[Elecraft] NOT de-Q ?

[Steven Gibbs]

A couple of years ago, when I received my first K2, (#01595) I was
interested to see the use of PLL reference & BF oscillator crystals
pairs in a parallel circuit.  The handbook (p.106) explains that
the crystals "de-Q each other to increase the VCXO tunong range".

As a matter of interest, recently when I built another K2 (#02723)
I initially installed just one crystal of each pair.  The 'PLL
Reference Oscillator Range Test' (p.57) gave
Ref. High Freq: 12100.58
Ref. Low Freq:  12089.41
Range (kHz):       11.17  (just outside spec low limit of 11.5kHz)
The BFO Test (p.59) gave
BFO High Freq:   4916.40  (OK)
BFO Low Freq:    4914.28  (too high, must be <=4912.7kHz)
Range (kHz):        2.12  (not enough, should be >=3.6kHz)

The BFO was obviously unsatisfactory, since it would only be good
for CW Rev. or USB, so I quickly installed the other crystal.

Despite the range being just a little low, the PLL Ref Osc would
have been OK, since it completed the VFO Linearization Test
without complaint.  Nevertheless, I installed the other crystal.

My question is this:  what is the trade-up when crystals are NOT
de-Qed in parallel?  Better temperature characteristics?  Or better
stability?  Less phase noise inherent in the crystal oscillator
itself (I'm not talking about PLL phase noise) - and would any
benefit in less crystal phase noise still be masked by PLL phase
noise?  Should I have continued to operate with just one PLL
reference crystal?

I'm NOT suggesting that you try removing crystals from your K2!

73, Steve GU3MBS, K2 #01595 #02723