February 2014: Here's the latest 4 metre conversion. Well, I've got to keep busy somehow, and doing the hoovering isn't up my street. It's an early 80's Icom 551 six metre transceiver from the same series as the 251 and 451. I picked it up for a good price and was surprised to find it was in almost pristine nick. Very smart!
The rig uses a single conversion architecture to a 9.0115 MHz IF. SSB and AM demodulation takes place at this IF, but FM signals are taken off just after the first filter (about 20 KHz wide) to the optional FM board....which I discovered is not present in this one. Aaargh.
The conversion follows the same path as my previous conversion job on a little FT690, so it's basically a transverter built into the rig. On TX, the low level TX signal is removed from the pre-driver stage and mixed with 20 MHz from a crystal oscillator. The output contains the wanted signal at 70 MHz and an unwanted one at 30 MHz, which can easily be sorted out with a bandpass filter. After this it goes into an amplifier to bring the level up to a usable few mW and another filter. The signal is now ready for final amplification.
The RX part of the main conversion is basically the reverse of the TX. The 70 MHz incoming signals are amplified, filtered and mixed with the 20 MHz from the xtal oscillator to produce 50 MHz and the rig deals with them in the normal way. This bit works too.
What about FM? Well, I have the circuit for the add-on board but some of the components and techniques are even more out of date than I am, so I'm using 20 year-old stuff instead of 30 year old! The RX part of Icom's circuit board contains a down-mixer (to 455 KHz), a crystal oscillator, a 455 KHz ceramic filter, a string of transistor IF amplifiers, a chip limiter amplifier, one of those Japanese ceramic resonator FM discriminators and a bunch of op-amps for driving the S-meter, centre zero meter and the muting.
I'm at a failrly advanced stage in the building of my own version of this, using an NE602 mixer and a good ol' TBA120 IF amp and quad detector (quiet at the back please). This chip has been obsolete since the Crimean war, but it works well and I still have a couple in the drawer. A couple of TL082 dual op-amps drive the meter etc. For TX the story is simpler - the FM is generated directly at 9.0115 MHz with a varicapped crystal oscillator, outputting around 1mW into the existing TX IF strip. There's a bit of diode signal routing to sort out.
Mid March has brought some sunshine and a nice Mitsubishi PA block from GH Engineering. I only ordered it on Monday PM and it arrived today (Wednesday) which is pretty good service by anyone's standards, especially as I ordered with a cheque!
Below is the finished and tested PA block, which breezes along at 10 watts. I'll be attaching this inside the rear of the rig with some fresh air blown in by a small fan. The PA module is obvious, and you can also see the bias pot, the low-pass filter and the antenna changeover relay. The IC551 uses pin diode switching, but it gave me the willies so out it came!
No crystals for the FM board yet.
Bit by bit it's taking shape. Below is the PA assembly mounted in the rear of the rig, up against the ventilation slots. There'll be a nice fan on the outside soon These PA modules are not particularly efficient and draw more current than you expect.
More progress. I couldn't resist lashing the various bits together and applying some volts. After a mini-panic (signal cables taken to wrong bits) it works. So far I have 20 watts of SSB - it will do more, but my 4 amp bench PSU collapses. The RX side of things works well too, so it's coming together nicely. No FM as the xtals are not here yet.
The fan is now mounted and the control circuit is a simple timer that starts the fan about 20 seconds after the PTT is pressed and lets it run for about a minute after transmission stops.
While I'm waiting for the crystals to arrive, I've been attending to a few other small jobs. The minimum tuning step on SSB is 100 Hz which is a little bit coarse by modern standards, so I decided to modify the RIT circuit to make it operate on both TX and RX. Unlike the RIT on my previous rig (the Trio/Kenwood 780) the control circuitry is simple and easily understood, so it was easy to see what needed to be done to achieve the desired result. Just one resistor and the job was done, so now the 551 has a fine-tune control that works on TX and RX and is on all the time.
25 March 2014. The xtals arrived today and of course I couldn't wait to fit them. The down-conversion one (9.4665 MHz to heterodyne the incoming 9.0115 MHz to 455KHz) went in and came up on frequency nicely. The one I thought I may have problems with was also well-behaved. This is the one that has to be pulled all over the place by the microphone signal, but a quick tweak of the series inductor and a small offset of the standing DC bias had it modulating merrily in about 10 minutes.
After that it was all systems go for a final installation and check. There was a buzz behind the modulation, but this turned out to be a poor earth connection on the PLL box on the main board. Below is a view of the almost finished rig awaiting a final tidy-up.
And here is the output spectrum at just under 10 watts out. My 40dB sampler is falling to bits and is reading slightly wrong, but the relative figures are correct. Marker 1 shows the second harmonic at -60dBc and marker 2 is on the 3rd at -70dBc. There are a few small spurs close in, but they are quiet enough to ignore.
I've had a couple of local FM contacts on this rig with good reports. One local was kind enough to tune around my signal and reported that he could detect nothing untoward. Relief!
I'll run the rig on the bench for a while and then find somewhere for it to live. Another job done (not very) cheap!