Yaesu FT8x7 Transceiver Operation on 2 m (and above) WSPR

As part of our ongoing external phase lock project that is designed to allow use of a variety of amateur transceivers and transverters on VHF/UHF WSPR, I have been collecting information that I hope will help apply the external PLL solution to a variety of equipment.

The Yaesu FT-817, FT-857 and FT-897 are different radios that were clearly designed and marketed for different function, operation and price points within the amateur market. However, they all use a similar RF architecture for generating a SSB signal on the bands they cover. This architecture uses a single 22.625 MHz master oscillator for generating 1^st LO, 2^nd LO and carrier/BFO signals. This commonality makes each of them similarly attractive for 2 m and above WSPR operation which requires good frequency stability and preferably, good frequency accuracy as well.

To complicate matters, there seem to be two versions of the 22.625000 MHz master oscillator, an “old one", and a newer “high stability” version. The old one looks like this:

The Nylon connectors on either end provide both mounting for the oscillator as a module and also connections to whichever radio they are plugged into. The schematic looks like this.

As can be seen, the electrical connections are just ground, power and RF out – these really are just modules.

The newer, high stability version looks like this

This photo was taken by Joe, KJ6QBA from his FT-897D. From what I understand from Doug, WW6D, this newer version of the oscillator, which no longer has an “exposed” crystal and oscillator circuit, has now become production standard on all three Yaesu models. You can see the oscillator module in the left side of the photo right near the back wall of the transceiver.

The older version has been adapted, with some effort for use with an external 10 MHz frequency reference using a now-obsolete TAPR phase lock system. Though it may be easier to adapt for use for 2 m WSPR it is no doubt becoming less common. An article describing that 10 MHz modification is here FT817 External Reference

Since the high stability version now seems more common, it appears to be the one to address for use with our external PLL solution. What we really are trying to achieve is a VCXO rather than just a high stability oscillator, TCXO or otherwise. It turns out that the very things that tend to make a high stability TCXO useful also make them harder to “get to” in order to steer their frequency. They generally use a good crystal but also one that has been temperature compensated and isolated from external physcial and electrical access. We don't mind the high stability and the (probably) good phase noise performance that a good master oscillator produces but we need a way to push the average frequency of the oscillator around to lock it precisely in phase with an external 10 MHz reference as well as enough tuning range to make up for any residual drift that this oscillator has. Yaesu specifies this high stability oscillator as having “.5 ppm accuracy”. It's not clear over what time, temperature and other conditions these specifications apply but in any case they aren't very good compared to the .001 ppm or so that we would like to have for WSPR use on VHF and UHF.

I asked Doug, WW6D to make some measurements of this oscillator in his FT-897D to see if we might be able to coax it into enough QSY to meet our needs. The first and obvious choice to make a VCXO out of this TCXO is just to vary the one line, other than the RF output, that we have easy access to – the Vcc input. Doug has done this and the results look promising. Over a reasonable voltage range for which the oscillator continues to operate, his unit did this:

This plot shows more than 100 Hz of tuning range (at 2 m) for ~3-7 VDC on the Vcc line. Doug has been watching (and adjusting) his FT-897D on 2 m WSPR for some time and the total drift he sees over days and weeks appears to be less than this tuning range. This is with his oscillator module removed from the radio and wrapped in plastic shipping material to isolate it from short term temperature variations. This isolation probably does not radically impact the slow changes over a day or season-season. It should be noted that this is only a sample of one oscillator and there's no guarantee that it will work for all, but at least it's hopeful.

It would seem a rather easy task to insert a linear voltage regulator into the Vcc line and to modulate the regulator's adjustment pin to create a VCXO from this oscillator that can then be locked precisely on 22.625000 by our external PLL solution.

If we can't reliably QSY this high stability oscillator enough to cover it's drift (presuming that we can tune it mechanically to be centered) then we will either have to open it up, requiring unsoldering, or else just build our own 22.625000 VCXO - which will cost us for a decent crystal.

The crystal, and perhaps even much of the oscillator circuit is inside the same metal, 4 pin, SMD circuit seen in both this picture and the first version. It appears that the frequency control adjustment is a pot which implies voltage control of the basic TCXO, wherever it resides. I haven't yet been able to determine the particular components, particularly the 5 pin one. The cover came off pretty easily so it would actually be fairly easy to get a tuning voltage connection - though I think most Yaesu owners would cringe at the thought. We'll see if I can repair it. I'm not too optimistic.

More to come on this as we have it.

Glenn n6gn