This news has made AMD'ers happy

[KR Williams]

(e-mail address removed)
says...

I wonder if the Seagate tech support guy mentioned all the patents
that applied. The Convolve press release sounded like they were more
concerned about Seagate using specially-shaped drive pulses.

I can only go by the article you referenced. The only
information there was about changing the performance
characteristics based on the desired sound level (IOW, the UI).

There is no word in the the claims of *that* patent regarding
performance tuning (it is taught though). The bottom line is that
they're going to have a tough time proving they invented
performance tuning of disk drives in 1999. They're about forty
years too late.

....and there ain't no "Square waves" in there. The waveforms are
altered (energy controlled) based on distance of flight. This is
nothing at all new.

 

[Robert Redelmeier]

(e-mail address removed) says...

I think that's an entirely simplistic view. I'd like to
see you substantiate this.
http://www.maxuk.net/241mp/geforce-image-quality.html

There's a *lot* more to EMI reduction than "LC filters".

Of course there is. I'm sorry if I expressed myself poorly.
What I meant was that poor LC filtering would reduce available
image quality.

Furthermore, a mfr that was doing a poor job of EMI managment
on the PCB level (trace routing, layers) might be tempted to
compensate by excessive LC filtering.

-- Robert

 

[KR Williams]

Of course there is. I'm sorry if I expressed myself poorly.
What I meant was that poor LC filtering would reduce available
image quality.

Please, "LC filters" aren't used for EMI reduction *AT ALL* in
this application. Al LC filter simply rings at its resonant
frequency, it doesn't attenuate anything. If you're thinking of
ferrite beads, they're not an 'L' at all. They're 'R's at the
frequency of interest. It takes an 'R' to dissipate power.
That's my point. There aren't LC circuits used to attenuate
anything, unless they're *REAL* morons. Perhap;s they can't pick
the right ferrite, but that's another issue (if they cannot even
follow the reference design...).

Furthermore, a mfr that was doing a poor job of EMI managment
on the PCB level (trace routing, layers) might be tempted to
compensate by excessive LC filtering.

I don't buy it. The ferrite on the I/O won't help one iota if the
card is radiating elsewhere. Sure there are crappy designs, but
I'm not buying your LC filter theory. ...at least as stated.

 

[Robert Redelmeier]

Please, "LC filters" aren't used for EMI reduction *AT ALL* in
this application. Al LC filter simply rings at its resonant
frequency, it doesn't attenuate anything. If you're thinking of
ferrite beads, they're not an 'L' at all. They're 'R's at the
frequency of interest. It takes an 'R' to dissipate power.
That's my point. There aren't LC circuits used to attenuate
anything, unless they're *REAL* morons. Perhap;s they can't pick
the right ferrite, but that's another issue (if they cannot even
follow the reference design...).

Well, maybe my terminology is off. I see three sets of "L"
components and "C" components on vidcards near the VGA connector.
They are most probably connected to the RGB analog output lines
and reduce the sharp corners, overshoot and high rise rate of
digital signals for analog transmission.

I don't buy it. The ferrite on the I/O won't help one iota if the
card is radiating elsewhere. Sure there are crappy designs, but
I'm not buying your LC filter theory. ...at least as stated.

Ah, but if the card is otherwise good, or nearly so, then
won't L & C omponents on the output stages help reduce EMI
from the monitor cable (aka antenna)?

-- Robert

 

[KR Williams]

Well, maybe my terminology is off. I see three sets of "L"
components and "C" components on vidcards near the VGA connector.
They are most probably connected to the RGB analog output lines
and reduce the sharp corners, overshoot and high rise rate of
digital signals for analog transmission.


Ah, but if the card is otherwise good, or nearly so, then
won't L & C omponents on the output stages help reduce EMI
from the monitor cable (aka antenna)?

No. You really want *resistors* there (ferrite beads are really
resistors at the frequency of interest). The idea is to absorb
the energy, not so much "filter" it. Who knows what silly
people do though. Lesse... <opens G550 box> ...nope no L's
there!