Friday, September 18, 2009

Review of the Elecraft K3 - 18 months on

As my Elecraft K3 is the most important and most expensive item in G4ILO's shack, some readers may be surprised that there is no review of the K3 in the Reviews part of my main site. The reason is that when I first received my K3, I was quite disappointed with it in many respects. Many of the advertised features that were among my reasons for buying it were still to be implemented when the first radios were shipped. Any criticisms that I made at the time would, I hoped, eventually be obsolete. So I have been waiting for the K3 to be finished, or at least to perform as I expected, before giving my verdict.

More than two years on from its inital announcement, there are still things I'm waiting to be finished or put right. So I am publishing my review in this blog, representing my thoughts about the K3 at this point in time, rather than on my main site.


My reason for buying a K3 in the first place was largely just that it was an Elecraft. I had owned a K2 since 1999 and had built several of Elecraft's other kits and modules since then. I have great admiration for Wayne, Eric and Lyle, talented radio designers who are not afraid to take an innovative approach in their products. They are also enthusiasts for their products, happy to engage with their customers through the medium of the Elecraft email reflector and respond to criticisms and suggestions. Elecraft has a first class reputation for customer service. None of what I am about to say is intended to detract from any of that.

However, the K3 exists in a marketplace alongside many other radios. It has to be judged by how well it compares with its competitors in its price class, and those below it. The K3 is lauded as the radio with the best receiver, able to pick out weak signals in close proximity to very strong ones better than anything else. It is claimed to be better even than more expensive radios. For many keen DXers and contesters that capability may be important enough to justify overlooking other failings. However, for the average Joe Ham with poor to moderate antennas in a QTH with a noise floor filled with emissions from electrical devices such receive characteristics have little value except perhaps to give bragging rights. It's how a radio looks, feels and sounds, how comfortable it is to use, and how suitable it is to use for modes other than CW or SSB that many of us ordinary Joes like to play with, that are more important. And in many of those respects the K3 does worse even than some of the cheaper alternatives.

To start with the look and feel. In pictures, and at a normal operating distance, the K3 looks quite attractive, if stylistically somewhat dated. However, its compact size suggests a medium to budget price radio rather than high end. Close up, the many screw heads scream "kit built", which of course my K3 actually is. The poor paint finish, plasticky knobs and gritty feel of the smaller rotary controls also say "cheap", which the K3 certainly isn't. Some of those complaints are a direct result of the K3's design aims to be user assembled (if required) and light weight (for portable / DXpedition use.) But most K3s are bought ready assembled and never move from the shack desk, and those buyers may not accept the compromise. The K3 is a radio to use, not drool over.

A frequent criticism of the K3 is its poor ergonomics, especially the lack of separate band and mode buttons. It's a fair point: other radios in the K3's price class are bigger, and can afford larger controls and more buttons. Both my K2 and my FT-817 select band and mode by up/down buttons, so I felt that was something I could live with. Still, every time I change band instead of mode, or vice-versa, or go the long way round when making a change, it's annoying. And it's an annoyance that's easier to forgive in a radio as tiny as the FT-817 that has no choice in the matter than in one that could easily have been made a bit bigger.

My biggest irritation with the K3 is its implementation of memories, which is frankly awful. In every other radio I have used, you press a button labelled V/M to switch between VFO and memory mode. In memory mode, you immediately go to the last selected memory band/frequency/mode etc. The VFO or some other control then acts as a rotary switch which you can use to change memory, and as you change memory the transceiver immediately changes to the band/frequency/mode/repeater shift etc. that has been programmed into that memory.

In the K3 there's no such thing as memory mode. You're always in VFO mode. You can save frequencies to memory and you can recall frequencies from memory back to the VFO but when you press M>V the first time, the radio stays where it is. You rotate the VFO and the memories change in the display, but the radio still stays where it is. You're just selecting the memory you want. You have to press M>V again before it will actually change to the band/mode stored in the displayed memory. So you can't use the memories to quickly check for activity on a list of different frequencies. This probably isn't something SSB or CW operators want to do. But FM users will, for sure.

To try to overcome this criticism and provide the opportunity to scan a selection of memories Elecraft introduced what they call channel-hopping memories. This smacks of an ugly workaround that is only a small improvement over the standard behaviour. To make a memory a channel-hopping memory you must give up the first of the characters in the alphanumeric tag for each memory and make it an asterisk. The K3 will then scan a block of consecutive memories tagged with an asterisk almost as if you're in memory mode, actually changing the frequency and listening there. You can even rotate the VFO to scan through them manually. But - and here's the killer - it ONLY changes the frequency, not the mode, the repeater shift, the access tone or any other attribute stored in that memory. To fully load the memory into the radio, you still need to press V/M again.

The K3 has the best support of any radio, bar none, for using transverters. You can have up to 10 of them, and the K3 will display the actual VHF or UHF frequency when you are using one. Elecraft has even brought out an internal 10W 2m transverter option. So this is a radio that many people might choose over an IC-746PRO or TS-2000 to use not just on HF, but also to chat on their local 2m simplex and repeater channels. You'll need to buy a crystal filter for FM, which will cost you as much as a budget 2m HT, but you'll get a receiver whose adjacent channel rejection and signal to noise ratio on weak signals is second to none, and certainly better than the aforementioned Icom and Kenwood radios.

But the K3's flawed implementation of memories makes FM operation a nightmare. You program your local repeaters and simplex channels into a block of channel-hopping memories. You then select a repeater and make a call through it, successfully. Then you turn the dial to cycle through the other memories and hear someone calling on a simplex channel. You call, but they won't hear you because you didn't remember to press M>V so you still have the repeater shift on from your previous contact. Or if you hear someone on another repeater and call him, you may have the wrong access tone. The K3's memory implementation is completely unintuitive. Unless Elecraft completely redesigns it to work like other radios, go for the Icom or Kenwood if you plan on doing much FM operation.

The disaster that is the K3's memory implementation suggests a radio that was designed by people with little experience of anything besides HF CW and SSB operation - or indeed experience of how the competition did things. Certain other aspects of the K3 tend to support that suggestion.

The K3 could have been the digital mode operator's dream radio. There is built-in support for encoding and decoding both RTTY and PSK31, though you must send using a Morse paddle rather than a computer keyboard, and the K3's display doesn't provide much space to show the decoded text in a scrolling window. In the DATA mode, the K3 implements a "soft" ALC that lets you set a comfortable audio drive level and then keeps the power output constant across almost the entire range of the audio spectrum. Click the cursor on any signal in the waterfall and just call. No more constantly tweaking the PC mixer slider level to control the power level and keep the drive below ALC to maintain a clean signal. It's PSK heaven!

Like most modern transceivers, the K3 offers both transmit and receive equalization: perfect for voice modes. For sound card digital operation, though, you want a flat audio response with no equalization. Other radios that have both a DATA mode and equalization turn the EQ off when the data mode is selected. Even radios that cost half the price of a K3 can do this. But not the K3. So if you're a keen digimode operator and don't want to have to manually change the EQ settings whenever you switch from phone to data, you'll just have to leave the equalization flat. Elecraft knows this is an annoyance, but apparently it is difficult to change due to the way the firmware has been written. So we're still waiting.

Which brings me back to the point that more than two years after its announcement, and more than 18 months after I first received it, I'm still waiting for some aspects of the K3 to work as expected. Believe it or not there are still advertised features such as synchronous AM detection that have yet to be implemented. I purchased the high stability TCXO option and there is a facility, described in the manual, for you to enter in temperature calibration data to increase its stability. Except it does not exist. It's still to be implemented.

In a nutshell, my verdict on the K3 is: great hardware, shame about the firmware. Two years since its first announcement, it's still a work in progress.

PS: You can now read a 'lite' version of this review at eHam.net.

Thursday, September 17, 2009

Just say no to Yahoo!

If you want to discuss a specific aspect or product related to ham radio, the chances are there is a Yahoo! group about it. In fact, for certain subjects there are likely to be several, due to some people's habit of starting a group before looking to see if one already existed for that topic.

Why are Yahoo! groups so popular with the ham fraternity? I hate them! First of all, I hate Yahoo! as a company, for reasons that I won't go into. Second of all, Yahoo! groups are essentially mailing lists, a technology from the early days of the internet before search engines and forums were invented. When my mail client shows I have new messages, I tend to interrupt what I'm doing to see who has replied to me. I don't want to be disturbed by someone's contribution to a thread I haven't participated in and may not even be interested in. Consequently I read Yahoo! groups on the web. And compared to 21st century forum software, Google Groups or the free forum platform Nabble, Yahoo!'s web interface is slow, cumbersome and just plain sucks.

The third, and really major reason why I deplore the use of Yahoo! for ham radio discussion groups is that they tend to be moderated by despotic paranoid individuals who make them private and then apply their own idiosyncratic rules as to what is or is not admissable. The problem with making them private is you have to join to see what is being said, because none of the content can be found using Google (or Yahoo! Search, for that matter) which would add to the openly available knowledge on that subject. Personally I like to see what the tone of a forum is before subscribing. But often I just have a casual interest in what is being said about a particular transceiver or whatever and have no desire to join and participate.

It wouldn't be so bad if joining was simply a matter of clicking a few buttons and you're in. But the despotic paranoids who run some of these groups make you jump through hoops before they'll even accept you. You have to wait until a moderator reads your request to join, and if he doesn't think your reasons are convincing enough he won't admit you.

One Yahoo! group - about magnetic loops, if I recall correctly - had some rule to keep out "lurkers" that you were expected to post 4 times in a fortnight after joining. The group didn't like what I posted as I was raising concerns about RF exposure levels from magnetic loops, so I was kicked out. Another group, recently started to discuss an as yet unavailable new transceiver, includes the following condition: "Application for membership waives any "free speech" rights, and implies that the applicant is willing to submit to the rules set by the group moderators and owners." I'm afraid I find that unacceptable.

Can't we make more use of open discussion platforms like Google Groups or Nabble for specialized ham radio forums, and do moderators need to act like Stalin?

No tropo

Well, it seems like I have the answer to the question of whether it is possible to work VHF DX during tropo openings from my location behind the Lake District Hills.


The map above shows the 2m opening that is apparently going on as I type this. What I can hear in my receiver with my beam pointing east is... precisely nothing. :(

I was delighted to be able to work down into Spain and Portugal in June during some big Sporadic-E openings, even though I heard and worked a lot less than people elsewhere in the country. I guess that was possible because Sporadic-E clouds are high enough to reflect signals over the hills to me. Tropospheric ducts occur in the atmosphere at heights of only a couple of thousand metres, so I suppose a clear path is still needed.

I would really like to move back to Essex!

Tuesday, September 15, 2009

Voices from space, whispers on 80

It was one of those crystal clear evenings when the stars shone like jewels from a black sky. N2YO's satellite tracking site showed that the Space Shuttle would pass over at about 9.20pm at magnitude -2.0 - brighter than Jupiter - so my wife and I went out in the garden to watch it. Olga spotted it first, rising over the roof of the neighbour's bungalow.

I had the VX-8E tuned to 145.825 to see what APRS I could catch, but nothing was heard, so I flipped down a channel to 145.800 and heard an astronaut's voice, clear as a bell. We only heard one side of the conversation, but it was still quite something to watch the space station go over and know that the person talking about why he had wanted to become an astronaut was actually in that bright, moving star, up there.

Meanwhile, up in the shack, my FT-817 was doing some WSPR testing on 80m using my Wonder Loop magnetic loop antenna. To get it to tune 80m I connected an 820pF silver mica capacitor across the loop / variable capacitor terminals. The tuning was nice and sharp, and although the SWR did not seem to be particularly low it only raised one bar on the FT-817 meter.

There were not many people on 80m WSPR but I was consistently received by LA2XPA, while I received G7JVN several times. For comparison I replaced the Wonder Loop by the ATX Walkabout and received two more spots from LA2XPA at similar signal strength. I also put out a number of CQs using 5W of PSK31 but got no replies. I wasn't even spotted on the PSK Propagation Reporter site.

I didn't really expect the small Wonder Loop to work very well on this band. However, earlier in the day I had stumbled across a site I hadn't seen before showing a small diameter magnetic loop with plug-in tuning modules covering 160m to 6m, so I was interested to see how far I could extend the Wonder Loop's coverage.

A taxing morning

A few days ago I received a letter from the Inland Revenue (the UK equivalent of the IRS) reminding me that I should pay tax on my bank interest. This caused some hollow laughter, since the amount of interest the banks are paying on savings at the moment is so paltry (0.1% p.a. on some accounts) that the tax on it amounts to almost nothing. However it seems to be a sign of the times that in order to try to fill the massive hole in the public finances caused by falling tax revenues and bailing out the banks, no amount of tax is deemed to be too small to be worth collecting.

A couple of weeks ago I ordered a USB programming cable for the VX-8R from Valley Enterprises in Idaho, via eBay. The cost was $39.95 plus an extremely reasonable $3.96 for USPS First Class International Shipping (total 26.50 British pounds.) The package was posted on 3rd September. This morning, 15th September, a card was pushed through the letterbox to say that a package was available for collection at the post office and there was £11.63 (about $19.25) to pay!

I couldn't think what the package was, because such a large amount suggested that the value of the item was much higher. But when I went down to the Post Office I found that the package was indeed the programming cable. The sticker on the envelope explained that the VAT was, as expected, £3.63. On top of that there was a Royal Mail International handling fee of £8.00!

This is simply outrageous! I have no objection to paying VAT, or even customs duty where it may legally be charged. But then to have to pay a fee for collecting it, a tax on a tax, a fee that is more than twice the value of the tax itself, is ridiculous! Does it really cost more to collect this tax than to physically transport an envelope all the way from the USA to here? This is legalized robbery!

Surely the Revenue and Customs department should pay for the tax collection, not the consumer? After all, when I pay my income tax at the end of the year I don't pay an additional fee for the privilege of having them collect it from me. Unfortunately, other than expressing my indignation in this blog there is not much I can do. Our politicians are only interested in the fat salaries they can get once elected plus the extras they can make on expenses. So who would I complain to?

In total, this programming cable has cost me about £38 (the equivalent of $63.) On top of that I still have to pay ten quid to G4HFQ for the software. The equivalent Yaesu ADMS VX-8 software and cable from a UK dealer would cost £61 plus shipping. So I've still saved a bit, but not as much as I expected.

Once upon a time if you ordered low value items from overseas and they came by post no-one bothered to collect the tax on them. But now that the postal service can make a profit from collecting the tax it seems that no item's value is too small. I shall think twice before making any more purchases from outside the EU.