[Elecraft] K2 14 crystal replacement results. K2 system gain results

[Lenny Wintfeld]

Hi

I just finished replacing the 14 IF filter crystals in K2 s/n 1520 with 
Elecraft's K2KSB2XTLS kit (7 cw filter crystals and 7 ssb filter 
crystals). The old crystals were marked "4.91-0195" for the cw crystals 
and "4.91-20" for the ssb crystals. I did before and after runs on 
Spectrogram (saving the "before" runs to disk).

After re-doing the CAL FIL setup for the new crystals, I ended up with 
just about NO before to after differences. I was interested to see if 
 the filter loss would be any smaller on the cw filter and if the OP1 
bandwidth would be any wider for the ssb filter. There was little 
measurable difference for either of these parameters after the crystal 
swap.

Spectrogram showed the following results for cw filter loss (with 
respect to the .7 kHz bandwidth)

B/W      Old Xtals     New Xtals

0.7          0 (ref)           0 (ref)
0.4          -3 db              -2 db
0.25        -5 db              -5 db

Measurements were done at 7.04 MHz with a N0SS noise source, preamp off, 
agc off and taken at the peak of the filter response (in my case at 680 
Hz, my desired sidetone frequency). I set the gain once then didn't 
touch it, and made my measurements. Note that I didn't include filter 
loss measurements on, or with respect to my 1.5 KHz cw filter. It's too 
lumpy.

The ssb filter graphs for OP1 look very similar before and after. I 
don't know how to measure the bandwidth because the passband  isn't all 
that flat. I've asked Gary of  Elecraft to look at my spectrogram graph 
and give his guess as to whether the graph looks "typical" based on his 
long experience with lots of K2s. I hope to hear from him in a few days. 
I can say this though: if I draw a straight line up the filter skirts 
and join them with a horizontal line where about half the passband 
amplitude is above the line and half is below, the bandwidth comes to 
about 1900 Hz. I have no idea if this is a valid way to measure.

Finally, to check my cw filter loss measurements, I did a system gain 
check of the whole receiver (only with the "after" crystals). The 
results are as follows:

B/W    Rx Gain
-----     ----------
1.5       109.9 dbv
.7         108.6 dbv
.4         105.9 dbv
.25       103.9 dbv

This squares pretty well with the numbers from spectrogram.

I used an HP 8640B sig gen set to -107 dbm (1uV ) and tuned the receiver 
for peak voltage across the speaker (used a scope across the speaker). 
Multiplied the speaker Vp by .707 and took the voltage ratio. (The 
resolution of the numbers above is greater than the accuracy, but the 
differences should reflect the filter loss pretty well).

Measurements were made with agc and preamp off and the af and rf  gains 
all the way up

While I was set up, I did a system gain measurement, as above but with 
the preamp turned on. The results were as follows:

B/W    Gain
-----     ----------
1.5       125.6 dbv
.7         123.0 dbv
.4         120.3 dbv
.25       117.4 dbv

Which shows filter loss (between .7 and .25, bandwidths) within 1 db of 
the preamp off measurement and it also squares pretty well with the 
spectrogram numbers. BTW the preamp gain is about the number shown on 
the K2 schematic.

Hope this helps those of you who have s/n's near 1520 to decide whether 
to get the 14 xtal upgrade. As always, even with a close s/n, ymmv. BTW, 
the xtal swap-out was not too hard. Took about 1 1/2 hours, not counting 
the CAL FIL afterwards.


73,

Lenny W2BVH