Sound Card power consumption

[Devonavar]

I've never seen any testing done on modern sound cards to see how much power they draw. The usual estimate is 10W, but I wanted to see how accurate this estimate is, so I decided to actually test it. I've been playing around with a Creative Labs Audigy 2 Platnum Pro ZS (or something), so this is the card I used for testing. This is the highest end version of the Audigy Line, which is pretty much the standard sound in use. Most other cards are either derivitives of this line (including cards from other manufacturers), or from VIA's Envy24 chipset, which I am guessing is similar in power requirements.

For testing, I used a Seasonic Power Angel to measure the change in AC wattage between various configurations. The rest of the system used the following components:

Antec TruePower 380S power supply (from a Sonata case)
Gigabyte GA-7ZM motherboard
AMD Duron 750 GHz
3 x 256 MB PC133 RAM
ATI 3D RAGE IIC
Maxtor D740X-6L
LiteOn DVD-ROM
Generic NIC

At idle, this drew 74W AC from the wall — higher than I expected. This established a baseline power level to which the sound card was added.

The Audigy 2 has a breakout box to allow various sound I/O to be plugged in, but not all models come with it. My first test added just the Sound Card itself — the breakout box was not connected.

With just the card installed, the system idled at 79W — a 5W (AC) increase. A small load was placed on the sound card by playing back an MP3 in iTunes. This pushed the power consumption up a single watt, to 80W. This task hardly changed the power level of the system at all. In fact, the increase in power draw may have come entirely from the ~5% CPU load needed to decode the MP3. A more challenging test was needed.

I decided to play back a DVD-Audio disc that came with the card. This music was 96 kHz / 24 bit, and was encoded in 5.1 Dolby Digital. The card's hardware decoder was used to decode the file. First, however, I needed to compensate for the power that would be drawn by the DVD drive itself. The disc was inserted, and the drive was allowed to idle without spinning down. This produced a power draw of 85W: 6W above idle. I then began to play the DVD-A, and power consumption rose to 87W — a net 2W that could be attributed to the sound card.

Now it was time to plug in the breakout box and see if it affected the power draw. It did: Power consumption at idle was 85W. The MP3 and DVD-A tests were re-run, and the net gain was again 1 and 2W respectively. This put the total AC power draw of the sound card and the breakout box at a maximum of 13W AC.

For the sake of argument, I will assume that the efficiency of my TruePower 380S is 70% at the load I was using. This is almost certainly high, as the original SPCR test showed it to be 65% efficient at 90W DC load. As such, my final estimate will err on the high side — a safe choice if this data is used for sizing a PSU.

The end result of my testing shows that my Audigy card drew the following loads (in DC watts):

Idle, no breakout box: 4W
Max Load, no breakout box: 5W
Idle, with breakout box: 8W
Max Load, breakout box: 9W

Because of the extremely rough and ready nature in which I conducted this test, I would estimate the margin of error to be at least a watt (~10%). However, my results appear to confirm the standard "10W" response whenever anyone asks how much power an audio card draws. If fact, most cards appear to draw even less.

One caveat: I did not test the power draw under recording conditions because I do not have the appropriate recording equipment. I would expect that, since my card doesn't provide Phantom Power, power draw is similar to playback situations. I would also expect power draw to rise slightly if multiple sources are played back (mixing) or digital effects were used (DSP, common in games). However, after playing around with EAX effects, I do not think that either of these will significantly affect power consumption.

Another caveat: I could think of no way of controlling for CPU usage, so the load measurements may be even smaller than they appear to be. CPU usage for both MP3 and DVD-A playback was typically ~5% above idle.

[pipperoni]

If you really want to be scientific about measuring the power consumption of your soundcard, my suggestion would be to tape over the power leads of your sound card and use an isolated supply to power your card.

It is very ambitious, but Xbit labs did something similar in their Graphics card power measurement. Here is a list of the PCI pin out. Safe to say it would be very ambitious.

All in all though, I'm not sure how valuable the data would be. My general rule of thumb is if it doesn't have a heatsink and its not moving, its not using enough power for me to care.

[Pilot]

Thanks for the figures, i was about to run into the same experimentations. I measure everything these days with my power meter.

In fact, I'm rather disappointed because of my UPS and my modem router. While it's understandable the UPS is constantly loading, (i'll give my figures if interested) the modem is not : it still runs full speed whether there are computers connected or not.

AMD Duron 750 GHz

How mighty !!! No, was just kidding.

[jojo4u]

Nice findings, Devonavor.
You might be interested in getting the disconnect to work on you Duron. This could save you couple of watt.

http://forums.silentpcreview.com/viewto ... re+cooling

[Devonavar]

If you really want to be scientific about measuring the power consumption of your soundcard, my suggestion would be to tape over the power leads of your sound card and use an isolated supply to power your card.

Way too much effort for a night's work, but perhaps worth thinking about for a project later. There's also an external power connector that would have to be tapped. I was just looking for an estimate of the power usage. Now I have one.

[~El~Jefe~]

some say its 15 watts. which is impossible. so there's my useless ungrounded tip for the day.

[EvoFire]

Audigys don't have heatsinks right?? I've heard people mentioning they get fairly warm while running for prolonged periods.

[Funny Man]

If you really want to be scientific about measuring the power consumption of your soundcard, my suggestion would be to tape over the power leads of your sound card and use an isolated supply to power your card.

Not as hard as you might think. Many cards use a voltage regulator for their analog section, which means there is a single trace to follow. On some of my cards it's straight from 12v with through a handy removable 0ohm link, likewise for the digital 5v power rail. Flaw in this method is: PCI and control components are excluded, and complex sound-cards=YMMV.

Actually for a useful suggestion, Devonavar, I would try to find a utility that can put selected devices into S3 power down, to get a more accurate baseline measurement.
Also try turning the volume up, with un-amped speakers plugged into the "jumpered" speaker socket (if you hadn't already). Max db might add 2W/eff.
More tips for removing other factors, like CPU use:
Put (wait for) the HDD into sleep mode (also lets you know no apps/background tasks are running)
Play tone://100 in fb2k.

[Flandry]

That's some great work. What i'm wondering is how Creative's new monster audio processor cards will measure. Next thing you know, we'll have to silence the fans on the audio card!

[~El~Jefe~]

Yeah I know, i was thinking about the new X-Fi cards with their much much larger cpu's. I would only assume that they use more power.

[steveo]

This may be a bit off subject, but I have been in the market for a new sound card, for home recording. I can't find a single brand or type that is what i want (which mainly consist of a 5.25" front interface with 4 inputs). Does anybody on SPCR record, or have a favorite sound card they think i might like?

[Myth!]

depends on the level of quality required really. The minimum pro-audio cards are about £70 which gives you breakout cables and hardware effects on many channels etc. If you need gaming effects too then an audigy of some kind is probably the path to take.... any card over £20 will be an enourmous improvement over onboard audio ESPECIALLY when recording line/in and mic.
Just swapped in my £20 SB live value thingy and the difference on playback through 5.1 headphones is great... no more hiss and HDD noises.

If you dont need gaming sound and are serious about your audio.... think about something which goes on at length about its 'pristine DAC's' , 24bit/192khz, and gives you all the right connectors (1/4" jack sockets for instance)

hope this helps