View Single Post
Old 22nd Oct 2020, 5:35 pm   #45
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
Default Re: Hallicrafters SX-43 2nd I.F. won't peak

The crystal phasing control in its neutral position provides cancellation of the case capacitance of the crystal proper. Yje phasing capacitor, the crystal case capacitance along with a push-pull IFT winding create a bridge. The phasing control thereby creates a null of the floor of the crystal filter response.

It's clearer if you have a sweeper working on the thing and you watch the display as you twiddle the phasing capacitor.

This makes it sound like there is really only one 'correct' position of the phasing control... well, yes, but. In real world use with congested CW signals, the distortion of the passband that off-tuning the phasing control gives can be rather useful. However, you have to remember that this comes at the price of throttling back he skirt selectivity at the other side and throttling back the ultimate skirt selectivity om both sides.

For a bit of scenery...

A full lattice crystal filter section has a second crystal in the place of the phasing capacitor of the half-lattice version. It's trying to get stopband attenuation by cancelling the current in the case capacitance of one crystal with that of the other... So now you know you need to get crystals with matched case capacitances. You also plant the series resonance of one on the parallel resonance of the other.... so you need those controlling to suit your target bandwidth. There's a limit to how fat you can plan this way, so you instead get a copy of 'Zverev' and calculate the pole-zero patterns for your ideal filter. Collect poles in pairs so you can order crystals for your crystal pairs for a series of full-lattice sections, and voila! you have a classically designed multi pole lattice filter.

A lot of work. AND the ordering of a lot of crystals with requirements on parameters only very specialist (read expensive) quartz firms would touch.

So people looked for an alternative, and the idea arose of having a series string of crystals with shunt inductors or capacitors between each pair of rocks. It was found that you could synthesise useful filter designs if you had a bag of identical crystals.
But it was also found that away from resonance, it degraded into a capacitive ladder attenuator with the case capacitances doing the damage. You could have either an upper sideband filter or a lower. What if you wanted symmetry, or better stopband attenuation?

Actually it's easy, and it's where this digression is going.

You could stich an inductor across each crystal, value chosen to resonate with the crystal's case capacitance at the centre frequency.... you now get a good, symmetrical bandpass filter embedded in a broader (LC) bandstop filter. It's very useful.

There is a plan-B where you add inverting transformers down the line to add little phasing capacitors to make a null with the case capacitance. Away from the centre frequency the nulling goes sour, so again you wind up with your bandpass surrounded by a much broader bandstop filter. It's also very useful

Look on them as a modified crystal ladder filter

Look on them as a series of simple crystal filters with phasing capacitors

Both views are true

And they shed a bit of light on what that phasing control does.

So if you buy a more modern fancy crystal filter in a box, or a transceiver with several of the expensive things, what's in the box is a simple derivative of what's in that Hallicrafters/HRO/AR88.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is online now