Fixing the Orion II warmup drift

My Orion II Fan Mod for Improved Warm-Up Frequency Stability

 
There have been reports about the warm-up frequency stability of the Orion.  Being the type who likes to know exactly “what is the frequency, Kenneth” (remember the 1986 Dan Rather knife-point incident?), I was a bit concerned about this.  When my O2 arrived, I placed a temperature probe in the A10 board compartment near the 44.5 MHz Siward TXO32 master TCXO and tracked the heat and drift from start-up.  I noted that it does get quite warm inside, reaching about 115 degrees F in a 70 degree F shack.  The frequency drift seemed to track the temperature rise over much of the time until full stabilization was reached, which took over two hours.  This is consistent with the results in Martin, AA6E’s original blog. 
 
It occurred to me that drift could be mitigated simply by adding a small CPU-type fan in the enclosure, set to blow cool air on the TCXO.  I started with a 40 mm 12 VDC brushless fan hooked to the bulkhead with wire-ties positioned over the TCXO module.  The results were striking.  The maximum 20 Hz drift on 15 MHz WWV was reduced to 8 Hz.  The only problem was that the acoustic noise was louder than I was willing to tolerate for the long term.  The second attempt used a 50 mm fan, carefully attached with thick ‘double sticky tape’ applied between the edge of the fan and the bulkhead, and an addition piece of standard foam weather-strip (one-side sticky) backed up with more double sticky tape on the stationary flat side of the motor, attached to one of the larger modules near the TCXO.  Much better!  The noise was nearly eliminated, and the increased airflow brought the unit to stabilization in a much shorter time.  With the fan mod, after a 15-minute warm up period, the radio is no more than 3 Hz off 15 MHz WWV, and after 30 minutes it is always with in +- .5 Hz, an astounding .03 ppm after stabilization.  Of course, if the shack changes temperature, the unit is likely to drift a bit, and there will be long-term crystal drift requiring that the alignment procedure be performed every six months or so, especially in the first couple of years.  Power for the fan is obtained by snaking a wire out the back panel through the #2 Band Data connector opening and using the 13.8 V RCA connection on the back panel which is protected by a 2-amp, self-resetting, thermal circuit breaker.
 
The pictures below should help in demonstrating how the fix was done: