Friday, 7 February 2014

28MHz oscillator and mixer Proof Of Concept

Due to the odd troubles im having with Wispy, I decided today I would make use of one of my stock of SBL-1 0-500MHz diode ring double balanced mixers, and make a 'Proof Of Concept' mock up of a 28MHz DSB generator/receive mixer. As the SBL-1 is double balanced and entirely self contained, it means unlike Wispy theres no toroids to wind, and no diodes to match, so the isolation and carrier balance should be much better.

After a false start with a Colpitts oscillator design, I went with this circuit designed by the respected Wez Hayward W7ZOI, and blatently pinched from AA3SJ's website

Now this is designed for 32MHz, so I used the various online calculators provided by Amidon etc, and recalculated L1 for 28MHz. This proved to be 19 turns, which just fit on the same specified T37-6 core.

So, I mocked this circuit up, minus at first the L2 tap and the 5dB pad. And of course I couldnt get it to work!

So why wouldnt it work? Its a W7ZOI design, so I'd every confidence in it. Perhaps something was missing? I decided to check the volts on the base of Q1 - nothing? Hmmm, what about the collector - nothing?! Glance towards the PSU - BUGGER! Turn the flippin' power supply back on! Ah, look at that! 28MHz on the frequency counter!

At this point I came back in and made a cuppa, and had a miniroll, and sent an apology to the G-QRP club forum for being such a muppet!

Back at the bench, I had a brief play with trying to pull the crystal, as the output im getting is a tad over the 28.070MHz it should be. No success there. Not to worry, this is a proof of concept, the actual frequency doesnt matter at this time. So I added the 3t tap to the toroid, which now causes the oscillator to be a bit more picky over the setting of the trimmer capacitor! Too far one way and the frequency first jumps down a dozen kHz, then oscillation stops, too far the other way and it starts operating at 5th overtone on 56MHz!

I then added the 5dB pad, and the SBL-1 DBM

It was my intention at this point to feed it a 28MHz signal and see if I could get audio out (i.e. a bare bones DC receiver) but the 2955B has decided to pack-up again! So instead, I connected the spectrum analyser to the SBL-1's RF port, and fed the IF port with the two-tone audio oscillator. The result was what seems to me to be a DSB signal. As my spec. an. doesnt go low enough in spread to show the sidebands, I have to just hope thats the case.

I did however, notice that the analyser easily picked up the oscillator with the probe near the circuit. This is probably why I had so much trouble with Wispy. So the next version of this will have the oscillator screened.

But, the concept was indeed proven. A simple crystal oscillator and an SBL-1 can make a very easy DSB generator for WSPR or PSK use, at least on Tx. I'd like to know how sensitive it is on receive.

One problem though is the pullability of the crystal, which is pretty much none. For a converter this would be fine, but here I need to shift the frequency a bit, so a fundamental crystal is needed. Club sales have 28.070MHz fundamentals, so I will order one and another 14.070MHz crystal, and see if I can get a decent pull from the first, if not I will have to go with a doubler and start at 14MHz.

40m WSPR beaconing over night showed some good spots across the pond. Ive switched to 10m again today, again at 500mW, but propagation seems very poor and ive yet to be spotted, and am only hearing a few European stations so far.

No comments: