DOS/Windows and PC clocks]

Andre' Kesteloot andre.kesteloot@ieee.org
Mon, 26 Mar 2001 10:23:51 -0500


James Moritz wrote:

> Dear LF Group,
>
> Although it may be possible to access the PC's internal clocks one
> way or another, I can't see it would be much help for timing
> accuracy - after all, it's all derrived from a very ordinary, un-
> trimmed crystal on the motherboard somewhere.
>
> I did some work at U of H a while ago with Dave Lauder, G0SNO,
> investigating PC clocks. In fact, a lot of PCs (and other things) are
> now using 'dithered clocks', where the clock is frequency
> modulated to a deviation of a fraction of a percent, by a spiky
> looking triangle wave. The reason for doing this is to spread
> radiated noise emissions from the PC over a wider range of
> frequencies, which in turn reduces the noise power seen by a
> narrow band EMC receiver when noise emissions testing is being
> performed. This slightly perverse idea is basically a fiddle to allow
> PC's to generate more RF noise, while still passing the EMC
> regulations. It saves the manufacturers some money on screening
> and filter components. I'm sure it does not do much for timing
> accuracy, though!
>
> Cheers, Jim Moritz
> 73 de M0BMU