May 2007: Here's a good one - the conversion of an old Yaesu FT690 6-metre multimode rig to 4 metres. This was meant to be a long-term project but as usual I went at it like a bull at a gate, resulting in failure within days. Then I calmed down and started a 'Mk2' version...

Here's the Mk2 transverter board with the 20 MHz oscillator, NE612 balanced mixer and 70MHz bpf. Fed with 50MHz from my sig gen, it worked a treat at this stage.

The transverter board almost finished, and still working! Those unused pads and blobs of solder are left over from an earlier failure. The smaller board is the 200mW PA - it used a VN67 fet, but it was plagued with instabilities.  So was the next version. I suspect the VN67 is not a good choice for this job, but I'm too thick to realise it.

The final PA, using a good old 2N3866 bipolar. In fact, this amplifier was designed as a wideband power amp (you can see the little ferrite transformers on the input and output) with a 75MHz lpf on the output.  The finished board has the antenna change-over relay on the end.

A view of the transverter installed in the battery tray, with the PA in the compartment at the top.

Starting at the top right, we have: RX converter (2 dual gate fets and 3 70MHz tuned circuits), bottom right is the screening box containg the 20MHz osc, mixer and 70MHz bandpass filter. Over on the left is the tuned fet amplifier and just above it (the little black dot) is a MMIC amplifier. This delivers about 10mW to the PA.

Spectrum analyzer plot from 50 to 150MHz. The 70MHz output is at 0dBm and the second harmonic is -68dB.

Nothing else is visible.

Just to make sure I ran another scan, this time looking at the closer-in spectrum of  60 to 80MHz. There are a couple of small spikes, but the one on the left turned out to be an internal of the analyzer, and the one on the right is 70dB down.

Good enough for me!

Thoughts: One of the problems has been the sheer amount of power gain necessary to raise the output of the transmit mixer to a useful level. The NE612 is a very useful device but its maximum output seems to be very low, no matter how much oscillator and RF is applied to its inputs. I've noted that the old 'Meon' transverter used an SO42 mixer - I must look up the specs and see if there is a difference.

The block diagram. The TX input to the transverter is taken just after the dual-gate amplifier/AM modulator in the 690.  On receive, the down-converted output from the transverter goes directly to the 690's receiver input.

23 July 2007: I got a very interesting 'phone call today. It was Enrico (I've forgotten your callsign!) from Italy, who has read this story and was wondering if there was a way to directly convert the 690 to 4m - that is without transverting. I'm sure someone with a bigger brain, better eyes, greater knowledge and steadier hands than me could tackle it, but it would be one hell of a job!
First of all, the synthesizer would have to be persuaded to go to a new set of frequencies, although this might just boil down to a crystal change and a coil replacement or two. Once the new frequencies were coming out of the PLL block, the transmit strip would need a retune, and this wouldn't be a simple 'twizzing' job. All those little cans are tuned to 50MHz and would need turns removing from them to reach 70MHz.  The transistors may or may not work at 70MHz either. Of course, the RX section would need similar work.
The transverter method leaves all the original circuitry intact (and much of it unused!) and is by far the easier of the two options.
Anyway, thanks for taking the trouble to contact me, Enrico - our short chat was very pleasurable and the surprise of receiving a call from Italy made some of the prawns fall off my sandwich, but the cat dealt with those with great efficiency! Please drop your callsign into my guestbook when you can spare a minute, Enrico.