[Elecraft] New Crystals or Not
John, KI6WX
[email protected]
Wed Oct 30 20:42:01 2002
Steve;
I worked with Elecraft on measuring the crystals, both the original ones and
the new ones. The original test procedure for selecting crystal sets did
not always select ones with the same series resonant frequency. A second
problem was that the motional inductance of the original crystals varied
more than desired. Both of these effects can create suboptimal response of
the SSB and CW filters. Many K2's are fine, but in some the variation
became great enough that the filter response is not as good as it should be.
In some cases this could easily be heard by listening to the radio.
The best way to determine if you would benefit from replacing your crystals
is to measure the filter response using Spectrogram on the audio output and
an RF noise generator at the RF input. Details on using the earlier
versions of Spectrogram can be found at my web site,
http://home.pacbell.net/johngreb/KI6WX.html . I will publish a note on
using Spectrogram 7 in the near future, but you can still use it for filter
response measurements since it is similar to the earlier versions.
SSB Filter - The correct filter response should have a width of about 2.0 to
2.2 kHz. There will be some ripple in the passband and a more prominent dip
of 3-4 dB in the middle of the filter passband. The response of the SSB
filter can be seen in "Improving the Performance of the KSB2" which is
posted on my web page. If the filter response is narrower than 2.0 kHz,
replacing the SSB crystals will widen its bandwidth. If the ripple is
larger or one side of the passband seems to fall off slowly, the crystals
should be replaced.
CW Filter - For bandwidths less than 700 Hz, the filter should show a single
peak. As the filter bandwidth is narrowed, this peak will narrow and the
amplitude may drop slightly, particularly at a 100 Hz bandwidth. The CW
crystals should be replaced if the filter response shows multiple peaks or
the amplitude drops off sharply with decreasing bandwidth. This includes
all 5 crystals in the CW filter and the 2 crystals in the second Xtal
filter.
Your current set of crystals is fine if you don't see these types of
problems in Spectrogram.
-John
KI6WX
> OK. So I'm somewhat confused. Some with "early" crystals are AOK, while
> others have K2's with apparent problems.
>
> Is there a definitive, non-subjective test that can be conducted to verify
> the performance of the crystal filter to determine if replacement is
> warranted? I'm looking for something more concrete than reports of "poor
> transmit audio" or other qualitative judgement...
>
> Ideas?
>
> 73,
> Steve
> aa8af