Tír Na HiFi

jkeny wrote:
It seems to me that the better sound quality that everybody is getting with all these mods fanless, SSD, SATA cables, etc. are really got to do with stabilising the main PS as much as possible.

that would make sense but does the vibrations of these fans not have an affect also?

As I have been saying for a while now - we are dealing with a very complex device (PC) & trying to simplify it - really the ultimate realisation of this is a much simpler base system i.e ARM CPU & simple surrounding devices.

are you think of mobile hardware feeding into a dac because yould still need a operating system and jplay like software

nige2000 wrote: that would make sense but does the vibrations of these fans not have an affect also?

Don't know? I think most things come down to the stability & cleanliness of the PS in matters of low signal level audio. Digital audio should really be treated somewhat like phono audio systems - the signal levels are probably as low & any disturbance on these seems to be an issue but shhh, don't say this to the digital brigade who think that digital signalling is somehow magical & that a bit is a bit is a bit :) I believe our experiments further prove the flawed premise of this bit brigade's viewpoint

As I have been saying for a while now - we are dealing with a very complex device (PC) & trying to simplify it - really the ultimate realisation of this is a much simpler base system i.e ARM CPU & simple surrounding devices.

are you think of mobile hardware feeding into a dac because yould still need a operating system and jplay like software

Well, I don't want to do the Jplay guys down as they are doing a fantastic, trojan job but the foundations will begin to limit how far they can take this i.e the complexity of the PC.
Some efforts were already made in this KISS approach with SD card reader audio players & early efforts sounded pretty good but not as good as what we hear currently with Jplay on PC. It might be time to revisit some of these SD card players? They are limited by user interface (but we are not too bothered with this :)) & storage limitations but SD card storage has multiplied since then.

jkeny wrote:

nige2000 wrote: that would make sense but does the vibrations of these fans not have an affect also?

Don't know? I think most things come down to the stability & cleanliness of the PS in matters of low signal level audio. Digital audio should really be treated somewhat like phono audio systems - the signal levels are probably as low & any disturbance on these seems to be an issue but shhh, don't say this to the digital brigade who think that digital signalling is somehow magical & that a bit is a bit is a bit :) I believe our experiments further prove the flawed premise of this bit brigade's viewpoint

i thought a digital cable been a cable was carrying more than bits such as noise which works its way round the system
a bit maybe a bit but even a usb cable carries more than digital.
i consider myself a skeptic but i can perceive how a digital signal can get poluted

nige2000 wrote: i thought a digital cable been a cable was carrying more than bits such as noise which works its way round the system
a bit maybe a bit but even a usb cable carries more than digital.
i consider myself a skeptic but i can perceive how a digital signal can get poluted

Yes, but the theory of digital is that it is relatively immune to noise when operating in the digital domain i.e once the bits are communicated correctly between devices then the information is correctly transmitted. This works & is proven to be correct or else our computers would be useless & communication systems would be unreliable.

However, when we cross into the analogue world of audio, noise is very much an issue - everybody understands this. The problem is that the place where this digital to audio conversion takes place is at the DAC device so the theory goes that it is only in this last step that noise matters. And here's the really sticky bit that's difficult to answer - how can changing the PS on a USB stick change the sound at the DAC. The USB stick is so far removed from the final digital to analogue conversion that it seems to be impossible that it could have an effect. The answer lies in looking at this problem on a system-wide basis - so any PS fluctuations on the USb stick can cause system-wide perturbations in PS & therefore have an effect at the final stage DAC stage. This is why the SOTM USB card seems to be favoured - because it can be used to provide clean PS at the final connection to the DAC - the USB connection. But again even doing this does not seem to be enough - there still seem to be ways that the final audio quality is effected.

I imagine that some real deep system thinking is needed to tease out the variables here BUT computer audio is so complex a system with so many variables involved that this level of system thinking is nigh on impossible. Hence my call for a simpler platform for all this

jkeny wrote: BTW, playing a song from the Win-to-go (WTG) stick seems to me to be not nearly as expansive as playing the same song from an SSD.
I put a song on the WTG stick & no internal SSD - booted from WTG & played it.
Shutdown & then put the SSD drive back in but booted from WTG & played same song from SSD. Result = sounds more expansive. When I then played this same song from WTG some loss of sound stage, I feel?

But when I boot from SSD & play the same song, there's a little bit of sibilance creeps in so booting from WTG seems the Way To Go (WTG) :).

All done with Ultrasize (US) = 200 (I need to check this again). I'll experiment further with reducing US

tried the experiment with the wtg usb3 and playing from ssd and i think your right it is a bit more expansive than storage and OS on the WTG

can u give me instructions how to give the WTG stick a clean power supply, id like to try it, i probably have enough stuff lying about and im not afraid of the soldering iron!

Finally found that post I mentioned where the guy modified his core i3 desktop PC to 8.5W at idle.http://ssj3gohan.tweakblogs.net/blog/61 ... r-(english

It's an interesting read - he even talks about the PicoPSU a little

Nige, I had one I made a while ago - what I used was a male USB A type plug & a female type plug which I soldered back to back but left the 5V Vbus connection unsoldered. I soldered a red wire which will be the new 5V supply to the female Vbus pin connector (it's either pin 1 or pin 4, I can't remember) & a black wire to the ground pin. This pic may help
Alternatively, you could take a male to female USB extender cable & open it up & splice into the ground wire; cut the red Vbus wire & supply your own 5V supply?

Something like this is a neater off-the-shelf solution http://www.streakwave.com/Itemdesc.asp?ic=5VUSB
Or this with it's own PS http://www.aqvox.de/usb-power_en.html

I too have been thinking about the power supplies ever since my experiment on the CPU fan.

I wonder if it is possible that there are 2 orders of effects due to power:

1: Instances where power is routed through the motherboard - eg USB, fans, CPU power, onboard LEDs, video chips and so on

2: Instances where power is derived from the same basic supply, but not routed through the motherboard eg SATA drives, separately powered video cards

My hypothesis:

magnitude of current demand x frequency = measure of effect

Does that sound reasonable or likely to any of you?


Fran

Fran wrote:I too have been thinking about the power supplies ever since my experiment on the CPU fan.

I wonder if it is possible that there are 2 orders of effects due to power:

1: Instances where power is routed through the motherboard - eg USB, fans, CPU power, onboard LEDs, video chips and so on

2: Instances where power is derived from the same basic supply, but not routed through the motherboard eg SATA drives, separately powered video cards

My hypothesis:

magnitude of current demand x frequency = measure of effect

Does that sound reasonable or likely to any of you?


Fran

I'd agree with your final statement I X F = "magnitude of effect" which might mean that some of this effect is to do with RFI?

But I'm not sure I understand what you mean by points 1 & 2?
Edit: Ah, maybe the penny dropped - the first order effect is to treat any devices that suck power from the motherboard
but I'm not sure what the second order effect is? Does SATA not suck power from m/b?

BTW, I hope to try external power to my SATA drive & see how this compares to external power to my WTG USB

Points 1 & 2:

some things are supplied power directly via the motherboard, most likely through DC-DC switching/regs from the 12V line - even the guy with the 7.5W pc mentions this, in his graphs at the end he shows CPU, platform controller hub (which I think is the controller for the other chips on the motherboard) and the RAM as the major power users. It is conceivable that when there is a large current draw by the CPU, say, then the other components are affected. If you take your USB experiment, this sort of thing could be why you are hearing differences. USB 3 supposedly has better power management, and can handle higher total draw (900mA vs 500mA) so that also may be a reason why you are hearing differences

Outside the motherboard, but coming from the picoPSU, you have the SATA drives. Note that many users suggest that SSD sounds better. Is this because they draw much less current than a spinning drive? These might have a different effect on sound quality as they are not routed through the motherboard itself?

On my test bench, the fan draws ~120mA. Thats more than some SSDs at idle - its not surprising that the fan affects sound, just as the SSD does.


Don't know if I'm making any sense to anyone other than me!! but it comes down to this:

Huge part of SQ = power supply

Other huge part as evidenced by jplay = memory management.

The jplay boys have done their bit, now the rest is up to us.


Fran

Ah, right, I'm with you!
Yes, reading that guys page shows how many inefficient power conversions there are on a motherboard & how much proper PS handling is ignored by PC designers. I guess it's because they don't have to bother too much with the quality of the PS because it's digital :)

Maybe the move towards lower power consumption might focus their thinking somewhat in this area & as a result we will get better designed motherboards?

The fact that he was able to cut down a 20W? consumption at idle to 7W shows the inefficiencies in current motherboard PS.

As you say "the rest is up to us" & this thread is turning out to be a great resource for this venture!