开云体育

Hi

Hans' comments (see below) are very appropriate to the measurement of Crystal frequencies.? I too found that if you solder the crystals in place, it may be necessary to wait several minutes for things to cool down to where frequencies are stable.? This applies to crystals in the 10 mhz filter as well as those in oscillator positions.? When soldering the crystals in place, it helps to use a heat sink on the leads between the solder joint and the crystal.? Just gripping the lead with a pair of pliers while soldering is usually adequate.? Wrap a rubber-band around the pliers handles to keep them on the lead if you need three hands to hold the pliers, hold the solder, and hold the soldering iron.

If you build the AF section, the IF-to-AF mixer, and the BFO before you build the IF Filter it is possible to use these as part of your test equipment in alignment of the 10 mhz filter.? Build the VFO but temporarily replace the VFO resonant circuit with a spare 10 mhz crystal and a series variable capacitor.? Use this oscillator as the signal source to "test" your IF Filter.? When the "VFO Crystal" and the BFO are tuned you will find a zero-beat if they are both at the same frequency.? Tuning either of these oscillators off-frequency will result in an audio tone that represents the difference in frequency.?

Using a simple diode detector and high impedance voltmeter you should be able to plot the bandpass characteristics of your IF Filter (turn the BFO "OFF" for these measurements).? The ideal bandwidth of the filter would be about 2.1 KHZ wide at the -6 db points and 4.2 khz wide at -60 db.? You may not be able to achieve quite that good a response with the simple ladder filter used in the BITX series, but it is the goal to strive for.? The slope at the sides of your filter should be as steep as possible.?

Once you have a picture of your filter bandwidth you can adjust the BFO frequency to position it at the -20 db point on one side or the other of that pass-band.? Which side of the passband you use is determined by which sideband (USB or LSB) you wish to transmit.? Why put the carrier at the -20 db point...because that helps to attenuate some of the lower audio frequencies in the desired sideband.? These lower audio frequencies are not really necessary for intelligibility, but they do use up a significant amount of power if they are transmitted.? Filtering out the lower audio frequencies ( 0 hz to about 350 hz ) allows more of your transmitted power to represent only the audio spectrum that is necessary for communication.? This same characteristic helps improve the apparent quality of received signals by reducing received audio power that might be devoted to these lower and unnecessary frequencies.

ALTERNATIVE FILTER PLOTTING METHOD:
If you do not have a frequency counter for monitoring the 10 mhz signal source, there is an alternative method for plotting the bandwidth of your filter.? With the BFO "OFF", use the RF voltmeter arrangement and your tunable VFO Crystal oscillator to determine the center of your filter pass band.? Set the 10 mhz VFO crystal oscillator to the center of that bandwidth.? Now turn your BFO "ON" and adjust it for a zero-beat with the VFO crystal frequency.? Now comes the hard part!? Slowly adjust your VFO Crystal frequency to find the estimated 1000 hz tone location on both sides of this zero-beat.? Carefully mark those frequencies on the dial of your VFO Crystal tuning knob.? Now you can turn the BFO "OFF" and proceed to use your "calibrated" signal source (the VFO Crystal oscillator) to make an approximate plot of your IF Filter bandwidth.? Note, that this is only an approximate way to tune your filter, but it will get you close enough for most operations.? The signal level through the filter should decrease very rapidly as you tune beyond the 1 khz points ( +1 khz and -1khz represent a total of 2 khz and should be the approximate edges of your filter bandwidth).

The IF Filter is probably the most confusing part of the BITX design, but it is also the easiest to duplicate if you use exactly the same components that Farhan specified.? Departing from Farhan's design specifications is what seems to cause the most confusion and doubt.? My recommendation would to build at least one BITX20 that is exactly like Farhan's original design.? That exercise provides the necessary experience to support deviating from his design, and it will provide you with a reference transceiver that may be used for comparison type troubleshooting.

73's
Arv
_._

From:? Hans Summers <>
Date:? Wed?Jul?14,?2004? 2:38 am
Subject:? RE: [BITX20] CRYSTAL MEASUREMENTS


?
> Frequency reading is still moving / drifting,
> although circuit has been warming up about 5
> minutes.
> Q: How do you "note/catch" the frequency in this situation ?
?
Your circuit and method look Ok to me... how bad is the drift, and what happens if you wait longer than 5 minutes? The drift could also be in your frequency counter (or PC if you are using that as a frequency counter). The crystal itself will also drift, we tend to think of crystals as rock stable but they aren't at all. Relative to an LC oscillator maybe, but not in absolute terms. Just read about the problems I had getting a stable frequency in my 30m QRSS beacon (?and ).
?
In my BITX20 evaluation, I managed to find a socket which fit the crystals. If you are soldering the crystals in, you might find that the initial drift is rather high due to the crystal being so hot from the soldering iron and needing to cool back down to room temperature. A socket is better for testing. My socket was a proper HC49 socket temporarily "borrowed" out of a different unfinished project. But you can hunt around the junk box and another type of socket which will work Ok. The wires of the crystal are long enough that you can bend them slightly to fit in whatever socket you find to be suitable. Even with a socket, there was some initial drift due to the temperature stabilisation after me holding the crystal in my fingers.
?
73 Hans G0UPL
http://www.HansSummers.com