August 15, 2009
KR1S Synchrodyne MW Receiver, Part 2
What's Not To Like?
The homodyne/synchrodyne circuit has possibilities. Carrier injection can reduce fading effects, and selectivity is less dependent on circuit Q (I believe) -- as long as the oscillator is running. The standard late-1940s homodyne/synchrodyne receivers used oscillators that ran continuously. This block diagram from the Radiotron Designer's Handbook illustrates the common arrangement.
More information on the history and theory of these receivers is posted on The Valve Page. The Webmaster didn't include the author's information, but I believe it was D.G. Tucker, mentioned in the Radiotron Designer's Handbook article.
Let's review the claimed advantages of the homodyne/synchrodyne:
- Carrier reinforcement reduces fading effects
- Less-critical tuned-circuit Q (compared with straight regenerative circuit)
- High-fidelity audio
- Injection-locked oscillator (less chance for drift)
In the superheterodyne and regenerative receiver, a drifting oscillator required listeners to frequently "touch up" the tuning. This is much less a problem with solid-state receivers. Obtaining high Q is relatively easy with regenerative receivers. High-fidelity audio is less important in these days of talk-radio AM broadcasting. The greatest advantage of this design, then, is carrier reinforcement as an antidote to signal fading.
Because selectivity is not obtained through tuned-circuit Q, and because the oscillator is running at the signal frequency, adjacent-channel stations cause high-frequency heterodynes, which are reduced by an RC low-pass filter. Compare this with the selectivity achievable from a regenerative receiver, where it's possible to clip the sidebands. Also, for carrier-modulated double-sideband (aka "AM") reception, the regnerative detector is adjusted just below the point of oscillation. Thus it does not heterodyne with adjacent-channel carriers.
The homodyne/synchrodyne idea offers some possibilities to the DX listener and those interested in exploring divergent receiver designs. Short's design falls into the latter category, but is not, in my opinion, a worthy DX receiver. Its simplicity is a serious handicap, in that the oscillator inductor also serves as the receiver's antenna. My condo is an effective rf shield; outdoors, the reception is much improved, but I was unable to couple an external antenna without detuning the receiver. An inductively coupled antenna like those sold for use with portable receivers would work, but I'd rather have the option of connecting an antenna directly.
A second disadvantage of this design is that the oscillator inductor cannot be shielded. The radiated field is not extremely strong, but it exists. This is not good practice. Further, the oscillator doubles as the amplifier; notice in the Radiotron Designer's Handbook block diagram the use of separate amplifiers and oscillator. An oscillator running at a constant level, phase locked to an adjustable amplifier, can provide a constantly optimal signal level to the demodulator/mixer stage.
I don't mean to diminish Short's design. It's a nifty alternative to the traditional array of hobbyist receiver projects. It works, but it has too many limitations to be more than a novelty. Adjusting gain also adjusts oscillator level. Perhaps in Short's area there weren't as many rock-crushing signals as I have near me. The oscillator certainly locks onto the carrier, but strong stations dominate several channels on either side of them. It would be helpful to have the oscillator isolated from the antenna, so it could be better shielded. Short hints at the problem when he urges readers to closely follow his receiver's physical layout. That's a sign that he had stability problems. Being able to isolate the oscillator from the amplifiers would have gone a long way toward solving them.
Toward A Better Design
If I ever take on the synchrodyne (the proper term for a receiver of this type in which the oscillator is phase-locked to the carrier) again, and I might, I will look more closely at the Valve Page article referenced above. An untuned rf stage up front is essential. Now the oscillator and subsequent amplifiers are electrically isolated from the antenna. Diode-ring mixers are easy to make or buy, and that's what I'd use for the demodulator/mixer. This still leaves the problem of adjacent-channel heterodynes. Fortunately, low-pass active filters are also easy to make, and would likely do a better job than the passive RC networks used in the past.
Elsewhere on this site I've mentioned the desirability of using low-impedance tank circuits in regenerative receivers, for smoother operation. This is hard to accomplish on medium-wave frequencies. To obtain a tank-circuit impedance of 100 ohms at 540 kHz requires a capacitance of 2947 pF! As the Q of s synchrodyne oscillator tank is irrelevant and all the amplfiers are broadband, any LC combination that provides the desired tuning range will work. This may be the circuit's greatest advantage for medium-wave listening.
Redesigning Short's receiver to use an inexpensive, available transistor was an interesting project. It grieves me to see builders slavishly follow an obsolete design. I'm sure this modified version works as well as Short's. I just think Short's design falls short on performance compared with a regenerative receiver, though the synchrodyne concept has merit and should be further investigated using modern devices. This block diagram shows what I'm contemplating:
There is no need to run the oscillator at a high level, as in Short's design. In fact, running it at low level makes it easier to injection-lock the carrier. With 120-dB of limiter gain the oscillator should be able to lock on even weak signals. The oscillator is amplified to drive a diode doubly-balanced mixer, but there is probably no reason you couldn't use an active mixer like the SA612. That would eliminate the need for the 20-dB signal amplifier, which might not be required with the diode mixer, either.