12525	2016-03-22 10:52:13	Ashhar Farhan	and about the lsb carrier crystal for a ladder filter...
Message	Date	From	Subject
in the last few months, i must have helped about 20 people build LSB ssb rigs (don't ask why). we used a set of five matched crystals (within 100 Hz) at 12 MHz. Invariably, each time, the crystal came up at almost the exact frequency required for the carrier without needing any trimming cap in series. is there an explanation to this? when we build the ladder filter, from all crystals of x frequency, is the passband below the frequency of the crystal just so that the crystal will sit on the upper skirt? i know that there are many variations (like what is the specified loading of the crystal) but these aren't of consequence as only relative frequency of each crystal is of consequence. my understanding was that the mesh capacitors add up more of parallel capacitance in addition to the Cp thus pulling the parallel resonance down and hence notch at the parallel resonance moves down. but is there any such rule derivable from the models that will predict this behaviour? XLAD didnt turn this up, btw. - farhan
12527	2016-03-22 13:39:11	Bill Carver	Re: and about the lsb carrier crystal for a ladder filter...
In the usual "LSB" filter topology the crystals are in series between capacitors to ground. Lets say those capacitors are 100 pF. The resonant frequency of the "mesh" of crystal and coupling capacitors is the filter CENTER frequency. It's raised above the series resonant frequency of the crystal by the effective 50 pF of the two coupling capacitors in series. When you do a design on the computer you can see that as you make the filter wider the bottom edge of the passband doesn't move much.....the top end is moving. If you made a ladder filter with those 100 pF coupling capacitors, and the carrier oscillator environment is 50 pF, then the carrier oscillator position will be the center frequency of the filter. If the carrier oscillator presents more than 50 pF to the crystal, the carrier moves lower toward the low side of the filter passband. Conversely, less oscillator capacitance will move the carrier oscillator higher, toward the high side of the filter passband. We usually try to arrange things so the carrier oscillator frequency is just above the upper edge of the filter, to make a lower sideband generator/detector since the filter attenuation is much steeper on the high side. You would be very lucky if you find that the filter coupling capacitors that produce the desired Bill
12528	2016-03-22 13:50:28	Bill Carver	Re: and about the lsb carrier crystal for a ladder filter...
On 3/22/2016 2:38 PM, Bill Carver wrote truncated his replay to Fahran: We usually try to arrange things so the carrier oscillator frequency is just above the upper edge of the filter, to make a lower sideband generator/detector since the filter attenuation is much steeper on the high side. You would be very lucky if you find that the filter coupling capacitors that produce the desired........... produces....the desired carrier frequency. At 12 MHz you have a wider tuning range so almost certainly even if it does not start on exactly the right frequency, the oscillator capacitance can be changed until it is. Also an interesting philosophical thing to consider: it does not matter one bit whether the crystal is "series resonant", or "18 pF parallel" or "32 pF parallel". They are all the same crystal, just moved a few KHz to produce resonance with the indicated capacitance. They will all make filters and carrier frequencies, the only difference will be a shift of a few KHz in the exact frequency of the passband. As long as all the crystals are from the same surplus batch, or are were all specified in the same way....you should be able to position the carrier with respect to the filter. Now if your filter crystals and your carrier crystal us specified differently than the filter crystals it might not be possible to move it far enough. Of if your filter is at, say, 3.1 MHz you might find it difficult to move the carrier oscillator to the filter edges. In that case you might have to have a crystal made or use an Si570, to get the carrier where it needs to be. Bill