[Hallicrafters] SX-100 stability

[Dave Moorman]

Very informative, John.  Thanks.

Dave

on 2/17/02 3:04 PM, johnf at johnf at kc.rr.com wrote:

> A few thoughts:
> 
> The N150 designation indicates the amount of negative chg in capacity.  It
> is expressed in negative parts per million per positive degree C change.
> Inductor drift is positive with temp....Formulas?  Never did find any.
> 
> Check the back of the ARRL handbook for a listing of N-drift color codes.
> 
> The dogbone temp caps used by Hallicrafters are notorious for being the
> source of trouble & drifting after 40+yrs.
> 
> Chging to all NPO zero drift wud not stop the drift as it wud not offset the
> inductor drift.
> 
> The N-drift caps need to follow the temp chg of the inductors.  If the coil
> form is ceramic I have sometimes found it necessary to put the N-cap in
> physical contact with the ceramic form by using clear heat shrink tubing
> over the cap & form.  Just hanging the N-cap out in the  breeze does not
> always have a good result.
> 
> Invest in a can of freeze mist to assist you in locating which
> parts have the greatest effect on freq vs temp.  A few squirts can save a
> lot of time.
> 
> Dipped silver micas have a small neg drift characteristic which can be used
> to advantage.  When you need a bit more N drift, chg a NPO disc to silver
> mica.
> 
> The original values are a good place to start in the effort.  However, you
> likely will not find the same cap values & N-drift readily available.  Free
> running oscillators are the distant past.  Direct synthesis is the here and
> now.
> 
> Hence, you must use what you can find and compute the effective Neg drift of
> the NPO + Neg caps in parallel and look at it as a "block" of Neg drift.
> 
> I have re-comp'd Collins 70K-2 PTOs, Hallicrafter VFOs, Heath LMOs, etc.  It
> takes time to sort out things and a good counter to measure the result of
> any change.  Slug tuned Inductors tend to drift higher in freq w/age.  The
> slugs tend to lose permeability w/age.  So, sometimes you must increase
> total C to put the osc back on freq.
> 
> 
> Cheers!
> 
> John
> K0LFA
>