enough

Hmph!...  Spread Spectrum clocks do NOT make the shielding any
easier, it's just a fudge for the accountants who won't fund a proper
job in the first place.  It only "fools" the QP detector in a measuring
receiver into showing a lower value, it does not "Fix" the problem.

It could be said, it only fools naive engineers into thinking it can
cure an EMI problem.

In any case, most of the problems that could be caused by CPU clock (or
derived signals) leaking out and into something else, are easier solved
if the blessed sprog so created is not spread over a large area.

As to spam.  It's OK between two slices of bread, with a good helping of
pickle!  It was also one of the best Monty Python sketches ever, in
fact, it's where the name was adopted from, for junk email.

Regards to All.

Dave Baxter.    Soon to go and try to fix a couple of multi kW RF amps,
a customer has let the high pressure recirculating smoke out of,
again....

And I thought I was the only person in the world who noticed that bit
of subterfuge. It actually makes EMI worse, artificially raising
the test limit by smearing the signal to get past the FCC's spectrum
analyzer-defined peak limit.

The Part 15 limits for such things as the FM broadcast band, on the
other hand, are defined by field strength. That's the only
cheat-proof way to specify emitter power testing.

--

--David Forbes, Tucson, AZ
http://www.cathodecorner.com/

Oh yes you are not alone :-)) I have posted two or three times to EDN on
this topic.
Alan G3NYK

----- Original Message -----
From: "David Forbes" dforbes@dakotacom.net
To: "Discussion of precise time and frequency measurement"
time-nuts@febo.com
Sent: Thursday, November 26, 2009 5:03 PM
Subject: Re: [time-nuts] Spread Spectrum Spam!

https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts

Alan Melia wrote:

Regardless, there is a limit to just how many designs can use jittered
clocks. For a number of cases one goes in the other direction. It's not
that tricky to design to be within limits if one spend some thought into
it. Usually, what is wise for signal integrity is wise for EMC.

Cheers,
Magnus

The MIL-STD-461 EMI specs have a different mask for broadband sources than
for narrow band sources. The spread spectrum clocks merely take advantage of
that. I am not familiar with the FCC spec, but I suppose they probably have
a similar handling of narrow band versus broad band emissions.

If you do not like it, do not blame the technology, blame the people who
wrote the specs.

For some applications, particularly narrow band applications where the
discrete spur would otherwise fall inside your passband, a spread spectrum
clock is actually a good thing. For other applications, YMMV.

Another reason why it is popular with manufacturers is that it is typically
harder to identify the source of a broadband interference compared to
narrowband, so when a customer has a problem with his or her wireless
device, if he or she cannot readily identify the source of the interference,
the customer is more likely to blame the device than an unidentified source
of interference.  In a society where profit is king, this is like music to
the ears of management and the finance guys.

It bugs me to no end that as a power supply design engineer, I came up with
the idea in 1986-1987 and even breadboarded a "random frequency converter"
(as I called it then) by using a pseudo-random generator (shift register and
a few gates) driving a DAC driving a current source driving the clock of a
PWM chip in a power converter. The corporation that I worked for at the time
had no interest in even checking if it was patentable (they did not have a
patent department). It did exactly what it was supposed to do, reduce peak
emissions by 10-20dB depending on the depth of modulation and the particular
test (I was more interested in MIL-STD-461 than FCC at the time). I still
have the breadboard...

Didier KO4BB

Interestingly some amateur radio 13.8V switch mode supplies have a QRM (interference) reduction control. This shifts the PWM frequency slightly so you an shift the noise off of the signal you are listening too! Complete fudge. In Europe the CE EMC tests use Quasi-peak detectors and slow scan rates that respond less to spred clocks. The spreading can be turned off in the BIOS on most machines if needed.
 
Robert G8RPI.

--- On Fri, 27/11/09, Didier Juges didier@cox.net wrote:

From: Didier Juges didier@cox.net
Subject: Re: [time-nuts] Spread Spectrum Spam!
To: "'Discussion of precise time and frequency measurement'" time-nuts@febo.com
Date: Friday, 27 November, 2009, 3:02

The MIL-STD-461 EMI specs have a different mask for broadband sources than
for narrow band sources. The spread spectrum clocks merely take advantage of
that. I am not familiar with the FCC spec, but I suppose they probably have
a similar handling of narrow band versus broad band emissions.

If you do not like it, do not blame the technology, blame the people who
wrote the specs.

For some applications, particularly narrow band applications where the
discrete spur would otherwise fall inside your passband, a spread spectrum
clock is actually a good thing. For other applications, YMMV.

Another reason why it is popular with manufacturers is that it is typically
harder to identify the source of a broadband interference compared to
narrowband, so when a customer has a problem with his or her wireless
device, if he or she cannot readily identify the source of the interference,
the customer is more likely to blame the device than an unidentified source
of interference.  In a society where profit is king, this is like music to
the ears of management and the finance guys.

It bugs me to no end that as a power supply design engineer, I came up with
the idea in 1986-1987 and even breadboarded a "random frequency converter"
(as I called it then) by using a pseudo-random generator (shift register and
a few gates) driving a DAC driving a current source driving the clock of a
PWM chip in a power converter. The corporation that I worked for at the time
had no interest in even checking if it was patentable (they did not have a
patent department). It did exactly what it was supposed to do, reduce peak
emissions by 10-20dB depending on the depth of modulation and the particular
test (I was more interested in MIL-STD-461 than FCC at the time). I still
have the breadboard...

Didier KO4BB

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Some spacecraft have a requirement that all the PWM supplies by
synchronizable to an external source.  The idea is that at least the spur is
in a constant location, rather than wandering around (PWM natural switching
rates not being the most stable in the world)

On 11/27/09 12:40 PM, "Robert Atkinson" robert8rpi@yahoo.co.uk wrote:

Some synth sig-gens use the reference divided down to drive the PSU (Marconi
2955 for instance) for the same reason I think
Alan G3NYK

----- Original Message -----
From: "Lux, Jim (337C)" james.p.lux@jpl.nasa.gov
To: "Discussion of precise time and frequency measurement"
time-nuts@febo.com
Sent: Friday, November 27, 2009 10:03 PM
Subject: Re: [time-nuts] Spread Spectrum Spam!

Some spacecraft have a requirement that all the PWM supplies by
synchronizable to an external source.  The idea is that at least the spur is
in a constant location, rather than wandering around (PWM natural switching
rates not being the most stable in the world)

On 11/27/09 12:40 PM, "Robert Atkinson" robert8rpi@yahoo.co.uk wrote:

so

rates

BIOS

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Alan Melia wrote:

Some Rubidiums lock their pump frequency to the OCXO. This can cause
harmonics on the output. I once saw it on a rubidium, the jitter was
unexpectedly high for compared to what the phasenoise plot would make
you expect, but when scoping the signal it was clearly seen.

Cheers,
Magnus