[time-nuts] new clock
oldmath101 at gmail.com
Tue Jul 22 12:34:41 EDT 2014
Started following the discussions recently and am learning a lot. Found
the temperature sensor thread interesting. Measuring contact
temperature (enclosure, heat sink) is a different problem than air
temperature. For closing temperature control loops the absolute
accuracy is often less important than things like speed of response
(phase lag), both short and long term stability, and sensitivity (noise
margin). Note that the NTC sensors, while having many positive
attributes, are quite non-linear over an extended range. Calibration at
the actual set point is necessary for absolute work.
I am building the typical beginner clock project. A "surplus" Trimble
VCO is used to clock a microprocessor. the micro has an internal 8x PLL
which adds resolution in timing and simplifies the code. A GPS PPS is
the reference and a DAC closes the loop. (Info if anyone interested).
Instrumentation is limited.
This all works ok and I am getting out old control texts to try to
improve the performance but the voltage required by the VCO to maintain
lock is dropping at an alarming rate - about 30 mV / week. Is this
normal or do I have a lemon? What are good sources for non-lemon quartz
Thanks much in advance for any suggestions.
More information about the time-nuts