Power conditioning for computer audio

If you are using a sound card in a desktop PC, the ferrites may add significantly to the jitter by making the power supply sluggish.

Hi Steve,

Could you explain why that would be so?

I would expect, given that we are talking about a digital output of the computer (as opposed to an output whose current demands may fluctuate widely) that capacitive energy storage and decoupling on the sound card, and perhaps the motherboard as well, would isolate the sound card circuits from any power supply sluggishness. If a lot of bits are toggling at once in parts of the circuit, then yes there may be a sudden spike in current demand for perhaps a few nanoseconds, but it seems to me that assuring that demand is met during those few nanoseconds is the responsibility of capacitors on the card, not the power supply.

And I would also expect that by far the greatest demands on the power supply in terms of dynamics would be due to the cpu, its associated chipset, and the video card, none of which would be heavily taxed in an audio application.

Thanks, and happy holidays!

-- Al

Hi Guys:

While snowed in today, I was surfing around the net and came across this AC regeneration product from PurePower. It looks like the product is used to clean up chaff from power utility lines. Can someone discuss whether and how so called Power Gremlins (see attached URL link) affect sound quality from stereo gear.

Does it matter if the gear is SS or tube?

What impact (if any) does a regulated power supply in a piece of equipment have on the quality of the unit's internal current?

BTW, the PurePower AC regenerators are not cheap!! See the attached URL.

http://www.purepoweraps.com/Top5.htm

Bifwynne, you might find this thread and this thread to be of interest.

While snowed in today, I was surfing around the net and came across this AC regeneration product from PurePower. It looks like the product is used to clean up chaff from power utility lines. Can someone discuss whether and how so called Power Gremlins (see attached URL link) affect sound quality from stereo gear. Does it matter if the gear is SS or tube? What impact (if any) does a regulated power supply in a piece of equipment have on the quality of the unit's internal current?

From a technical standpoint I don't think it's possible to generalize or predict the effects of a power regenerator or a power conditioner on any given system. There are far too many variables and unknowns involved. Just to cite one small example, rfi may to some degree bypass the power supply, couple into arbitrary internal circuit points that are in the signal path, and intermodulate with the signal resulting in unpredictable effects at frequencies that are audible.

That lack of predictability is unfortunate because, as you point out, these things can be quite expensive.

Regards,
-- Al

Thanks Almarg. The second thread was right on point!

If power Gremlins are really finding their way into the final output that goes to a speaker, from a layman's point of view, that looks like pure distortion. I take your point about rf getting in the middle, but wouldn't that always be the case. It would seem that power-line Gremlins just crap up the signal even more?

I'm still looking for an answer from the forum, in layman's terms, as to whether it matter if the gear is SS or tube? And, what impact (if any) does a regulated power supply in a piece of equipment have on the quality of the unit's internal current? IOW, does an electronic power supply in a way also act like a power regenerator??

"Could you explain why that would be so?

I would expect, given that we are talking about a digital output of the computer (as opposed to an output whose current demands may fluctuate widely) that capacitive energy storage and decoupling on the sound card, and perhaps the motherboard as well, would isolate the sound card circuits from any power supply sluggishness."

Sorry, but there are no voltage regulators out there that reject input voltage variations like this, even the best ones. They are all imperfect. Even lots of power decoupling will not do it. Regulators all have finite response times, so regulating large current transients is a tall order too. All of these things are helpful, but still imperfect.

When it comes to generating a low-jitter digital output, the power must be extremely clean with fast responding regulators, and isolated from everything else.

Steve N.
Empirical Audio