Superhets

For my money, a dual conversion superheterodyne is the way to go. True 50 years ago, and holds up nicely today.

Better than direct sampled conversion (to baseband) for HF bands and aggressive RFI environments. This is not just my opinion! If you get an IF where you can slip in a high rejection analog hardware filter at radio frequency, it lowers the stress on the second conversion and or directly on the A/DC, particularly against blocking compression for signals out of the narrow RF filter, thus lowering both intermod generation and AD/C saturation.

The second conversion is often digital reduction to a "processing bandwidth." This can be by discrete LPF followed by simple decimation.

My rig, the Yaesu FTDX10, moves RF to an analog 9 MHz 1st IF. I used to think it then beat that down to a near-baseband analog 2nd IF, for the comm(unication)s passband at least. This is apparently wrong.

The FTDX10 apparently fast samples directly at the crystal filtered 9.005 MHz fc of the 1st conversion to an FPGA. You can do this A/DC economically at 20Msps, 16 bit samples, these days for a near-2X-Nyquist real-sample-only conversion. Or you can undersample/downconvert to obtain higher bits/sample. 16-bit samples at a 10X oversample rate is pretty high resolution. Either way, the widest onboard crystal filter BPF (of 12 kHz PBW) protects the resulting digital SNR (to sampling noise floor) reasonably well. Thereupon, apparently, the digital samples are split into two separate paths: one to the wideband scope DSP probably used directly, the other to decimation serving the narrowband comms DSP. Each path then becomes a segment of an FFT: relatively wide (up to 1 MHz) for the scopes, and relatively narrow (up to the full 12 kHz of analog BW) for comms. The latter comms BW being the widest available, necessary at least for 6m FM. So that second digital downconvert constitutes the second conversion, for a dual conversion superhet, the crystal filter killing off the 1st convert image. In a legacy dual-conversion superhet, the second conversion is simply to an IF where analog detections are made, envelope for AM and beat frequency for SSB/CW. (And discriminator for FM). It was often 455 kHz way back, then up in the low MHz more recently.

As to the analog conversion of the FTDX10, at least one smart guy thinks the FTDX10 is actually a single conversion superhet! (Not Yaesu!) Well, it becomes a matter of interpretation. I go with two conversions, as the digital is not direct to DC baseband.

 

 

Razzle Dazzle

Commercial rigs these days seem to razzle dazzle with graphics, principally a spectrum scope, once the near exclusive domain of Elecraft's great (and still great) K3. My Yaesu FTDX10 does this 3D thing. 2D, 3D - really doesn't matter. What matters is a clear view of the band, now and very recently (say ten seconds ago). The tap-and-jump to a frequency in this rig is convenient. With it, you can descend instantly onto some as-yet-unknown signal, or just poke your way across the band with a selectable scope span! Maybe in other rigs, too. The Yaesu FT950 was a very nice rig that showed up just before the wave of spectrum scope rigs, making it a near steal on the used rig market - you could get one today for $575 with all the original box contents, including a hand mic. That's a kilobuck cheaper than my rig with otherwise about the same performance. Yes, it's that spectrum scope that raises the price to $1K and up for most any premium 100W HF/6m rig on the market. If I was looking to spend no more than a kilobuck, I'd have to say a used market FT991, a current product, is the most bang with almost the same receiver and modes as the FTDX10, and you get 2m/70cm to boot. I've worked a bunch of these and their owners always seem happy with them.

Yaesu recently created a direct sampled "variant" of the FTDX10 in the form of the FT-710. It is not just a slimmed down DX10, in that it abandons the hybrid receiver architecture, and so lacks crystal filters and super-rest RMDR and Blocking performance. But if you're not bothered by really big signals close by (Field Day, ham neighbors), why not?

Back to the DX10, I actually use my rig's oscilloscope display, called by many superfluous razzle dazzle. In CW, you instantly see waveform of a received signal. I like seeing keying rise/fall times and durations. I can do this with the audio turned down to zilch, revealing approximate WPM right off. Very convenient if seeking practice signals or avoiding calling someone because they're 10+ WPM faster than I can handle. I find the amplitude on the o'scope a better indicator of both signal "strength" with respect to the noise floor, as it reveals an approximate SNR right off, but also the S in RST that various folks say is more to the point of that report (than the S-meter, which can be pretty arbitrary). And it discriminates the CWs from the digitals or the RTTYs right off.

The DX10 supports a wide array of digital plug-ins on the rear panel. Other than the PC/Mac USB, which is really useful for all things digital, the two "stacked" USB's are for an ordinary mouse and keyboard. Honestly, I tried out the "mouse thing," and it's almost worthless in increased utility. Maybe the keyboard somehow enables RTTY Send without an interface, but I haven't tried that. And of course there is the HDMI for a big screen reflector of the little screen. Com'on, who really needs that? We're not presenters on cable TV! The little screen is just fine.

Finally, Yaesu has recently issued a firmware update that permits receiving and transmitting "ESSB" or bandwidth-extended SSB. It apparently allows up to 4kHz of voice bandwidth, if this be your cup of tea. I would call this razzle dazzle, but that's just me. A full 3kHz of voice, actually only about 2.9kHz in practical receive due to the scratching (at least) of the lower 100Hz in every sender's transmit signal, "standard" in the earlier firmware, is just great IMHO.

 

 

Missing in Action in the Yaesu FTDX10 (as of this date)

First and foremost is a vector network analyzer for SWR sweep across any covered band. This differs from the SWR at-a-frequency already in the rig. It would be so easy to add to a modern software-defined rig, especially one with a direct digital synthesizer (DDS), which I suspect about all of them have these days. Most everything you need is already in the box. Maybe adds $25 to the retail. I had to cough up $175 for a touch screen VNA to tune my antennas to the measly 3:1 SWR range of the rig's internal ATU. Nobody's rig seems to do this yet. It has to be on the A list of many buyers. And it may be essential for field ops.

Second is an output power selection for 5W and below. QRP. Also on the A lists. Yeah, you can build one with some resistors and a heat sink, but again, most everything's in the box to begin with. You just attenuate the drive signal and bypass the PA. The rig's TUNE function already does the like. And for those who would WSPR, essential. 5W WSPR is downright antisocial. Just drop the new mode in with a couple relays (for safety) to a selector from 1 mW (which is about what my nanoVNA puts out) up to maybe 100 mW in beacon mode, or up to that 5W in QRP mode. This rig's already got TUNE and QSK relays galore, what's a couple more? Again, maybe $25 atop retail.

Third, in a rig with more modes than you can shake a stick at, the clever MPVD outer ring dial should be capable of more comprehensive custom settings than the CS buttom provides. (It either provides one pre-set function, or by holding the button in, access to a menu of functions, that requiring additional screen presses, etc.) It would be quite straightforward, firmware-wise, to allow an entire operating configuration (mode, band, frequency, filter settings, etc. etc.), say based on capturing the current operating state, to be selectable via the MPVD ring. That would be so much more useful. It only takes a "macro" mindset, a la MS Excel or many other apps, to see what you could do with such a feature, especially where time is of essence (contests, DXing, etc.). Bottom line, I actually never use the MPVD dial except as a fast VCO frequency slew. But "multi-poking" the screen in wide-scan is actually easier! Yaesu has missed the boat with their wildly inexhaustive possibilities for the MPVD ring. Com'on, this is supposed to be an SDR hybrid. Mr. Sulu, make it so (please)!

Fourth is just a plain 'ole complaint! (Yeah, I know, No Snivelling!) (But the Complaints Box is Open.) In the MENUs, some parameters can be rolled up or down by holding in a (virtual) button press on the display. This is often OK, but a real pain in other cases. The one I highlight is MAX POWER for the various operating facts (HF, 50M, AM, etc.). Note: this differs from ACTUAL POWER, which is rapidly adjustable in practice by activating the CS (Custom Setting) button and assigning POWER to the MPVD ring (enabling a rapid spin to any power you like, less than the MAX POWER for that mode!) But I am kevetching about MAX POWER in the OPERATION SETTING subset of the FUNC Menus. Holding the button in for MAX POWER just results in a blindingly slow roll up or down. Takes forever to get from 100 W down to 50 W, let alone 5 W, in HF, for example. Why can't software do the obvious? Hold down for a second or so, you step 1 W; hold down beyond that, and the numbers scroll fast, so you can get to 5 W (from 100) in less than 5 seconds. MIss that 10 W setting? Then just get close, then punch it up or down to hit 10 exactly. The radio guys could take a few pointers from the smartphone guys.

Be careful in digital mode. The FTDX10 could pump out 100 W of FT8, but don't do it! The final LDMOS transistors may (will) overheat, then fail. Now you've blown it. Never try to go past the AM design limit in digital, all of which are key down for up to 14 seconds or longer if you're in the old PSK modes. I wouldn't suggest even 30 W in FT8, but I have used that and the amp is still working. Who is trying to out-"weak"-signal whom? Why not just stop at 25 W?

 

Ah, S-Meters

Don't get me started! IARU defines S9 as 50 uV across 50 ohms at the antenna connector, and 6 dBV (3 dB power) per each S unit step down from S9. So what's the S in a rig with not one, not two, but three selectable RF receive gains? Nothing in my rig manual! And at low S, it seems most rigs become very nonlinear (even 1-2 dB in steps between successive low end S units) as to be worthless in forming an RS or RST report. Yes, we all know the dBV of a signal depends on everything from their ERP at you to band conditons to multipath to antenna efficiency and gain. It's all relative to the best signal in the band, I suppose, right then and there. Or is it? Is that guy a 549 or a 599 when he's merely top dog on a particularly QRN-quiet 30m afternoon, but the band is faded out? You're on your own to make it up. This is one reason "5NN" became the default for "I can hear you" in contests or DX. It does seem that nobody in DX, WAS or contesting actually cares about the S in RST so long as the QSO gets logged! S-Meters ...