Build a dedicated Audio PC

[nige2000]

Great work, Nige - must get to hear it sometime.

My suggestion for your final config
- Use 3 or 4 pole switch instead of multiple relays - try these NKK switches with max 10milliohm contact resistance http://ie.mouser.com/ProductDetail/NKK- ... wI2AA2%2fK
- You could also put a switch between battery & Pico

im sure it will be at a meet at some stage

Switches i can see it now


be like an aircraft cockpit if i get the sequence wrong there will be smoke

was hoping for something a little more foolproof

Haha - it's retro :)

No, I mean 3 pole or 4 pole switches i.e one toggle turns on/off 3 or 4 circuits! The correct terminology for them is 3 pole, single throw or 3PST or 4PST - have a look on good ol' wiki http://en.wikipedia.org/wiki/Switch

didn't need to google 4 pole switch (was just a bit of banter)
ive a few 2 poles here i can experiment with

thing is i think it comes down to this:


or



im leaning towards this but much testing is required to insure the placement of the switch does not interfere with the maximum SQ gain

i have tested the 3.3v line with this circuit with a noticeable improvement,
whether SQ would be better again without the switch still needs to be tested

[jkeny]

....didn't need to google 4 pole switch (was just a bit of banter)

Ah, OK

ive a few 2 poles here i can experiment with

thing is i think it comes down to this:


or



im leaning towards this but much testing is required to insure the placement of the switch does not interfere with the maximum SQ gain

i have tested the 3.3v line with this circuit with a noticeable improvement,
whether SQ would be better again without the switch still needs to be tested

Yes, I think the second diagram is more elegant.
Do you mean your experiments in the past showed a noticeable sonic hit to having a switch in the 3.3V line compared to no switch?
Although these switches are low resistance 10milliohm they not behave ideally at high frequencies. But then neither would a relay be guaranteed either as it is just a switch too.

[nige2000]

I meant I have not as yet tested the 3.3 line without the switch between the 3.3 line and the battery

Do yo think 3.3 wire from pico and the 3.3 wire to mobo socket would be a better stabilising effect soldered direct to battery terminal/ends rather than having a switch in between

You might be able to guess with I did here for testing which was a positive result but maybe direct would be better

[jkeny]

I meant I have not as yet tested the 3.3 line without the switch between the 3.3 line and the battery

Do yo think 3.3 wire from pico and the 3.3 wire to mobo socket would be a better stabilising effect soldered direct to battery terminal/ends rather than having a switch in between

You might be able to guess with I did here for testing which was a positive result but maybe direct would be better

I doubt that switch or no switch will make any difference with the length of wires you are using. High speed really should have very short wires (even for PS) but I wouldn't worry about it too much. I know this is not high-speed signalling but if the current draw on the PS is in high speed pulses then somewhat of the same issue apply. You would also be wise to look at RFI emanating from wires delivering high-speed pulses at these voltages.

I'm sure DaveF would be able to elucidate some more?

[nige2000]

I meant I have not as yet tested the 3.3 line without the switch between the 3.3 line and the battery

Do yo think 3.3 wire from pico and the 3.3 wire to mobo socket would be a better stabilising effect soldered direct to battery terminal/ends rather than having a switch in between

You might be able to guess with I did here for testing which was a positive result but maybe direct would be better

I doubt that switch or no switch will make any difference with the length of wires you are using. High speed really should have very short wires (even for PS) but I wouldn't worry about it too much. I know this is not high-speed signalling but if the current draw on the PS is in high speed pulses then somewhat of the same issue apply. You would also be wise to look at RFI emanating from wires delivering high-speed pulses at these voltages.

I'm sure DaveF would be able to elucidate some more?

Ah now no need to set the dogs on me!!!!

As you can see from photo RFI was a top priority :)
any suggestions welcome
bottom line sound better with this system than without
the rfi will be a battle for another day

might go ahead with the 4 pole switch for stealth mode for my jet fighter!!!

[Sligolad]

A report and a few thoughts on how the Audio PC route is progressing based on recent events.

Another interesting night at Tony’s pushing the boundaries of what can be achieved from PC based playback, and it was also partly a rematch against Simons wonderful Wadia 581 which easily thrashed the PC based contenders the last time around.
This time the Meitner MA1 made an appearance to help the cause but as with all DAC’s they are only as good as the source feeding them.

The competition on the PC based side was as follows I think: Tony’s Zuma build with batteries going in to his Ciunas DAC, Nigel’s PC build with batteries/linear supplies going in to his Ciunas DAC and my Zuma build with linear supplies going in to the Meitner DAC.

The usual PC messing started trying to make sure everything was connected but at least Tony’s was set and ready to go, mine was set and ready to go or so I thought until I could not get mine to run with all the same problems as I had the last time at Tony’s. Luckily this time I was convinced my build was rock solid and I began to point the finger at Tony’s mains supply, pennies began to drop and Tony mentioned he had lighting issues and an issue had been raised with the local ESB.

Well Tony fortunately has an power regenerator which I use myself for regenerating a clean mains supply and when I plugged my PC in to this supply everything worked straight off, just goes to show you can never take your mains supply for granted!!

Finally Nigel eventually got all his kit up and running after a more test bench like build on the floor as only Nigel knows best….a sight to behold for anyone with a technical bent.

I think we had some good ears in the room excluding mine with Ivor, Simon, Nige and Tony pretty much in agreement that the gap between a Ciunas DAC and single audio PC was now much closer that previously to the Wadia listening to some good vocal and jazz tracks. Tony and Nige were running MQN for playback against the Wadia and for builds including DAC coming in at about one sixth the cost of the Wadia they acquitted themselves extraordinaly well.

Moving to my PC and Meitner against the Wadia it was difficult to call on some tracks as both had their positives and I will leave it to the guys to give their thoughts separately. I felt the Meitner had a more musical and organic sound but saying that I had just added a new USD and SATA cable the previous evening and I expect these to improve over the coming days.

Bottom line was we could all live easily with the sound from both the Wadia and Meitner PC combo and not forgetting the Ciunas PC combo but the big take away for me was that all the Audio PC’s there had all it takes to deliver an extremely high quality output for any DAC you partner them with. My firm belief after this meet is that it is now possible to put together a Wadia beating PC/DAC based system for less than a third the cost of a top end CD Player. It is just a matter of tidying up the package and selecting the right DAC to suit your tastes for a serious front end to your system.

Finally on the PC’s we were using mainly Server 2012 R2 with no networking and me using mainly Jplay Mini 5.1 for playback and some tracks with MQN for reference. There was very little in it between MQN and Jplay Mini but I will continue to monitor progress with MQN.

Thanks to Tony and Anne again for the genuine hospitality and of course the usual Tea/Coffee and biscuits in which they excel!!

[Diapason]

Quite right on all fronts Pearse, and thanks indeed to Tony for hosting and to Anne for putting up with us!

Moving to my PC and Meitner against the Wadia it was difficult to call on some tracks as both had their positives and I will leave it to the guys to give their thoughts separately. I felt the Meitner had a more musical and organic sound but saying that I had just added a new USD and SATA cable the previous evening and I expect these to improve over the coming days.

I would have to agree that the Zuma/Meitner combo came out ahead of the Wadia, giving a really lifelike sound with slightly more solid bass, lovely timbre, and just an overall sense of "rightness" that was very impressive. I'm looking forward to hearing it in my own system at some stage to find out what it can do in a more familiar set-up. Well, maybe it would be better if I didn't hear that, but I need to know!

As Pearse also said, the gap in performance between the Wadia and PC/Ciunas combo seems to have narrowed considerably. In fact it was definitely one of those evenings where the overall sense was "it's all good". Some differences in presentation sure, but you wouldn't be feeling cheated with any setup that we tried. Well, as long as I didn't insist on playing that Patricia Barber track again...

Thanks again Tony.

[Ivor]

I concur with all the above. I'm also interested in hearing a PC source in my system but, like Si, am worried it might lead me down the path of no return!
Thanks as always to Tony and family for their hospitality and confectionary. Having your house infested with geeks isn't easy, it's very much appreciated.

[nige2000]

Thanks to Tony and Anne for hosting

great to hear the wadia, stiffer competition than i imagined and some what prettier than my tangled looms of wires and batteries
(but it was built for testing SQ benefit of individual components)

Wadia seemed marginally but noticeability better than either pc/ciunas (no pang),
the real competition for me seemed to be between the Wadia and the Panged Super Zuma/meitner combo, which i thought edged out the wadia overall but some aspects seemed better on the wadia like a slightly lower noise floor (might have been just the tracks we played)

After we did some component swapping which i tried a differently set server R2, which we all agreed sounded like a completely different presentation,
we tried the meitner with mine which closed the gap a bit more (might have a bit to do with buffer settings) and then the panged super zuma with the ciunas ( we used the same OS stick for this to remove OS influence

Moral of the story the pang gear seems to add something to the SQ

We should have done a pang usb card swap to see what that brings, next time?

we might need to get studio monitors for SQ testing as its getting tighter every time

All in all theres allot of improvement in pc audio this year

maybe close to a stage where we can down tools and hear whats beyond 10 seconds of a test track

Although the 10 second rule would have been nice for Patricia Barber,
don't think i heard half of that track because i was just waiting for it to end

Power was a little more than suspect, luckily i came prepared with batteries

(electricity supply like that is really hard on appliances, it can break stuff )

anyway

To pang or not to pang?

[Ken Moreland]

Sounds like a good evening was had by all, I heard that stragglers were kicked out in the early hours.
The Pang USB card and Meitner must be doing something well, especially as Pearse's set-up was using JplayMini 5.1. Last session the comparison between Jplay streaming and MQN on identical pc's favoured MQN and MQN has come on much more in the last month. I tried JplayMini 5.2beta on KS single pc but MQN was much better although when I compared Jplay 5.2beta in Wasapi it was closer but still MQN was better.
KM