Huff and Puff Stabilisers - they really DO work!

In an earlier post I alluded to the fact that I was building an X-lock by Cumbria Designs with a view to incorporating it into my 30-year-old Yaesu FT-107M as that particular radio has (or had) a bad case of "driftitis".

X-Lock is a software based derivation of the "Huff and Puff" oscillator stabilisation technique whereby the frequency of an oscillator is sampled, compared with a crystal or other high stability reference, sampled again and frequency nudged in the opposite direction to any detected drift.

A couple of sessions with the soldering iron saw the kit completed without any issues, and then followed quite a bit of "circuit-bashing" (study of schematics!) plus reading up of other implementations of X-lock into old equipment.   After all this I had on the back of an envelope a plan for interfacing the two pieces of equipment.

Last weekend I lashed up a test which looked promising, and finally, this weekend the project was completed and I now have an "X-locked FT-107M".

The X-lock fits very snugly into the radio, and all of my interconnections are such that no mechanical or electrical changes have been made, and it could all be removed without a trace, if required.

This will all be written up properly in due course and posted on my web site - here and now isn't the place for detailed explanations, but the following two "Argo" screen grabs illustrate just how effective the unit is on received signals.

They both show reception of the 14MHz beacon IW3ICH and were grabbed withing minutes of each other, one with X-lock controlling the frequency, and the other with just the bog-standard built-in circuitry in control.

The above trace is with X-lock disconnected, in other words how the "vanilla" FT-107M performs even after the radio has been powered for some considerable time.

Now look at the same signal with X-lock in charge :-

I would say that's quite an astonishing result and the screen grabs speak for themselves!

One of my long-term objectives has been to improve the frequency stabilityh performance, if possible of my old radio gear to the point where they could be used without any embarrassment on modern modes such as WSPR, and I would say that the X-locked FT-107M is now "good to go" on WSPR.  I just might try it today ...

Classic Hybrids and WSPR and ...

A very interesting thread has sprung up on the "Fox Tango" Yahoo group prompted by an Italian amateur with an interest like mine in classic hybrid radios - in his case a Yaesu FT-901.

Apparently his radio developed a couple of faults, in particular the failure of the 6V regulator which feeds (inter alia) the DC supply to the VFO circuit.

Not surprisingly the original IC regulator is obsolete, and having nothing else to hand he replaced the faulty one with a bog-standard 7805 unit, and even though the previous 6V supply is now only 5V he claims that the VFO is now "rock solid".

You will realise from my www.g4fui.com web site that old radios are also an interest in mine, and it so happens that over the last couple of days I have been looking at the performance of my own FT-902DM and in particular the VFO stability (or rather the LACK of VFO stability) of this particular set.

I had been musing upon the supply regulation to the VFO as it appears that the real problem is that the VFO drifts as the set warms up, and as it is a bulky, heavy radio it really NEVER warms up properly (not in a reasonable operating period anyway), consequently it NEVER stops drifting.

It drifts so much that you couldn't use it sensibly for a mode such as WSPR.

I had come up with a theory that the original regulator, which claims to be a "low drop-out" device (ie it will work with the input and output voltages relatively close together, really was working just TOO close to the wire with only a theoretical 8V input being regulated to 6V out, such that the thermal changes in the environment were affecting the output volage to a significant extent.

Having a bunch of 7805s kicking around in my junk box, I just might give this a try. OK the output voltage may be significantly lower, but assuming any static frequency change can be trimmed out, the extra volt of dropout voltage, coupled with a more modern device may well make a noticeable difference to the performance.

Another contributor to the same thread suggested putting a diode in the common lead of the 7805 to lift the output voltage, but I would be tempted to try it without as (a) it would reduce the dropout voltage and (b) I suspect that the temperature coefficient of the diode would exacerbate the drift problem. Having said that, it would be very interesting to find out!

Another alternative would be to fit a 7806, and possibly take the input voltage from somewhere else, eg a 12V or similar rail in the radio.

Another item added to my "to do" list, so watch this space!