Build a dedicated Audio PC

[tony]

bit is bit. bit is perfect
repeat after me....
bit is bit. bit is perfect
fair is foul and foul is fair

beat already made billions with digital sound.
http://www.usatoday.com/story/tech/2014 ... s/8868913/
are you sure analog sounds better?

That's worth a new thread jc
Bad move on apples behalf
Kinda makes me wanna sell my mac book

Oh dear mac books with beats audio label on it
Kinda feel sick now

Could be the starting lyrics to a new class A rap song from China?

[jesuscheung]

G400 has regain my place of a audiophile mouse
cannot install setpoint software or that gaming software thing as driver. too much useless junk

need to manual install the mx512 driver for a G400 mouse.

mx512 driver > G400 driver > Logitech HID wheel optical driver > Logitech HID optical driver > MS mouse driver

in general
one should try Logitech HID wheel optical driver.

i extracted the Logitech mouse driver,
download here
https://drive.google.com/?usp=folder&au ... jc1QzcxeTg

[Sligolad]

Got a nice improvement this past few days on my audio PC by finally adding all grounds to a common connection array.
Up until now I had a common connection shared with the Mobo grounds for most of the supplies bar the USB supply and the SSD for Music. I am assuming adding the USB card ground with all the others has been the source of the improvement.
So to sum up I have 2 x 5v iUSB supplies, 2 x 5v Anker batteries, 1 x 3.3v Teddy Pardo Linear Supply, 1 x 5v Teddy Pardo Linear Supply, 1 x 12v Teddy Pardo Linear Supply all sharing a common ground array with the Mobo Ground and it sounds excellent.
I would recommend it to the more adventurous among us who have direct feeds to cards and SSD's to connect all the DC grounds together and see if you hear a difference.....if you blow it all up then sure it was just a crazy idea anyway :-)))

[nige2000]

what way had you routed the gnds before?
all individually back to their ps?

most gnds will merge at the motherboard only ssd gnd wont,
maybe merging at the motherboard level exacerbates the noise

how does the ankers compare to the iusb supplies?

did you get comparing the b85 to the z87?

still having teething problems with the multiple clocks but i still have faith
hope to have something special out of it

[Sligolad]

Hi Nige, I always had all Teddy Pardo linears initially on a common ground with the Mobo with the switch on these in the Zuma.
I added the OS ssd and the clock PS supplies a while back and then recently the music ssd and JCat USB card PS supplies to this same ground system.
I can only suppose the more there are from all the various types of clean supplies and voltages that they seem to act to smooth any weakness in any of the supplies....it is all too weird and strange for me to work out why but it does appear better than when I had individual supplies to each piece in the chain which goes against logic but logic for me went out the door a long time ago when I started mucking about with DC voltages.
There is still one bit of weirdness I have not figured out yet but one of these days I will try a bit harder to figure it out....
I found recently that the 12v positive connection (yellow) to the ATX connector seems to act as an energiser for parts of the Mobo when all grounds are disconnected, almost as if the 3.3v positive from another linear gets flowing once it sees a 12v positive switched in on the board even when each do not have a ground connected anywhere.

I pulled the usb cable to the DAC in case it was finding a ground path through the USB cable and still there was a 3.3v lighting an LED on a board on the PCI rails, the Zuma case has no other route to ground so all a bit strange but I am sure I am missing something.

So now when I hit the shut down button on the Zuma and windows exits I hit the ground switch which disconnects all grounds from linears and batteries but I also switch off the power on the 12v Linear at the unit itself so no power is trickling anywhere through the Mobo.
Next time we get together I will explain and demonstrate better as I am sure this is difficult reading.

Never could really tell any difference between the Ankers and the iUSBs but the iUSB hold 5.00v bang on when you put a meter on them.

Have not seen any difference or heard any diffierence I can speak of between the 2 MSI boards, looks like they just added a heatsink and maybe some slightly better regs for overclocking but other than that the Z looks the same as the B.

Good luck with the clocks as it appears to be the final frontier....poor Ruben on JPlay forum is having a tough time with clocks, hope he gets sorted soon.

Thinking I am at the end game for my audio PC now, just waiting for the Paul Hynes PS and that will be it for this year :-))
Cheers, Pearse.

[jkeny]

guys, it's always good practise to bring grounds together at a single point - a star ground configuration.
This arrangement is designed to avoid having different ground potentials between supplies - i.e current flows between ground supplies.
The other important concept with grounds is to try to avoid PS ground currents polluting signal ground currents - in the case of a PC it's complicated but probably best to think in terms of what are the supplies with big current draws & what are the supplies with small current draws - also consider what supplies have fast switching currents. So try to keep these fast switching heavy ground currents (SSD, mobo, RAM?) from polluting the smaller, slower changing ground currents (USB, clocks).

I would try to ensure that the ground return loops were as tight as possible for each of these separate PSes

[nige2000]

All my gnds are merged back to the batteries
But I run extra shorter gnd returns from sensitive devices such as the usb card, usb o's and ssd etc
And all clocks will have too when I'm done

[jkeny]

All my gnds are merged back to the batteries

Yes but it may be good practise to interconnect all the grounds of the batteries together at some star ground point - otherwise you have floating grounds which could have voltage differences between them?

But I run extra shorter gnd returns from sensitive devices such as the usb card, usb o's and ssd etc
And all clocks will have too when I'm done

Yes, with individual devices like isolated cards & clocks you can design in short ground loops - it becomes much harder for areas on the Mobo - hopefully there is a good ground plane which connections can be made to. If making connections to this ground plane from multiple PSes, do it all at one point on the ground plane i.e all your PS ground wires should connect at this point (including the ones from the USB cards, SSD, clocks, etc.).

So you now have individual devices with their own local ground planes that have tight ground loops on it & these ground planes are then each connected back to the mobo ground point through a ground wire.

[Sligolad]

Never was one with much time for theory John as I always wanted to confirm for myself but it appears on this one the theory is proven out through practice.....I was the one at the back of the classroom yawning and wanting to go out and make it or break it before I would believe what I was told!!

DC is a funny animal though and I am still not finished with poking at it for a while longer to see what it has got in store for me.
Hard to see how to keep these fast switching heavy ground currents from polluting the smaller, slower changing ground currents when the ATX connector seems to allow all the grounds on these connectors to spread all over the board.

I should pay more attention in future though :-)

[nige2000]

All my gnds are merged back to the batteries

Yes but it may be good practise to interconnect all the grounds of the batteries together at some star ground point - otherwise you have floating grounds which could have voltage differences between them?

I've got two star type gnds one pre batteries and one after