[ntp:questions] Influence of temperature on PC oscillators
mills at udel.edu
mills at udel.edu
Fri Apr 20 16:38:35 UTC 2007
As a member of the amateur radio community myself, I would much rather
diffuse a spectral line over a wide range. This reduces the spectral
density and the power over a narrowband communication channel.
Once upon a time computer equipment was build using hefty shields,
bypass capacitors and finger stock that reduced or eliminated
electromagnetic interference (EMI). The FCC has rules about that, but
Part 15 of the rules has so many exceptions as to be essentially
worthless. For proof, put a shorwave receiver anywhere near a PC and
hear the birdies.
As further proof, consider the Manassass VA Broadband Power Line (BPL)
system that brings Internet to the home. It wipes out shortwave amateur
communications over most the the area and blatently violates Part 15.
Repeated complaints to the FCC have not found adequate response. With
that in mind, be happy there is provision to at least (potentially)
reduce the spectral density of PC interference.
While I am barking this tree, note that in the 21 years I have been with
UDel I have seen the incidental EMI due miscellaneous causes such as arc
welders, UPS systems and whatever else rise to the point where WWVB
receivers are now unusable and shortwave communications of any kind have
been rendered almost useless. I have calibrated the increase in noise
level over the years at a 30-dB increase. What used to be nominal zero
on an S meter is now reading S7.
David Woolley wrote:
> In article <iOadnZzu3ZnSSbzbnZ2dnUVZ_qjinZ2d at megapath.net>,
> hal-usenet at ip-64-139-1-69.sjc.megapath.net (Hal Murray) wrote:
>>I don't agree with the suggestion to disable the spread spectrum
>>stuff. It's there for a reason. Effectively, your CPU will run
> The reason is basically a workaround for poor quality combinations of
> case and motherboard. I believe the amateur radio community feels that
> it is a bad solution because it basically moves the generated radio
> frequency interference from a number of spot frequencies to a wide band
> of frequencies, without reducing it well below natural background levels
> Although spot frequencies are still a nuisance, most of the band is still
> available to weak signals, but broad noise wipes out weak signals across
> the whole band, even though the result complies with the relevant legislation.
> That legislation is a compromise, aimed at minimising manufacturer compliance
>>slightly slower. I haven't seen anything published to indicate
>>that the speed is sufficiently less stable so that ntpd will notice.
> I assume it is pseudo random, so the long terms stability will be as good,
> but the short term stability may be compromised, although I'm not sure of
> the timescale.
More information about the questions