Varactor tuning portion of this shortwave radio schematic. What dark magic is this?

Martin_H

I was at You-do-it electronics yesterday buying IC sockets, perf board and alligator clips. In the kit section I made an impulse but of a shortwave radio kit. It's well made and mostly describes how it works, but there's a bit of handwaving too. It is described on this person's blog:

https://frrl.wordpress.com/2008/11/15/direct-conversion-receiver-making-friends-with-the-signetics-sa602/

The sa602 is essentially the guts of the radio, and the input tank circuit and audio sections of the schematic below makes sense to me:



But the varactor tuning portion looks like dark magic. It's obviously driving an oscillator which the sa602 uses to demodulate the RF input, but what sort of wave form? Lacking an oscilloscope I don't have any insight into what's happening across pins 6 and 7.

I'm curious if that portion of the circuit could be removed and replaced with a microcontroller generating the input to 6 and 7. That way you could precisely generate the demodulation signal over a much broader range than the current schematic is capable of doing.

Mike Green

Pins 6 & 7 along with C1 and C2 are part of a Colpitts oscillator. C3 is just a DC blocking capacitor that can be mostly ignored. L2 is the tank inductor for the oscillator and the capacitance is made up of C4 and D1 in series ... in parallel with C1 and C2 in series. The capacitance of D1 is set by the tuning voltage.

The NXP datasheet for the SA602A says that the "oscillator" can be configured as a buffer for an "LO" (local oscillator). A microcontroller could serve that function although you want to generate a sine wave or as close to a sine wave as you can get. All of the harmonics in the LO signal mix with the input to the SA602A and the results appear in the audio output.

Publison

Martin,

No tubes?!?

Heater will be disappointed.

Phil Pilgrim (PhiPi)

Although you could output a square wave to pin 6 of the SA602 mixer, you would receive stations not only at the square wave's fundamental frequency, but also at its harmonics. It would be better to control the voltage on the varactor via the Prop's DUTY-mode output and monitor the oscillator frequency on one of the Prop's inputs to close the frequency-control loop.

-Phil

Martin_H

Thanks Mike. It sounds doable, but it also sounds like an oscilloscope would be needed to ensure the filters applied to the square wave produce a respectable sine wave.

@Publison, yeah but no transistors either. They're hidden inside the IC's so they don't count.

Martin_H

@Phil, that's an interesting idea as it eliminates the need for a scope. But aren't you are limited by the range of the oscillator? A microcontroller direct into pin 6 could go from the AM band to the FM band if it could generate the sine wave.

Phil Pilgrim (PhiPi)

Martin_H wrote:

But aren't you are limited by the range of the oscillator?

That's certainly true. And a 1N4002 is pretty limited as a varactor.

-Phil

Heater.

No tubes, I'm disappointed.

But hey, a varactor is a diode with a variable capacitance (True of all diodes but exploited in the varactor). It's capacitance is varied with the applied bias voltage. Which means you can tune a circuit with voltage control.

Brilliant! Can't do with tubes.

Martin_H

Phil Pilgrim (PhiPi) wrote: »

Martin_H wrote:

But aren't you are limited by the range of the oscillator?

That's certainly true. And a 1N4002 is pretty limited as a varactor.

-Phil

The idea of a Propeller voltage controlling the oscillator might still be useful if the range of oscillations was broad enough. Would a better varactor diode increase the capacitance range, and hopefully the range of frequencies?

Phil Pilgrim (PhiPi)

Martin_H wrote:

Would a better varactor diode increase the capacitance range, and hopefully the range of frequencies?

Certainly. But high-valued varactors for the HF bands, like the BB212, are no longer in production. Apparently, manufacturers have concluded that the only markets are for VHF and UHF. The alternative would be a multi-tap inductor, whose taps can be grounded via Prop-switched MOSFETs for band-switching.

Of course, a monolithic DDS chip like the AD9832 is always an option, albeit a rather more expensive one.

-Phil

Martin_H

By the way, the reason I became curious about the possibility of altering this receiver is becuase it is a direct conversion receiver without automatic gain control. This makes it similar to the NASA receiver used by the Radio Jove project:

http://radiojove.gsfc.nasa.gov/telescope/rj_receivers.htm

The reason these attributes are important is that the radio emissions from Jupiter are between 18 MHz up to about 28 MHz, but they're not normal radio signals. They're broadband noise where the changes in signal strength are what you want to hear. A good shortwave radio with AGC, narrow selectivity, and noise filters is actually less useful. But the Radio Jove receiver is $165 which strikes me as overpriced.

Martin_H

I have another question about the schematic above. What is the bandwidth of the signal processed by the sa602 and passed through the LM386?

For example, suppose the radio's oscillator was changed to 89.7 MHz (a local FM station), would the subcarriers pass through? For example would the L-R subcarrier at 19kHz be audible? What if you processed the output with a module to detect additional subcarriers for the RDS at 57 kHz, and alternate audio at 67 and 92 kHz?

I'm curious because Boston is supposed to have at least six SCA stations as subcarriers of various stations, so it would be interesting to try to tune into them.

I skimmed through both the data sheets of the sa602 and LM386 and could not find anything specific. But I am almost certain that the sa602 can do it because I saw a schematic of an FM radio using it for exactly this purpose.

Phil Pilgrim (PhiPi)

The circuit shown will not work very well -- if at all -- for FM demodulation. See this document:

http://physics111.lib.berkeley.edu/Physics111/BSC/Readings/ARRL-2012-Book/Mixers-Modulators_n_Demodulators-37_pages.pdf

Also, for demodulating the information in a subcarrier, you will need an additional SA602 whose local oscillator is tuned to the subcarrier frequency.

-Phil

Martin_H

Thanks, I will read that document on the T tomorrow.

I know an AM receiver can't demodulate the main L+R FM signal, but I thought the subcarriers were in side bands and only AM modulated. So if you were only interested in the subcarriers then it was irrelevant that FM was used for the main signal. Or is the entire 100 kHz wide channel FM modulated with the subcarriers AM modulated on top of that?

BTW I have the subcarrier decoding solved because I picked up an unbuilt Ramsey SCA decoder off eBay for $15. It's input is the unprocessed 100 kHz wide channel after it has been down converted by an FM radio. So it's basically an AM radio that tunes over a 100 kHz wide band. The instructions suggest either hacking an FM radio to get that signal before it is sent to the audio stage, or buy the Ramsey FM broadcast receiver which has that output connected to a jack.

Phil Pilgrim (PhiPi)

Martin_H wrote:

Or is the entire 100 kHz wide channel FM modulated with the subcarriers AM modulated on top of that?

That's my understanding, yes.

-Phil

Martin_H

Thanks. I guess that's why they're called subcarriers and not carriers. Not a major setback, I'll just shelve it for now and try their suggested approach later.