When out contesting with G3PYE/P, we often use Gavin, M1BXF's Icom IC-910 transceiver. A fine rig to be sure, but it doesn't have an internal voice keyer (or digital voice recorder) - a feature still lacking in the brand new IC-9100, for reasons best known to Icom.

Instead, we use SM3WMV's PC-based Voice Keyer software. This plays a pre-recorded sound file when you press one of the function keys, and puts the radio into transmit using the RTS line of a serial port.

So far, we've used this and a home-brew data interface to feed the audio into the data port of the rig, but there's a problem - this doesn't go through the rig's compressor, so average transmit power out is rather lower than using the mic.

To improve on this, we decided to build a small interface which accepts two audio sources (PC and microphone), two PTT lines, and routes the appropriate source to the radio when its PTT is activated. This avoids the mic being live when the voice keyer is transmitting, and prevents PC sounds being transmitted when using the microphone and foot switch.

The circuit is simple - a relay for audio switching, and a few transistors to drive the relay and pull the output PTT line low when the PC PTT is activated. A potentiometer on the front allows control of the PC audio level.

Here's the breadboard for an initial test:

And mounted in its box:

Under test with an IC-7000:

And the finished project!

This tuesday is the 23cm UK Activity Contest, where it will get its first on-air test. I hope to work some of you then!

Finally - another item off my TODO list!

My 12v bench power supply is a Watson W-30AM. It's a decent enough PSU, and competitively priced. However, they appear to have a design fault - the fan speed controllers fail... in the "fan off" state.

Mine did so a couple of years ago at a special event station operation, and consequently got rather warm, rather quickly. We performed some quick field surgery and hard wired the fan to +V, which took care of the problem, albeit rather noisily.

This evening I finally made the repair: the LM358 op-amp on the speed control board had let its magic blue smoke out (quite obviously). A quick replacement, along with the rather scorched looking 390 ohm resistor adjacent to it, and it's behaving as it should again. Total cost about £0.50.

I mentioned earlier that I thought it was a design fault: of the four people I'm aware of who have these supplies, three have independently failed in exactly this way...

(As an aside, I'd never buy another linear power supply. They weigh far too much, and leave nice dents in my car's metalwork when you drop them. Decent switch-mode power supplies are smaller, lighter, more efficient, and work every bit as well.)

Although the majority of my radio operation is out portable, I occasionally play a bit from home as well - both on HF and VHF. Living in a semi-detached house in a residential area, with a back garden just over 40' long brings inevitable compromises for HF antennas...

Initially, I ran a 1/4wave vertical for 40m using a fibreglass pole planted in the flower bed, and a good number of radials dug into the lawn at regular intervals. Nice and simple, but very noisy on receive, and not terribly good on transmit either - most of the radiation went straight into the surrounding houses. Including a neighbour's hi-fi system - not good.

The next attempt was an off-centre-fed dipole for 40m, fed about 1/3 of the way along its length. The feedpoint was supported from the alloy pole that holds my VHF colinear, on the side wall of the house. This fitted very neatly: the short end hooked over the gutter at the front of the house, and the long end dropped to the rear fence. The OCFD gave me greater flexibility in terms of bands (with an ATU) - effectively all of 40-6m were somewhat usable. Since putting it up, the neighbour also reported no further issues with his hi-fi. However, the noise was unceasing (S9), and transmit performance remained poor. Time for something better...

On a whim, I tried making a balanced, horizontal version of the 7.6m "Rybakov" vertical - 7.6m either side and fed with a 4:1 balun. The ATU matched it (but then, the LDG Z-11 will match pretty much anything), but TX performance was well down - probably 20dB. Modelling in EZNEC showed a feedpoint impedance that suggested almost all of the 100w was likely being dissipated either in the tuner or the balun!

Suffice to say, that didn't stay in the air for very long, and was quickly replaced by a hastily constructed half-size G5RV. The G5RV is something of a politcally charged antenna - there are those who swear by them, and those who claim they're little better than a leaky dummy load. My view is roughly as follows:

• It's a compromise antenna - just like pretty much all multi-band antennas are. It's never going to beat a 3-ele SteppIR, and you shouldn't expect it to.
• I suspect many people install them very badly, and they perform accordingly: ladder line cable tied to metal poles or coiled up on the ground, and long runs of poor coax before the ATU leading to horrible losses due to the high SWR on the coax.
• For its size, and convenience, it's a pretty decent compromise, if you can get it at a reasonable height, installed properly, and with the shortest run of coax before the ATU.

Given these, I'd done a reasonable job - the ladder line was pulled away from the metal pole, and there was only a couple of feet of coax before it hit the ATU. Performance was pretty good - the noise now at S5, and at least some of the RF leaving the radio seemed to be making it into the ionosphere.

However, the ladder line as it was looked messy from the road, and any good solution I could think of for fixing it more permanently would involve either lots of visual clutter, or introduce significant effort to any maintenance of the antenna. The ATU also seemed a little unhappy on some bands (particularly 30m), and there were some effects I can't explain.

So today, that's come down after only a couple of weeks in the air, and has been replaced by a pair of nested dipoles for 40 and 20m with a common feeder, the former also doing a reasonable job on 15m. The 40m dipole runs between the back fence and a convenient post at the front of the drive, and the 20m is tied off to the gutter at the front and back of the house.

After a previous bad experience with a 10m and 20m pair of dipoles where there was significant interaction between the two when trimming them, this installation was much easier, and they present a good match on both bands.

Noise seems reasonable (S5 or thereabouts); I've not yet tried firing it up on transmit. No doubt in a week or two's time, something else will be up there instead!

One of the things that keeps me occupied is acting as QSL Manager for the Camb-Hams callsigns - mostly GS3PYE/P and GM3PYE/P from our Scottish DX'peditions. I've spent the last few evenings catching up on a bit of a backlog of bureau cards, and in doing so, came up with a set of tips for anyone sending cards...

1. Get the size right

Starting off with something that bit us - make sure your QSL cards are the right size.

Historically, this has been described as "postcard size", but there's no standard definition of postcard size. We incorrectly went with 6x4", rather than the correct 5.5x3.5" (139x89mm) for our 2009 and 2010 cards, leading to the following problems:

• The RSGB bureau will longer accept oversized cards after 31st January, 2012
• Many of the return envelopes sent with direct QSL requests aren't large enough to fit your card in, meaning you have to re-write the address on a new envelope
• Even if they are accepted by the bureau, it makes handling throughout the system less efficient, and your cards are likely to get folded or creased

Our newer cards are the correct size, and the RSGB website now gives much clearer information on the correct size.

2. Print your callsign on both sides of the card

Through some thought and software, I can process, reply to and sort about 500 bureau cards in an evening. One of the steps is to sit with a pile of cards in front of me, type the callsign, briefly check the QSO details, press enter, and move on to the next card - normally spending only a couple of seconds on each card.

One of the biggest slow-downs in this process is if I have to search the card the callsign of the station sending it. That may sound obvious, but remember that the side of the card facing me will be that with the QSO details - so if your callsign is only on the other side, it can double the time taken to do this stage.

3. Make sure it's legible

A nicely designed card is a real pleasure to receive, and it's great to see the effort, artistic talent and creativity that goes into many cards.

However, don't get carried away and print your callsign in such an obscure font that I can't actually read it!

4. Combine multiple QSOs on one card...

If you've made a couple of QSOs with the same station, maybe on different bands, try and include as many as possible on one card, rather than sending multiple cards to the same station.

This keeps bureau costs down, and generally avoids wasting resources, so when designing your cards, allow plenty of room for more than one QSO.

Our outgoing cards can handle up to four QSOs per card. If I receive a card from a station, I tend to reply with the details of all the QSOs in our log, as often the card for other QSOs appears shortly after in the pile!

5. ...unless they're for different DX'peditions

This specifically applies to the sort of operations that I manage, but it's worth bearing in mind: each year, we go to a different island, and each year's operation has its own QSL card design. So if you've worked us once in 2009, and again in 2010, that will need two cards.

When I'm replying to cards, the first step is to separate the incoming cards by operation, then work through a pile of cards for one operation. This poses something of a problem when a single incoming card needs two cards in reply - I need to remember which card I've sent, and instead of moving it to the "done" pile, move it to the next "incoming" pile.

This last tip won't apply to all DX stations you send cards to, but it's worth considering.

(I should point out that if you do combine multiple years on one card, I'll still reply to you - it's just a preference on my part as it significantly eases the workflow!)

Finally, here's my lounge floor earlier this evening, with just short of 2000 cards sorted by DXCC entity, before being packed up and sent off to the bureau:

(I'll leave it as an exercise for the reader to guess which single country makes up nearly a quarter of all the cards! Guesses in the comments section below...)

Today, Phil M0DEG and myself entered the CW leg of the RSGB's 80m Affilliated Societies Contest as G3PYE. Today's contest forms part of the AFS Super League that the Camb-Hams are taking part in, encompassing contests from 160m to 70cm.

Neither of us had entered a CW contest in a serious way before - we'd both done a little casual search & pounce in the likes of CQWW CW, but never tried to run, and never actually gone for a competitive entry.

CW is something I've been wanting to improve for some time - I've periodically spent time practising, but life always seems to get in the way before I actually become competent! Phil is somewhat better than me, but it's similarly a relatively new skill for him.

Having set up in plenty of time, and with the (surprising!) absence of any technical difficulties, we started the contest in search & pounce mode to get warmed up. An hour in, we dared to hit the F1 key for the first time - "CQ G3PYE TEST"! Thankfully, no-one replied the first time :-).

Thereafter, we managed a steady run at about 60 QSOs per hour, and between the pair of us, we managed to copy most stations reasonably quickly - but many thanks to everyone who was patient with us if we didn't get it right the first time!

The occasional pile-ups were an interesting challenge, normally resulting in blank looks all round:

"Did you get any of that?"
"Nope."
"Me neither. There's probably a G3 in there though..."
[Sends "G3?"]
"Oh yes. All good." 

We finished the contest with 176 in the log, less a couple of dupes, and no doubt more than a few errors that we'll lose in adjudication, but with smiles on both our faces, and thoroughly exhausted after four hours of intense concentration.

As an aside, this morning also saw some work at the G6UW shack - the rotator cable for the main tower had become caught under a guying lug when the tower was wound down, and snapped when we then wound it up.Unfortunately, this meant we couldn't rotate the beams into the right orientation to be able to tilt the tower over. Gavin, M1BXF leant me his climbing harness, and a quick repair was made at the top of the tower, before heading off for coffee before going our separate ways to prepare for the AFS contest.

As it's just into the new year, I thought a quick post was in order, looking back at 2011, and forward to what's in store for the forthcoming year...

Like Keith, G6NHU, I started 2011 with the goal of making a QSO every day of the year - something to try and get myself more active from home. For the first four months of the year, I managed that, with no days missed.

However, the start of May saw the very successful Camb-Hams DX'pedition to Arran - where as a group we made 9,100 QSOs in a week from the not-very-rare DXCC of Scotland - and I realised that the "project 365" challenge was getting in the way of many other activities I was enjoying, so decided to put it on hold on my return. My hat goes off to Keith for completing his a couple of days ago - very well done indeed!

Another regular feature has been the Tuesday evening RSGB UK Activity Contest series, where I've joined Gavin, M1BXF, Lawrence, M0LCM, Joe, M0ZRN, Mark, M0MJH and several others throughout the year to put G3PYE/P for 48 weeks of the year. We've seen our results steadily increase, ending the year second on 4m, 6th on 2m, and hopefully in the top few on 70cm and 6m as well. 23cm has been a challenging band for us, but with some recent investigations, it looks like we're heading in the right direction there at last.

After a summer of various other contests, Raynet events (Cambridgeshire Raynet's busiest and most successful year for a long time - if not ever), September bought with it a week-long trip to Saint Pierre et Miquelon with Cambridge University Wireless Society (CUWS). Between the six of us, we managed over 15,000 QSOs from our four stations, with some wonderful pileups to go with the great food, company and hospitality. It was a pleasure to meet Rick, VO1SA and Jean-Pierre, FP5CJ, while we were out there, and it was a real trip to remember.

Barely time to unpack, and it was off to the RSGB Convention, where Cambridgeshire was well represented, with myself, Gavin and Dom, M0BLF all giving talks. As well as the entertaining and informative lectures, the social side of the convention was great, and the chance to meet up with some very familiar callsigns face to face most welcome.

Next up, CQ WW SSB with the CUWS (operating as M4A), where the fantastic HF conditions allowed us to anhialate the England Multi-Two record we set ourselves in 2010 - final results pending adjudication, of course.

And so into the winter months, with the return of the AFS Super League, and a great start there for the Camb-Hams: second in the 50MHz leg, and third on 160m. The 2m results aren't yet published, but it's looking hopeful for a good result there as well.

Finally, a few days off over Christmas allowed me to join my Dad, G3VFC and Fred, M0RFD operating as G5MW/P in the RSGB Christmas Cumulatives contest series, which runs over four days from 26th to 29th December. It's a fairly relaxed and cheerful contest, and it was a nice way to round off the year.

I ended 2011 with just over 600 QSOs (80 DXCCs, 90 all-time) in my M0VFC logbook for the year - a small fraction of the number I've made using other callsigns (G3PYE, FP/M0VFC, M4A, G6UW, G5MW and so on), but well up on the 210 logged in 2010.

So what of 2012?

2011 saw a lot going on - a little too much, at times - and there's lots of projects I just couldn't find time to finish. I need to make some more time available for these in the next few months, as well as maybe keeping a few more weekends free to "just play". Having literally zero free days for months on end certainly doesn't lead to boredom, but I do need to find some down-time sometimes.

That said, there's already a lot of exciting stuff planned:

• Three more events in the AFS Super League series
• Continuing participation in the Tuesday evening UKACs, particularly trying to improve our 23cm score
• Our usual entries in the various field days (HF and VHF) and other contests such as CQWW
• A return to the Isle of Mull with the Camb-Hams
• Overseas rallies in The Netherlands and Friedrichshafen
• Plenty more Raynet events
• Achieving DXCC from my home station (100w and very limited antennas)

In September, I also hope to be travelling to Tristan da Cunha (ZD9) with Martin, G3ZAY, where we'll be active for about three weeks. This is still dependent on getting a place on the ship, but it's looking hopeful so far. Assuming we do get there, it should be a real once in a lifetime experience.

I look forward to catching you on the bands!

73 and Happy New Year,
Rob, M0VFC

After yesterday's rather encouraging results, I've made some more progress on emulating the Yaesu FT-8900's front panel in software.

Firstly, I discovered a stupid copy and paste error was the reason behind not being able to drive it directly from a USB/RS232 (TTL levels out) converter, which is why I was using the Arduino last night. After inserting the missing leading byte, it's now much happier, and the demo below is driven directly from the PC.

That said, for resilience, I'm tempted to revert to using a microcontroller as an intermediate in any case - as I'm typing this, I'm rendering the video you'll see below, and the high CPU usage is enough to occasionally lead to a long enough gap between data frames that the radio assumes the front panel has been disconnected and shuts down.

I've also now got a complete map of all the front panel and microphone keys to their respective positions in the data stream, so I can emulate any key press. Interestingly, it looks like both the microphone keys and the two sets of four buttons on the bottom of the front panel (low / V/M / home / scan) are resistor networks connected to ADCs rather than corresponding directly to single bits. On the other hand, the hyper-memory (1 through 6) buttons are purely a digital encoding. The mixture of encoding methods seems a little odd.

For demo purposes (and to sanity check my logic), I quickly put together a primitive copy of the front panel in software, and it works well:

There's plenty more work still to go, but it's looking really encouraging at this stage. Next up is to do the same for the display data (that is, the data going from the body to the head).

73 for now,
Rob, M0VFC

First blog post for a very long time! I've been meaning to resurrect this blog for a while, and hadn't quite managed it. But here goes...

I have a couple of Yaesu FT-8900 radios I use - one in the car, and one that floats around in the shack or out portable. They're a great radio for RAYNET use, with cross-band repeat capability, and 10/6m as well as 2m/70cm.

However, like most non-HF radios, they have only very limited remote control ability - there's no CAT or C-IV control, for example, and this is something I've been wanting to add for a little while. The radio has a detachable front panel, so in theory, it should be able to emulate the front panel in software, allowing anything that can be done from the front panel to be remoted.

After a little work with a logic analyser, I've now managed the first step of this: getting the radio to run from a "fake" front panel, with some amount of control from a PC. In this instance, it's just a matter of changing the volume, but the rest should just be a case of decoding the rest of the protocol.

Here's a quick (and very rough) demo:

Ralf, M0LRB and I spent a few hours this afternoon working on Woop Woop 3's insulation:

We managed to get the first wall complete - definitely the easiest of the four (not to mention the roof!), but good progress nontheless. It was also pleasing to see the generator ran the circular saw without complaint.

The shed assembly is now complete, so it's in our hands now. Phil, M0DEG and I took Flossie, the Cambridge Repeater Group van down to Wickes, and got the first ten sheets of OSB and loft insulation. We should start getting it up over the weekend.

I was hoping to wood stain the outside tomorrow, but the wet weather has most likely delayed that.