Desktop vs. Laptop USB

serengetiplains · « **on:** 24 Sep 2006, 04:26 pm »

Since jumping into the USB fray some months ago (I own a DEQX unit modified by Steve with USB --> I2S input, more on the DEQX later), I've wondered whether the source --- the computer, and specifically the source's power supply --- has any influence on the sound of a USB audio setup. I use an Apple G5 desktop as my main music source. Yesterday I grabbed my daughter's PowerBook G4 and ripped some wav files into iTunes, and downloaded the M-Audio Transit driver to output music data to the DEQX. My initial impressions comparing desktop to laptop is that laptop, run on battery only with charger unplugged, is noticeably better. One CD I used for comparison is Leonard Cohen's "I'm Your Man." This recording is among the worst I own, which is particularly unfortunate since it's probably Cohen's best work, and no doubt is his best effort (he edited the lyrics of some songs 750 times). Played via the desktop, the rasp in Cohen's voice is harsher than played via laptop, the presentation from which is overall less edgy, softer, more delineated with better channel separation (typically an important key to sound quality), less phasy, etc.

I must say, I'm not surprised at this outcome, as IME *everything* affects the sound, so why not a super-noisy computer power supply even if it's serving duty only to shuttle music data into a USB line? Don't take it from me, though, as an experiment confirming or denying the sonic impact of the source psu is as easy as swapping one USB cable from computer to computer (though you might have to commit one arm to keeping a teenager from accessing one of the computers).

Has anyone else experimented likewise? I'd be interested hearing your comments.

serengetiplains · « **Reply #1 on:** 24 Sep 2006, 04:32 pm »

I meant also to say I'm of two minds regarding the outcome of this observation. I'm happy to have found another sonic upgrade, but I'm left wondering how to now configure my audio files. They sit on two hard-drives and consume a total 500MB of space. I could possibly create an external drive setup --- the 750MB Seagate should suffice --- and run the external unit from its own dedicated supply, probably battery-based. The Seagates evidently require a single +12V supply, but at 2.8A I wonder if a battery supply would falter.

Hobby audio is a full time job!

Occam · « **Reply #2 on:** 24 Sep 2006, 04:54 pm »

Is the usb cable power (interface) provided by the computer, as typical, or internally from the Deqx? Has Steve modded this? If it is from the computer, the preference for the laptop is certainly understandable. Even though both (I assume) use switching supplies, the battery supplied laptop can at least be viewed as a 'offline' supply, not subject to the vaugaries of line noise.

As you know, the Deqx is powered by a switching supply, and benefits tremendously from powerconditioning. JoshK uses a a Felicia on his quite effectively. Given your proclivities, should I assume its fed by your byzantine powercondioning chain?

serengetiplains · « **Reply #3 on:** 24 Sep 2006, 05:06 pm »

Paul, Steve modded the DEQX power supply --- swapped it out for a Mini-Pac Vicor unit. The USB interface runs off the DEQX supply. The DEQX, for its part, is plugged into my super-AC-conditioner supply. So the computer supply merely operates to shuttle data down the USB line and does not power the USB --> I2S converter.

Occam · « **Reply #4 on:** 24 Sep 2006, 05:43 pm »

Wrt the differences between the respective USB sources, all things certainly aren't equal. You simply don't know whether the percieved differences are ps related or specific to the actual USB hardware/software drivers on your respective computers.
If you took something like a parallel output to USB card you might find they both equally suck, reflecting the add on card and not the differences between the PSs. Then again, you might not.

serengetiplains · « **Reply #5 on:** 24 Sep 2006, 05:50 pm »

Drivers and software are identical. The laptop likely has different hardware outputting the USB signal.

texendo · « **Reply #6 on:** 24 Sep 2006, 06:05 pm »

I can't offer any experience with desktop vs. pc, but I can definitely comment on the power supply to a computer "simply shuttling data" to an outboard dac.

Vinnie from RedWine made a Reali-T for me that had an onboard usb powered dac. The dac wasn't of the highest quality in and of itself, but Vinnie's battery powering it and the fact that there were essentially no ic's (the tripath chip and dac were on the same pcb) made it into a rather spiffy piece of low cost audio equipment. Vinnie even hard-wired a usb cable to it for me, and the power lead was disabled (there was no power supply coming from the computer; all battery). My source was an iMac G4 800 running Apple Lossless iTunes. Pierre Spray (from MapleShade Records) had told me of the tweaks he had done to his own computer system, so I tried it for myself even though I didn't think it would make much of a difference.

I bought an IEC adapter for a VH Audio Flavor 4 power cord I had and plugged it into my iMac. I also upgraded the wall outlet to a Furutech model. The difference in sound was anything but subtle. There was much more dynamic slam to the presentation. It honestly seemed as if there may have even been more gain, although I don't see how this could be possible. Bottom line, it made a big positive difference.

So yes, even if your "computer transport" is merely producing 1's and 0's via usb (even with the power lead severed), the power supply still makes a very noticeable difference. This makes absolutely no sense to me, and I have no explanation for it, but the result was clearly evident. It's no surprise to me that your battery powered laptop sounded better than your ac powered desktop.

serengetiplains · « **Reply #7 on:** 24 Sep 2006, 08:00 pm »

Tex, I'm not surprised by your observations.

Wow, what interesting, far-from-subtle differences between the two computers in my setup. Listening to The K&D Sessions, Kruder & Dorfmeister, the song "Gotta Jazz" begins with a synth sound to which is added a cymbal and bass. The bass begins with an echo-like staccato blurp, and the blurp trails the bass line a few times. That blurp is almost entirely missing --- absent, gone --- on the desktop. A difference of that kind and magnitude suggests, at least to my mind, that the digital data is somehow subtly different, and rendered so in the ripping process (all ripping done to wav with error correction on). If the difference attributes purely to electrical and/or hardware, it suggests significant room for improvement on either or both those fronts.

serengetiplains · « **Reply #8 on:** 24 Sep 2006, 08:57 pm »

Well, a handful of CDs is enough listening for my ears. The laptop wins hands down. Go figure.

Joules · « **Reply #9 on:** 24 Sep 2006, 09:05 pm »

Does this mean we need linear power supplies for our pc's, now?

Doublej · « **Reply #10 on:** 24 Sep 2006, 09:44 pm »

Being naive in this area, is the USB DAC buffering and reclocking the signal? If it doesn't do both it makes senes to me that different computers (think different transports) would yield different sonic results.

If, however, the DAC is buffering the data coming in and generating its own clock then I am miffed as to why the two computers would sound different.

JDUBS · « **Reply #11 on:** 24 Sep 2006, 10:07 pm »

To isolate the USB signal, this is something I'm considering:

http://www.opticis.com/product_5.htm

It runs on a 5v. I would think that using this with either battery power or a nice linear power supply would be ideal.

-Jim

texendo · « **Reply #12 on:** 24 Sep 2006, 11:29 pm »

Quote from: JDUBS on 24 Sep 2006, 10:07 pm

To isolate the USB signal, this is something I'm considering:

http://www.opticis.com/product_5.htm

It runs on a 5v. I would think that using this with either battery power or a nice linear power supply would be ideal.

-Jim

Many over at audioasylum.com (PC Audio) have reported success using this application.

audioengr · « **Reply #13 on:** 25 Sep 2006, 05:43 pm »

Another possibility is that the desktop has a grounded power cord, so there is a ground connection between the audio system and the computer. This is not so with the laptop. The ground is floating. I believe this may be the explanation since the clock generated in the Off-Ramp I2S interface does a pretty good job of isolating the USB interface. The USB interface has very slow edges and actually quite rolled-off due to caps at the receiving end doing a LOT of filtering. This is to insure that a 5m cable works properly. This creates a LOT of jitter in the USB interface. Virtually none of this shows up in the I2S interface.

One way to test this theory is to plug both the audio system and the desktop computer into the same AC circuit and then float the ground on the computer power cord.

BTW - I have customers that compared the Opticis to the USB cable with active hub and the cable with hub sounded better. Your results may be different however.

Steve N.

Joules · « **Reply #14 on:** 25 Sep 2006, 11:40 pm »

Do you suppose removing the +5v and gnd contacts on the computer side USB cable plug might help, by perhaps reducing high freq trash form the PC power supply from coupling in to the remaining data lines in the USB cable?

serengetiplains · « **Reply #15 on:** 26 Sep 2006, 03:15 am »

Steve, I'll use your test idea tomorrow. By float the ground do you mean lift (eliminate) it?

Does your I2S interface connect the computer's ground to the audio system's ground?

BTW, my audio system runs from a dedicated line as follows: 120V + 120V --> step down Topaz isolation transformer --> Elgar AC conditioner(s) --> various audio components. This dedicated line has no connection with the "neutral" household line. The Elgar conditioners, for their part, establish a new connection between neutral and ground on their output.

audioengr · « **Reply #16 on:** 26 Sep 2006, 04:30 pm »

Quote from: Joules on 25 Sep 2006, 11:40 pm

Do you suppose removing the +5v and gnd contacts on the computer side USB cable plug might help, by perhaps reducing high freq trash form the PC power supply from coupling in to the remaining data lines in the USB cable?

You cannot remove the USB Ground. It is part of the circuit. The 5V line is already isolated (disconnected).

Steve N.

audioengr · « **Reply #17 on:** 26 Sep 2006, 04:32 pm »

Quote from: serengetiplains on 26 Sep 2006, 03:15 am

Steve, I'll use your test idea tomorrow. By float the ground do you mean lift (eliminate) it?

Does your I2S interface connect the computer's ground to the audio system's ground?

BTW, my audio system runs from a dedicated line as follows: 120V + 120V --> step down Topaz isolation transformer --> Elgar AC conditioner(s) --> various audio components. This dedicated line has no connection with the "neutral" household line. The Elgar conditioners, for their part, establish a new connection between neutral and ground on their output.

Yes, disconnect the ground pin on the computer power cord - use a cheater plug.

Yes, the I2S interface connects the computer ground to the audio system ground.

If your system islolates the AC grounds, then there should be no ground loops.

Steve N.

Joules · « **Reply #18 on:** 26 Sep 2006, 06:00 pm »

"You cannot remove the USB Ground. It is part of the circuit. The 5V line is already isolated (disconnected)."

Disconnected where ?

serengetiplains · « **Reply #19 on:** 26 Sep 2006, 07:21 pm »

Ok, I spent all morning (there goes my day job) listening to various permutations of my system. I first lifted the ground on my desktop, reinserted the ground, then lifted it again. No difference. The laptop was clearly better in this setup.

I then plugged the computer into a standalone Elgar AC conditioner that feeds my DEQX. *Nice* change for the better. In this setup, the desktop clearly bettered the laptop, which sounded a touch soft and diffuse, lacking bass drive by comparison, with overall less distinct sound, more hashy treble, and rendering low level information more difficult to hear. I was able to identify these differences very reliably in three tracks various parts and all of which I repeated over and over and over.

I then plugged the DEQX/desktop Elgar into my larger Elgar setup (this latter = isolation TX --> two 3006s in series, the DEQX Elgar being, then, the third in this series) and the desktop sound improved further. I am uncertain how much of this latter improvement derived from feeding cleaner power to the DEQX. My sense was that the desktop setup improved more than the laptop setup, which both share the DEQX of course.

I then ripped two of the three songs I was focussing my attention on, wondering if clean AC would affect sonics if just from the ripping process (my entire collection was ripped using dirty AC). I thought I could detect a very small difference for the better on the newly ripped material, but putting it that way tells you how large the difference could in any event have been.

By "clearly better" in the paragraphs above, I'm referring to a difference in the range of, and not too dissimilar in character to, but greater than the difference between a single polypropylene and a teflon coupling capacitor.