XXHighEnd

---

Edit :
The subject in this topic is obsolete because it was about the original version of the 24/384 NOS1 which by now has been replaced with the 24/768 NOS1 async USB version. With the latter this problem can not occur.

Peter

---

Because there is an NOS1 customer with problems in the PCI area using an AMD processor, I just wondered whether anyone is using an AMD processor ?

If so, what is the model/type ?

Thanks in advance for your responses,
Peter

AMD's Athlon 64 X2 3800+, but sadly no NOS1 ......and after a failure in a bios flash, now no system at all!

I guess it is either AMD CPU or AMD chipset on Mobo but my friend's Phasure works just fine on my Intel computer where it does not behave
properly on his AMD machine. I think he went out to buy new MoBo and Intel CPU today :)

I think there is more, and I guess it is not AMD related at all. But me and your friend are working on it.

But thank you for helping out !
Peter

Hello Everyone,

It is me that Peter is helping. Peter is so helpful that we have exchanged about 50-60 emails on this subject to date. One will never find any manufacturer more thoughtful than him.

Here is the situation that I explained to Gigabyte technical support in Taiwan :

Processor : Two configurations were tested:
     AMD Phenom II x4 quad core with Gigabyte mobo 880GA UD 5H
     Intel i7 2600 with Gigabyte mobo P67A UD5
General Hardware : ATI HD5770 graphic card, 2 x HHDs, 1 x SSD, 12 GB DDR 3 dual channel RAM, 650W PSU, water cooling.
Dedicated Hardware : Juli@ studio sound card :   http://www.esi-audio.com/products/julia/      external board version built into the dedicated hiend Digital to Analogue Converter - NOS1 DAC  :   http://www.phasure.com/index.php?topic=1560.0  External Juli@ requires an adapter card to be inserted into PCIe x1 slot in mobo. This adapter is linked with main Juli@ sound card in the DAC via standard DVI cable.
OS : W7 SP1
NOS1 DAC : Is perfectly fine. My NOS1 has been tested with at least 3 PCs flawlessly. One on very old AMD platform. Another is with Gigabyte X58 mobo and i7 950 platform.

Problems :
First configuration (AMD) : Mobo is able to detect the Juli@ sound card. Sound card driver was installed without problem. However, when running the dedicated music player XX Highend, extreme distortion with slow motion sound is encountered. Data run through the bus is way too slow. With unknown reason, I think I fried this mobo during the test. However, I later sent this mobo to Gigabyte repair center for a thorough diagnostic test. This borad is perfectly fine. No problem at all. This AMD system was configured to run dual boots for both W7 and Vista 64 bit.


Second Configuration (Intel) : DAC manufacture recommends to switch the CPU to Intel and i7 2600 is chosen due to its performance. P67A UD 5 was chosen. Finding and problem are as follows:
      1. Reinstall W7 OS and all required drivers except Juli@ driver.
      2. HD 5770 in PCIe x16 slot
      3. Juli@ adaptor in either PCIe x1 or x8 or x4 slots but DVI cable not connected to the NOS1 DAC.
4.  System boot normally and PC worked well. System did not detect sound card since only adapter is installed.
5.  When DVI cable is connected to both the adapter end and the NOS1 DAC end, system did not boot even with the DAC powered off. Note that the DVI cable is perfect and the ground loop problem is not relevant. Then boot misbehavior was that mobo run 2 second and off and then on for 2 sec and off and so on .........
6. With DVI cable connected at both end like #5 above, but this time the DAC was powered ON. The PC did not respond at all when PC power switch was pressed (even with the press on onboard power button).  The PC went like dead.
7. Flashing the BIOS with F6 did not help. Playing around with many BIOS setting by disenabling things randomly did not help. Switch back to factory default gave same results.
8. Try using another graphic card which run on the PC that works perfectly with NOS1 on P67A UD 5 mobo. Same unsuccessful result.
9. Ask Gigabyte dealer to exchange for new P67A UD 5 mobo. Swap the mobo and test again. Same result. Did not work.

Cuprit : suspect that it is a PCIe Bridge compatibility issue at the electrical level. Something like : the bridge is seen while it's powered off, but from there on it goes wrong.

Solution : Gigabyte gave me 4 other mobos on loan ranging from basic H67 chipsets to X58 hiend gamer one. I have done some test already but not all. We share my final finding shortly.

And my guess for now : the i7/2600 isn't compatible somewhere ...

Hello Everyone,

FIRST ROUND OF TESTING

With the courtesy of Gigabyte RMA office in Bangkok, I have the luxury of playing around with some samples and processor. They lend me the followings:

1. X58A UD9 : this is gamer grade outdated mobo running on Intel X58 chipset with very large form factor. You need 1366 socket CPU and  big PC case.
2. G1 Sniper : newly launched at CES show this Jan. Again running under X58 chipset with fancy onboard audio and LAN processors.
3. 890 GPA UD3H : AMD chipset with onboard GPU capability
4. H67M UD2H : new 1155 socket for CPU and H67 chipset. This is a basic small form factor mobo with onboard GPU.
5. Intel i5 2400 processor
6. Intel i7 930 processor
7. RAM (to test if I have RAM issue)
()
FINDING

The testings were conducted under the same general hardware configuration as mentioned in my previous post. Same OS ie W7 SP1. No other SW application. All hardware drivers were reinstalled fresh after the mobo was replaced. XXHE version 0.9Z-4-1.

1366 Socket

1. G1 Sniper + i7 930 worked perfectly fine without any change in BIOS from default setting. Installation of Juli@ driver was easy. But with one trick that is with the first detection of Juli@ device after installation of the 1095 extension card in PCIe slot, you will not be able to install the driver right away even you follow Peter's manual. The system will keep on giving you an unsuccessful dialog box. What you need to do is to reboot one time. And this time the system will detect and ask for the driver location. You browse through the location where you unzip the driver and check "search through folder". After you click next then the driver will be installed. This trick  is always true for X58 chipset. I don't have experience with other mobos.

G1 Sniper is a fast mobo that supports 32 lanes, meaning both PCI-E slots can run at x16, providing a sigificantly higher threshold for the bandwidth. A dedicated audio solution and LAN network processing unit are not only designed in order to free the third PCI-E slot for an additonal card, but also to reduce the workload of the central processor in order to improve overall performance. This is definitely ideal for game freak ! Oops !

2. X58A UD9 : Cannot test as my case is for standard ATX form factor. However, it should work perfectly fine like G1 Sniper since it run on the same X58 chipset.

1155 Socket

3. H67M UD2H + i7 2600 (and i5 2400) : Failed. With NOS1 connected as earlier posted, the PC did not pass through the initial boot and hanged there forever.

4. I have no RAM issue and this problem is not related to RAM.

I will share my second test on AMD processor later.

Cheers !
Praphan

Thank you for all your efforts Praphan, and your local Gigabyte dealer for giving the opportunity of course.

Is there any time left for listening to music ?
:)

Hi Peter,

To be honest , no !

But after finishing playing with all these hardware toys, I am sure I will enjoy the beauty of music I have never heard before. I ran NOS 1 on my son's outdated AMD platform, heard crystal clear sound. Bravo !

Praphan

Wow.... I did not realize all the stuff you went through since we last talked.
Just adding one bit, Praphan tried Phasure NOS1 DAC on my computer with Gigabyte GA-X58A-UD7 rev.2 with
no problem with installation. Very strange...

Hello everyone,

More report.

LAST ROUND TESTING

AM3 socket for AMD processor

The last mobo from Gigabyte I put on my test bench was 890GPA UD3H. This mobo uses more advanced chipset than my 880GA. Both has onboard integrated graphic capability. 890GPA has a better cooling heat sink than that of 880GA too.

With my 880GA mobo, I did not have any problem on interfacing with Juli@ sound card system.  But rather I faced the slowmotion sound.  But with 890GPA chipset, my test PC behaved exactly the same as mobo with 1155 socket. With the NOS1 connected to the interface card via DVI cable, after firing the PC , it hung in the BIOS stage forever without passing through to OS booting stage. Flashing the BIOS and updating all the mobo drivers did not help.

CONCLUSION

1. All the PC hardwares that I put on the test bench work fine without NOS1 in the system. No PC hardware problem in the equation.

2. NOS1 was tested by Peter before shipping. It functions perfectly with X58 chipset both on Sniper and Suteetat's music server. Only the interface card which sits in PCIe slot was brand new when shipped. Can this be a half cuprit.  It is hard to believe since there is only one SMT IC on it and I handled this with great care.

3. I assume that there is no one out there who is a NOS1 users are using P67 or H67 chipsets. This is because I am the only one who report this case. So let assume that my interface card and NOS1 is not partially malfunctioning ie working on X58 chipset but not on the new Intel chipset for 1155 socket.

4. Given no. 3 above, there is a good reason to believe that the problem is due to the firmware incompatibility. But I don't have any idea as to which one that actually cause this incompatibility. The one on Juli@ card or the one on mobo. Well the mobo design is supposed to be universal and the third party device must align with the industry standard.

5. H67 and P67 chipset both have Sata 2 data transfer issue. This goes all the way through PCI bridge as well. Intel just recently launched B3 version chipset that fixes this Sata 2 issue. Gigabyte is willing to lend me the new  B3 board for further testing but I feel exhausted with this project and wish to spend more time listening to quality sound through NOS1.

Well I will end up with G1 Sniper as it looks really cool and fast. It supports DDR3 triple channel so XX will run smoother.

I will end this project here and refocus on listening. My next project will be comparing NOS1 with Berkeley Alpha DAC.

Last but not least, i wish to thanks Philip and Chalee at Gigabyte head office for their support without which this project is not possible.  

Praphan            

Please allow me to adjust a few things, before a "conclusion" is a conslusion I don't concur with.


This latter has happened to a few others (not knowing the mobo brands), and so far they were all solved (by providing that list of BIOS settings to the people concerned). In your case Praphan, and if my memory serves me right, we never got around to finalizing this, because before we could your mobo was fried by extraterrestrial causes.  :censored:


You most probably wanted to say something which is different from how it looks; in any case, of course all was brand new.


I think I know what you are referring to (after by now 70 emails), but in the context of this post I guess nobody will;
What I said about this (and probably what you refer to) is that this card was in the mobo when it fried, so very far away we could think that the interface card may be "half damaged" and now only works in "some" mobos. Not much realistic, but still to keep in mind.


Impossible. With the DAC being off (but connected) your PC (and those particular mobos) won't boot. Of course something will be going on, but no firmware issue inside the DAC.
The interface card, perhaps. But then in a strange way.
In any event (and as told by email by now), the mobo will be physically able to detect the DAC even when the DAC is not under power. I like to opt for something going wrong there. Since it is the interface card passing this through, and since it works in other mobos, the option of the (firmware of the) interface card being the culprit is farther away. That a mobo is not able to deal with this signal properly, seems not a far away option at all.


May this little project indeed have ended here, then I want to thank Praphan for his seemingly endless efforts with testing things, exchanging more motherboards than most of us will ever do in our lives, and the guts to even start working on this all not having a clue what to do or what to do it for, in the process killing a fine system in the first place.
Let me finish by saying that this topic is here -and at this moment- to prevent others to make "mistakes" beyond their fault, which -again for this moment- seems to be about certain Gigabyte mobos somehow not being compatible. This is not about bashing Gigabyte who are in the loop just as well (I saw half Gigabyte Taiwan being copied in our emails), and can only hope that someone sees some light in this strange problem. On this matter, Gigabyte is provided a link to this topic.

Thank you all for now,
Peter

Hi Peter,

I wish to acknowledge and THANK you personally for your time and effort to help me. I must admit that I learnt from and enjoyed with this little project although it was exhaustive.

Anyone who is going to buy new Gigabyte mobo, please let me know. May be I can share some experience with you. If some happens to live in a country where they have a subsidiary like Thailand, you can try asking for a test first. They are very supportive.

Thanks again Peter.

Praphan   

Hello Everyone,

Just to let you guys know that I should be entitled to call myself MOBO GEEK.

Take a look at the mobos that I further tested on my NOS1. Here is the list : 3 Snipers, few X58As , two version of GA 880s, few P67s, ASUS oldie, three CPUs (AMD, i7 970 , i7 2600).

I am now still having some issues which needs Peter's help. I feel like I have a haunted machine and I need exorcist.  He he

Cheers !
Praphan

Wow. Thought you had opened a computerstore... Wishing you all the best. Per

Thanks Per,

After I fixed the issue with the help of Peter, I will share the findings in due course.

Anyone want to buy some of my mobos please let me know. Deep discount may be. But the shipping cost from Thailand and lack of warranty doesn't worth it .

Cheers !

I could use one of the 1366 boads ... :)
I mean, if you really have one spare, we can later talk about how to get it at my place. In the mean time I can promise that somewhat later you will be able to use any of these boards again, :secret: so be careful what you put to eBay or anything ...
:)

Best regards, and hunt down those ghosts !!
Peter

Hi Peter,

Thanks heap for helping me to hunt down the ghosts. I eager to see what does it look like.
 
On the mobos stuff, unfortunately I own only on each of the socket. Don't have spare. But will ask Gigabyte if they have any spare that they don't need in their workshop.

Ready and fully armed to catch those ghosts with you.

Best
Praphan
 

Hello everyone,

Ghosts are finally hunted down by Peter !!

I am not quite sure where to post this but thought it might probably be best here so this thread can be closed in a happy ending way.

Being a lucky one with some priority treatment from Peter, I just received my NOS1 USB back to my listening room. Installation of  Phasure v1.0.2 driver was simple and took only few minutes. I connected NOS1 USB with my X58 chipset music server using a plain simple printer USB cable.   NOS1 USB worked perfectly fine with all my gigabyte mobos (P67, AMD and X58 chipsets). Anyone who is going to use Sandy Bridge processor can do so with no more compatibility problem with NOS USB version too.

Now I hope Peter doesn’t mind me doing some anatomy to NOS1 USB just in case someone eagers to see what are inside.

You will quickly notice that Juli@ card is replaced by another new sound card with USB interface. The sound device is smaller with compact size clocks on the rim of the card which understood to be very precise (low jitter) like Swiss watch. So when you return your NOS1 for an upgrade, don’t forget to put back the PCIe interface card and the DVI cable into the box. Peter might have some use with those stuffs. At least put them back in the market on eBay. It has no use for you anyway. This new sound card  is powered by the same PSU as it predecessor hence saving your burn-in time. My NOS1 USB comes with SPDIF interface as a gadget but not everyone will benefit from it. The USB connector that must go to your PC is located in the middle between the SPDIF connector and USB connecting to the small screen on the right leg of NOS. Talking about this little screen gadget, I hope Peter will spare some of his valuable time to write codes to fire up this screen in a near future. If this is possible then we don’t have to bother with the computer monitor anymore. Just use your keyboard to control XX z-6 with Auto HK command. Your cover art will be on your NOS right leg during playback. How nice ?

The DAC board located at the middle of the case remains intact. However, its PSU located in the same compartment as the toroidal transformers got some minor upgraded too. Looking closely in photo, you will see blue thin sheet of vibration damper slided in between those four heatsinks.  Two large Mundorf caps are now held in place by Styrofoam glued to the H-shape cover. This upgrade is to protect potential damage from vibration during shipping. Moreover, vibration may degrade SQ as well during listening session. A word of advise: when you ship your NOS1 back , don't forget to secure these four heat sinks with foam sheets like it was first shipped to you. Do the same for Juli@ card too.

Your NOS1 USB will be shipped back to you with greater protection. Peter put your original box in another bigger box with cushion foam floating between the two boxes.

How the upgrade sound compared to the NOS1 version ?  One single word : BETTER. But to what extend ? Well I cannot tell since I am the one who do not get the NOS1 works smoothly since day one. But definitely worth the upgrade cost. I also own Berkeley Alpha DAC which is also a very good DAC. I will compare both in a separate forum.

I am glad to be able to close this thread happily. I guess that my problem with NOS1 did give some inspiration for Peter to expedite his USB interface project which ended up not only solving Juli@ compatability problem but also sound much better than the first version too. "Excellente !"

Praphan
Mobo geek

PS. Sorry...... it can be impolite to do the anatomy without Peter's consent. The photos were removed.

Hi praphan, congratulations on the USB NOS1 - I'm sure you'll love it. And thanks for the write- up.

But personally, I would remove the attached images. It's just too easy for the 'competition' to see what Peter has been slaving away at for so long.Also, there's thread on gearslutz.com at the moment where some posters are calling the NOS1 a 'scam'. You'll probably get some of these total idiots saying things like, "He claims there are 8 DAC chips but it's obvious from the images praphan posted that there are only 4". Of course the twits wouldn't realise that there are another 4 on the underside of the board! The same idiots (supposedly industry 'experts') believe that the PCM1704 only accepts 96KHz sample rates. One claims that Peter has got the chip name wrong and that he must be using another chip really!

Of course, this is totally your call. And perhaps Peter doesn't mind. But personally, I'd never post images of the insides of any gear I own, out of respect for the designer/manufacturer.

Cheers, Mani

Hi Mani,
Thanks and done.
Cheers !

Just took a look. Wow, not shy of stating opinions with not one person having ANY first hand experience of the device thing they are discussing  :)

I can understand people who work in the industry getting tired of the bigger numbers equals better sound pitch. However, I hope someone decides to get better informed about how the NOS1 does what it does, because it looks like many of these guys are the people responsible for creating music that we listen to. Who knows, we might even end up with hi-res files that deliver on the promise of better reproduction  ;)

Nick.

It was also my first time reading threads on Gearslutz and I feel that some folks in there are pretty arrogant. Not to be surprised since most of them are recording engineers and professionals who think they know a lot. They tend to be so obsessed with technical specifications and numbers. Good recording engineers should use their ears more but I think some are too obsessed with technical specification driven by market demand and mass consumer behavior. Take a look at the example of THD war during the 70 and ‘80. Mass market audio components were back then often designed to look good “on paper” – on specification sheet – sometimes at the expense of the sound quality. The lower the THD, the better the amp was perceived to  be. This led the electronics giants to produce products with vanishingly low THD numbers. It became a contest between all leading brands not to be named here.

Recording engineers in Gearslutz should have also developed a habit of judging the product virtue from numbers before listening by their ears.

In our case, we are in a true “high end audio” business. Hi-end audio products are not only designed by ear, but are often hand-built by skilled craftspeople who take pride in their work. The designers are often audiophile themselves, building the products with as much care as if the products were to be installed in their own homes. This meticulous attention to detail results in a better quality of construction, hence reliability and durability. Hi end products are often backed by better customer service than mid-fi. Because high-end manufacturers care more about their products and customers, they generally offer longer warranties and periodic product update or exchange policy for improved quality.

To me, Peter is absolutely a hi-end designer and manufacturer. 

Praphan