[time-nuts] Thunderbolt? (re simple gpsdo.)

David davidwhess at gmail.com
Sun Jan 1 20:43:06 UTC 2012


Jim Williams did this in one of his designs for measuring low
frequency reference noise.  The large value low leakage wet tantalum
capacitor he used was like $400 and it took 24 hours for the
dielectric absorption to settle:

http://www.linear.com/docs/28585

You can get the necessary time constant using a good 1uF film
capacitor with good design and construction in this case.

On Sun, 1 Jan 2012 15:11:04 -0500, Bob Camp <lists at rtty.us> wrote:

>Any real world capacitor will have a dielecric with an associated insulation resistance. It's a "more money gets better performance" sort of thing, but there are indeed limits. A 1000 uF cap that has a "good" insulation resistance number might cost you more than some new cars
.
>
>On Dec 31, 2011, at 11:54 PM, Chris Albertson wrote:
>
>> On Sat, Dec 31, 2011 at 5:56 PM, Hal Murray <hmurray at megapathdsl.net> wrote:
>> 
>>> I think the main problem in this area is building a low pass filter with a
>>> long time constant.
>>> 
>>> The time constant of the filter has to be:
>>> long relative to the noise from the phase detector
>>> short relative to aging of the oscillator
>>> short relative to environmental changes
>>>   (so the osc can track temperature and voltage
>>>     those changes may be in the PLL system rather than the osc)
>>> 
>>> If we are starting with PPS (rather than 10KHz), the filter time constant
>>> needs to be 10s or 100s of seconds.  How do I build an analog filter with a
>>> time constant that long?
>>> 
>> 
>> Time constant is just R*C.  If you have a 1000uF cap and a 1K resistor you
>> have 1 second.  In theory you could build 100s just by using a 100K
>> resistor but I think real world components are not perfect enough.



More information about the time-nuts mailing list