Tuesday, 26 March 2013

The Elecraft K3 as an IF for 472kHz

After a few days playing with my new PICAXE 20M2 project board I thought I'd put it on one side and go back to the 475kHz transverter project to try something I had meant to try previously.

I use an Elecraft K3 as my 630m band IF transceiver.

By putting the Elecraft K3 CONFIG:KXV3 to TEST it is possible to 'bypass' the normal restrictions on K3 coverage ( to some extent). It is then possible to tune the K3 to 10.472 kHz to 10.479kHz and then downconvert to 475kHz +/- by mixing with 10.000MHz. This can be either the shack 10MHz standard or alternatively a suitable crystal oscillator.

I found that by just replacing the 3.2MHz crystal in the original oscillator (G3XBM 472kHz transmit converter) with a good quality parallel mode 10MHz crystal the oscillator and buffer would go directly on 10MHz and could be netted onto 10.000000MHz quite readily!
In due course I will tap off the 10MHz shack standard instead of using the crystal for rock solid 472kHz operation. LO output power seems to be similar to the original arrangement ( to be checked again)

As the K3 produces up to about -12dBm of clean output at 10.475MHz, this is a good match to the transverter SAM1 mixer IF input without requiring further attenuation of the IF transmit signal. And in test mode the K3 power setting still works allowing a 10dB or so adjustment range.

Using my SMG signal generator I checked the K3 sensitivity at 10.475MHz and although less sensitive than at 10.1MHz, it is still capable of better than -120dBm MDS. This should be adequate for use with the 10dB gain of the transverter receive side.

Now to complete the transmit chain.


- Posted using BlogPress from my iPad

Tuesday, 19 March 2013

Further work on the 472kHz antenna

After last night's pleasing results on 472kHz I though I would make some further improvements to the antenna. It was a 7m ( approx) vertical, using the lowest four sections of my MFJ 40/80m vertical. As it is 472kHz, radials are not used as they would hardly be effective, being limited to 10m long.
Instead it had just one ground rod. The soil here is sandy ( the area is known as the Sandlings!) so it is very poor for an antenna ground. On the recommendation of Alan, G3NYK, I have increased the number of ground rods from one to 4 and used thick green/yellow earth wire to connect them to a common earth plate and hence to the antenna mounting ground.

Using my HP8753C network analyser I was able to measure the vertical self capacitance as approximately 90pF. This was a little higher than expected from a 7m vertical, but the capacitance was confirmed by resonating the vertical with a series inductor and measuring its inductance value. The ground loss resistance was far too high at around 200R! It was not possible to determine the real value due to 'noise' in the measurement. With the cursor sitting out on the edge of the Smith chart even small changes, e.g. due to antenna movement or interference, caused the value to change dramatically.

The photo shows the cursor out on the right side of the VNA display.

With an extra section of vertical taking the total height to 8.1m and the extra ground the antenna capacitance has gone up to 114pF and the ground loss resistance now measures approximately 110R. Coming down but more to do......!

For matching I am just using a substantial ferrite rod matching/base loading inductor, with 4 turn overwind to couple the coax into the inductor, in order to tune out the vertical self capacitance. This gives me 1.4:1 (38R) SWR at mid band. I will need to add another 1 turn to the overwind To further improve the match before firing up the transmitter.
The SWR is currently (just) within 2:1 across the band.

Just how critical the adjustment of the ferrite inductor core is came as a big surprise.
Who said LF adjustments were less critical than at microwave!

I will leave the receiver on overnight and see what WSPR stations can be received. Last night's was quite good.


- Posted using BlogPress from my iPad

Sunday, 17 March 2013

475kHz Transverter - an update

At last, today, I was able to put my 23cm and 13cm preamp building on one side and concentrate on my homebrew 475kHz transverter. In the few minutes I have been able to devote to it over the last few weeks I had become frustrated by low receive sensitivity and IF breakthrough from the 80m band. This was mean to be a therapeutic break from microwaves.

I had already decided to use a 3.2MHz conversion crystal, as G3XBM had done in his simple 475kHz transmit converter, for the conversion oscillator to the 3.675kHz IF. Indeed it uses the same circuit configuration as Roger. No Butler oscillators here!

However, whilst Roger's popular design passes the IF through to, e.g. an FT817 on receive, I wanted to use my K3 as the IF since it was already connected to the P3 spectrum display and via the in-built PC sound card to the K3 data ( line in and out) facility. A K3 doesn't normally cover down to 472kHz. This necessitated a complete transmit and receive converter effort as I couldn't find any suitable designs elsewhere.

I decided to use a Norton amplifier RF stage with 2N5109 followed by a five pole Chebi LP RF filter. Together these give a nice, flat topped bandpass characteristic with roll off below 300kHz due to the Amidon 68-2 transformer in the RF stage and above the LPF knee at 530kHz, thus allowing some coverage of the frequency range around 475kHz whilst providing excellent out of band rejection . No filtering is used in front of the rather 'strong' RF stage.

The mixer is a low level (+7) MCL SAM1. This may be changed to a +17mixer in due course.

Miniature relays are used to switch the signal path into and out of the mixer.

The low sensitivity turned out to be due to 'finger trouble'. This problem was also exacerbating the IF breakthrough. It was due to connector termination problems on the PCB groundplane. The ground wasn't properly connected........this happend as a result of a small design change the unwanted effect of which then went unnoticed at first.

With this problem corrected, but the transverter board still outside the screening case, there is still some evidence of IF breakthrough from 80m, as previously described ( elsewhere), but it is now MUCH reduced and the first genuine 630m signals have been heard. I was astounded at the strength of the BIA NDB this evening! A number of CW and digital signals have also been copied.

Most pleasingly, the bottom loaded antenna doesn't appear to be picking up much noise from the shack or house on 475kHz.

Now to do some further development of the transmit path.

73 de Sam

- Posted using BlogPress from my iPad

Monday, 11 March 2013

G8BHH/G4PBP funeral

I have just returned from the funeral of Russ, G4PBP.
Amateur radio was really well represented. I counted at least 20 radio amateurs in attendance and there were close to 200 family and friends overall. A really good turn out to see Russ off in style!
It was a cold, cold, day at the Bushbury Crematorium, Wolverhampton, but the reception back at the church was a fine affair with excellent food and really great chat with friends, old and new. It really was good to see so many faces from the past; Amateurs I had not seen for many years.
It would appear from comments heard that the QSL pin board was a great success and even non-amateurs asked a lot of questions about what it was and how nice and colourful it was.
A decent picture will appear in Radcom. For now, here is a smaller photo, as requested.

Vale Russ
Posted using BlogPress from my iPad

Saturday, 9 March 2013

Elecraft P3 build

Today I bought a P3 kit to go with my Elecraft K3 transceiver.

Up until now I have been using a SDR-IQ and PC with Spectravue to observe signal spectrum and waterfall activity.

However, the 22inch monitor screen was becoming a bit crowded when trying to display spectrum, 'KST, JT65 and VK3UM EME planner all at the same time!

Something had to change. The something was that a decision was made to purchase a P3 and initially view spectrum on its built-in screen. Later, a P3SVGA card will be added together with another large screen monitor dedicated to the P3. The P3SVGA will also enable me to plug in a USB keyboard and operate any of the K3 data modes without needing the use of the PC.
A second advantage of the P3 is that it is better integrated with the K3 than
The SDR-IQ and Spectravue. The ability to see the frequency of both VFO A and B at a glance looks really useful.

The SDR-IQ will be repurposed to be used with my TS2000X and IF output modification at 10.695MHz. When required, it can still be connected to the switched IF output from the P3 in order to measure noise in the continuum mode. Something sadly lacking in the P3 firmware and indispensible for EME cold sky to sun noise measurements.

This picture shows the first stages of assembly of P3 serial number 2555. On the left is the back panel and on the right the front panel with the colour LCD screen and various control buttons and function selection knob.

In all it took me about 90 minutes to assemble the kit. Mostly this was checking everything was supplied, that should be, and then screwing the various ready-assembled cards into the P3 case. There were no surprises and everything fitted perfectly. This was a pleasurable experience and very therapeutic!

As my K3 was built for me ( thanks Dave) I never previously had the pleasure of following the superb Elecraft assembly instruction. I think I can see why these have been favourably likened to Heathkit at its best!

Now that the P3 has been built up it is time to check it out.......

More anon.

73 de Sam

- Posted using BlogPress from my iPad