Strange digital noises from motherboards...

[KILOWATT]

Hi thanks for your time.

I have a 3 yrs old mobo (ASUS P4P800-VM) and i have strange digital noises
leaking from the mobo right into the SoundMax integrated digital audio chip.
I mean "blips" sounds when restauring/minimizing windows, rapidly ticking
sounds when moving the mouse that sounds like "Vrrrrrrrrrrrrrrrrrrrrrrrr". I
can hear those sounds very clearly when i use headphones, although i have to
obviously crank up the volume to hear those noises with the cheap plastic
desktop speakers. I thought that it was typical for computer audio, but my
friends who have cheap soundblaster value OEM cards or integrated audio on
old A-Open mobo don't have this issue! I've bring my headphones and could
not hear any noise on their system.

The same issue happens no matter if i use Windows98SE or XP's audio driver.
You may be tempted to say that my mobo is cheaply designed or defective, but
it's not the first time i'm experiencing such issue. With my old IBM Aptiva,
i had that "Vrrrrrrrrrr"'ing noise also when i did moved the mouse. Under
warranty, the tech at authorized service center replaced the motherboard.
Same problem. He then replaced the K6-2 CPU...without success. He concluded
that the board design was flaky and to solve my problem, i would have to
install a pci sound card. So to go to the point i wish to build a new
computer and use that P4P800-VM based one as a web/ftp server. For the new
one, i would like to know if anyone have experience with the integrated
audio on the ASUS P5LD2 ?
( http://tinyurl.com/lh9tb )

I'm tempted to rather buy the Deluxe version wich is not a lot more
expensive:
( http://tinyurl.com/zpqqo )
It seems to have a better integrated audio. Any comments about the sound
quality of those boards or comments related to what i'm experiencing
actually are welcomed. Thanks again for your time.


--
Alain(alias:Kilowatt)
Montréal Québec
PS: 1000 excuses for grammatical errors or
omissions, i'm a "pure" french canadian!
(If replying also by e-mail, remove
"no spam" from the adress.)

 

[kony]

Hi thanks for your time.

I have a 3 yrs old mobo (ASUS P4P800-VM) and i have strange digital noises
leaking from the mobo right into the SoundMax integrated digital audio chip.

It's because motherboard integrated sound does not allow for
enough board real-estate, and/or enough of a board budget or
perception of public demand for, the components necessary to
filter the power rail noise sufficiently.

The generic solution is to buy a decent sound card (not
necessarily an expensive one, but at least with a fair
amount of capacitors, linear regulation stage(s), and LC
filters. In other words, you might not get the full benefit
from some cheapo $8, barely populated C-Media (or similar)
based sound card, but most of yesteryear's or today's medium
grade cards should suffice for all but the most noisey of
system power rail problems.

I mean "blips" sounds when restauring/minimizing windows, rapidly ticking
sounds when moving the mouse that sounds like "Vrrrrrrrrrrrrrrrrrrrrrrrr".

These come from power rail modulated noise, and in many
cases, replacing the power supply will not help- it can
randomly seem to help but not necessarily from buying a
better, or worse, PSU, rather it could depend on the
resultant frequency of the ripple when powering the specific
board, specific load of the components on that board, and
specific tasks / % utilization, during the sound glitches.

In short, systems are being designed to provide merely
digitally-sufficient power, which need not be very clean,
while analog audio is another beast- requiring truely low mv
ripple for best results.

I
can hear those sounds very clearly when i use headphones, although i have to
obviously crank up the volume to hear those noises with the cheap plastic
desktop speakers.

The resolution of the amplification stage (Or lack of one,
the ultimate resolution IF it can power the load
acceptibly), can determine how well you can hear it. That
is, on a system exhibiting it to an extent loud enough that
any given person with accute enough hearing, can hear it.

I thought that it was typical for computer audio, but my
friends who have cheap soundblaster value OEM cards or integrated audio on
old A-Open mobo don't have this issue! I've bring my headphones and could
not hear any noise on their system.

There is no short answer here, what I wrote above still
applies, plus some boards do go to extra efforts to try to
improve the audio power filtration, and/or the parts being
powered induce less ripple on the power rails, and board
design can also mitigate the ripple in various ways,
depending on type, magnitude, several parameters.
Unfortunatey, boards are rushed to market, short product
cycles mean that if a board exhibits such a tendency there
is not going to be a reworking of the whole thing to fix
such a problem... especially since it may not even occur in
all system configurations.

In general though, the more high frequency components you
have on the power rail powering the codec, and the larger
the current swings of these parts (particularly those that
have sleep modes, or different video states), the more
likely to encounter the problem.

The same issue happens no matter if i use Windows98SE or XP's audio driver.

Sometimes changing some mixer settings can offset this, but
it's a fiddly attempt and often does not help, or help
enough.

You may be tempted to say that my mobo is cheaply designed or defective, but
it's not the first time i'm experiencing such issue.

Nope, not necessarily cheap relative to other boards, these
kinds of problems do crop up on many, many, many different
boards. I'm quite critical of such things and even hard-mod
my sound cards' filters in some instances just to get lowest
noise possible.

On the other hand, in the grande scheme of things, the
problem would not exist if the motherboard manufacturer had
the time (in their design stage, relating to
time-to-market), the will (to allocate sufficient space on
the board and cost for the filters), and the perception that
the market cared.

Unfortunately, reviewers seem fascinated by digital tricks
rather than the basics like clean output. They'll hook up
some speaker system they thought was great, but wasn't
particularly sensitive, and completely miss the problem.
Their review praises audio output that is inferior to 5 year
old, now $10 sound cards. Some of it is almost laughable
too, they assume such issues were fine years ago and there
is nothing to test, instead of realizing that they simply
lack the methodology (Or will, parts) to discriminate.

In the end, it's still a similar situation to other
motherboard-integral parts, that if you are a discriminating
user (regarding that function), you may want to upgrade to
an add-on card.

With my old IBM Aptiva,
i had that "Vrrrrrrrrrr"'ing noise also when i did moved the mouse. Under
warranty, the tech at authorized service center replaced the motherboard.
Same problem. He then replaced the K6-2 CPU...without success. He concluded
that the board design was flaky and to solve my problem, i would have to
install a pci sound card. So to go to the point i wish to build a new
computer and use that P4P800-VM based one as a web/ftp server. For the new
one, i would like to know if anyone have experience with the integrated
audio on the ASUS P5LD2 ?
( http://tinyurl.com/lh9tb )

I caution you to not consider any answers to this as
definitive. There may easily be people who own your present
board and don't find any trouble with it's audio!

For a reasonable level of certainty that your analog output
will have least noise, you need a medium-or-better quality
sound card. That need not be one costing a lot- some of the
costs of sound cards may go to the name (like newer Creative
stuff), or hardware processing for games (your call, it can
help if you use EAX/3D in games).

I'm tempted to rather buy the Deluxe version wich is not a lot more
expensive:
( http://tinyurl.com/zpqqo )
It seems to have a better integrated audio.

For your purposes, probably not. Even if a codec has better
specs in general on paper, the PSRR rating is seldom that
much different, and is one of the more relevant ratings
effecting your needs. It could easily be that the
particular parts you populate the board with, don't result
in as much rail ripple as on your current board- or it could
even be worse, paying a premium for a board doesn't guard
against this.

One part of motherboard design that can significantly effect
this rail noise is the number of stages, and frequency of
these, for powering the CPU. IE- the CPU VRM subcircuit, #
of mosfets and degree of offset used by the switching
controller. Not having the specs readily available, nor
being able to predict the resultant frequency for any
particular CPU, there are just too many variables to predict
this unless one has more, applicable evidence collected
already- which I do not. A generalization might be that the
lowest combined switching frequency of all elements will
tend to introduce the least, (most recoverable) errors.

Any comments about the sound
quality of those boards or comments related to what i'm experiencing
actually are welcomed. Thanks again for your time.

Buy the board you want ignoring the issue of sound output
quality, except in an attempt to get a board that provides
enough slots that you can add a sound card if later deemed
necessary. Another alternative is an external amp with
digital input, and of course compatible digi-out feature on
the motherboard. A 3rd option is USB sound dongle, though
they tend not to be the best for directly driving headphones
either, expect possibly some of the newer stuff (which I
have no experience with, but possibly someone like Zalman
makes something higher-end with USB out). Such a premium
USB solution is bound to cost more than a traditional card
though, but could be worthwhile if it also has other
features you want.

 

[meow2222]

kony wrote:

On the other hand, in the grande scheme of things, the
problem would not exist if the motherboard manufacturer had
the time (in their design stage, relating to
time-to-market), the will (to allocate sufficient space on
the board and cost for the filters), and the perception that
the market cared.

It really doesnt.

Unfortunately, reviewers seem fascinated by digital tricks
rather than the basics like clean output. They'll hook up
some speaker system they thought was great, but wasn't
particularly sensitive, and completely miss the problem.
Their review praises audio output that is inferior to 5 year
old, now $10 sound cards. Some of it is almost laughable
too, they assume such issues were fine years ago and there
is nothing to test, instead of realizing that they simply
lack the methodology (Or will, parts) to discriminate.

One should be aware that a lot of reviewers are just another cog in the
advertising process. This is particularly so of magazine articles. Mags
compete for advertising revenue, and advetrisers go with whoever hypes
their products the most, so the nicer they can be about the junk they
review the more money they make.


NT

 

[KILOWATT]

Thanks for your reply Kony. Since the reviews i've found yet don't tell a
lot about the audio quality of these boards, i'll simply buy the deluxe
version of the P5LD2 since there's anyway other integrated stuff on that
particular board that may be useful for me. @ $30 more, it won't be such a
pain to discover that the sound quality is no better than my P4P800-VM and
that i will need an add-on pci sound card.

You are right about the biasing of some reviewers, Meow. That's why i also
read when possible, user reviews.

 

[Mitochondrion]

I've had this problem, Nothing you can do short of enable spread
spectrum in your chipset options to reduce interference, and buying a
sound CARD, IOW: get the sound processing away from the chipset