m centered around 3.590, plus all of 40m, 30m, 20m, 17m, 15m and 10m. That's not bad for a situation that would have many people giving up the hobby altogether. I will just have to not be greedy for even more bands in future.

Spark into life

In this hobby, people often reinvent the wheel just because they can. Someone builds a transmitter or receiver from their own design, rather than copying someone else's, because they can learn from it, and because it is interesting. However, I'm not sure this approach makes sense for web-based projects.

This morning I saw a Ham Banner Exchange ad for Sparkgap.net - a ham community. I clicked on the link, and what I saw was a fledgeling site, still with a lot of bits to be implemented, where hams can chat to each other, upload photos, exchange files and so on.

It looks nice enough. But is there a need for it? We already have QRZ.com and eHam.net. There is Yahoo Groups, which hosts hundreds of ham-radio related mailing lists. Personally I hate Yahoo Groups, because I don't want my email box filling up with that stuff, and the web interface stinks. But most hams seem to be happy with it. Then there is 73s.org, which many object to just because of its name, but which seems to be doing what Sparkgap is seeking to do, and actually appears to be gaining popularity.

The trouble with web communities is that they need a critical mass in order to be successful. A year or so ago Thom K3HRN and I set up the Zerobeat Forum. We did actually have an excuse - it was originally meant to be just a demo of what a web forum alternative to the Elecraft email reflector could be like, though the Elecraft users, being a forward looking bunch, mostly shunned it. But since we had done the work, we decided to open some more general discussion boards and make it available as a more general ham radio forum. However, although there are nearly 400 members, the last post was more than two months ago. You need a lot more participation than that to keep people coming back, and when there are already active forums in existence, there is little incentive for people to visit another one.

It's probably time to pull the plug on the Zerobeat Forums. Whether Sparkgap.net will eventually suffer the same fate only time will tell.

Monday, May 04, 2009

Ten metre madness

My intention to try PSK31 beaconing on 10m to detect sporadic-E propagation turned into something of a nightmare. What happened was this. I saw traces on the waterfall and decided to kill the beacon macro and make some contacts. I called Roberto, IV3KFV, and he came back with a report of 527 and said that my IMD was <-9db! I switched on the IMD Meter and sure enough, it was beeping away reporting IMD of around -15dB. Something was wrong and I would obviously have to investigate.

What then followed took up more than a day and had so many red herrings and false hopes it seemed as if someone "up there" was determined to drive me crazy. Many times it would seem as if I had found the cause of the problem, and then the next moment it was back.

Finally I got to the point where I had the K3 with just the power supply and the antenna connected to it. Nothing else. Using the internally generated PSK I was still getting poor IMD when transmitting into the antenna. I then discovered that putting clamp-on ferrite chokes on to the power supply cable affected the IMD. With 4 chokes on the cable the IMD was perfectly acceptable. I moved the K3 back into the operating position and ... bad IMD again!

At this point I was out of ideas. Changing the antenna or moving it further away from the transceiver was not an option. Due to back problems it's difficult for me even to work on antennas. Operating from the ground floor where a better RF ground would be possible was not an option either. With a house this small and antenna restrictions you don't get a lot of choices.

I decided that the K3 must just not like something about the situation here, so I thought I would try all my other radios to see if any of them was less fussy. I tried the the K2 (both with its power supply and running from its internal battery), the Juma TRX2 and the FT-817 (from power supply and battery.) All radios exhibited poor IMD when transmitting into the 10m antenna, compared to when a dummy load was used!

The obvious conclusion is that the IMD Meter is giving false readings. However, that doesn't explain the bad signal report I received from IV3KFV. Having a local who could come on and listen to my signal would be a big help at this point. Unfortunately there are no other active amateurs here in Cockermouth that I know of. The nearest amateur to me is a few miles away, on the other side of a hill, not a strong signal at the best of times, and I rarely hear him on the air.

So the problem remains unsolved and I am giving up. There is only so much time you can spend trying to fix something and getting nowhere. If I can't work on 10 metres then I guess I'll just have to live with it.

Saturday, May 02, 2009

Using PSK31 to spot sporadic-E on 10m

There are a few mutterings online that 10 metres has been open for sporadic-E (a.k.a. short-skip) contacts in the last couple of days. These openings are, as the name suggests, sporadic, so they are impossible to forecast and easy to miss.

You can, of course, tune around and listen for the numerous beacons that are alleged to be operating on 10m. However, if tuning around listening to band noise for hours is not your bag of chips, and you'd rather do something more useful while waiting for 10 metres to open, here's a suggestion. Why not set up a 10m PSK31 beacon and use the PSK Reporter network to alert you when an opening starts?

All you need to do is set your transceiver to 28.120MHz, PSK31 mode, use one of the digimode programs like Fldigi or DM780 that can send spots to the reporting network, and program a macro that sends BEACON DE yourcall yourcall yourlocator every minute or so. Then sit back and wait for the results!

I'm beaconing now on 28.120. Anyone care to join me?