DIY Project High Performance Audio PC with high quality wiring

[Exocer]

@StreamFidelity Kudos to you for such an awesome build. I have been back and forth over PSU decisions for over a month now. My system is available in my sig. The main components are: i7 8700T, 8GB Apacer, Asus Strix Z390-I, Elfidelity USB add-on card with Gentooplayer OS. I do not upsample as I like the sound of my Yggdrasil A2 with standard CD quality streaming via Tidal. Would greatly appreciate your input.

 

I notice that you regard the Keces P8 highly and it appears to be the most convenient,available high quality options for me. Have you had an opportunity to compare it to any other highly regarded PSUs with similar continuous output at 19v?

 

Also,

I see that you have experience with both the HDPlex 400W DC-ATX and 800W DC-ATX converter. Of the two options below, which would you say was superior sounding?

 

1. 2x Keces P8 19v feeding HDPlex 800W DC-ATX Converter

2. 1x Keces P8 19v feeding HDPlex 400w DC-ATX Converter + Keces P8 12v feeding EPS directly

 

Thinking to start with 1 Keces P8 19v feeding a 400W DC-ATX converter and consider working towards either of the above options.

 

Thanks,

-Rob

[Exocer]

3 hours ago, elan120 said:

The peak demand during boot up sequence for either i9-9900K or KS is much higher than average consumption, based on observation for KS, it can go up to around 120W or slightly higher, so for P8 at 19/20V-8A, it works, but the 12V-8A model likely be insufficient.  Unfortunately, the option for a 12V linear power supply that is rated higher than 8A is quite limited, and a 12V power supply that works with i9-9900K or KS will likely need to handle peak current demand at 12A or higher, which is challenging to source one.

Based on this logic, dual Keces P8 19v feeding HDPlex 800W DC-ATX adapter seems like it would provide more headroom. CPU EPS coming from output of the DC-ATX. Does that additional headroom make it to the CPU if going through the converter while using only a single EPS input at the motherboard?

 

 

[Exocer]

2 hours ago, Nenon said:

I have been a beta tester for a while. Waiting to receive his latest prototype in a HDPlex H5 chassis to see how well the heatsinks on the H5 would do in this application. It may need to go in a case with better heatsinks at the end. The Asus SAGE / dual Xeon box I built is perfect for testing those types of power supplies - highly resolving and drawing a lot of current if unleashed. 

It would be awesome if this is the case he goes with. It is a perfect match for my server.

[Exocer]

2 hours ago, elan120 said:

Yes, it would, but will need to check both HDPlex 800W DC-ATX outputs carefully due to this concern.  OTOH, single P8 at 20V feeding HDPlex 800W DC-ATX powering i9-9900KS EPS work without issue.  It is my interim solution until I finish building my 12V power supply to power the EPS directly.

 

Understood. 

 

Do you still feel the need to power PCI-E expansion cards (JCAT Femto NET/USB) with dedicated LPSU if the PC is already powered by an lpsu at the Keces/HDPlex level or better?

[Exocer]

Xilllinx vs Startech to commence tomorrow. Will let some tunes play overnight for initial break-in. So far, it is sounding at least as good as the Startech.

 

Will start with my favorite SFP: Finisar 1321

 

[Exocer]

6 hours ago, MarcelNL said:

@exocer

Is the result so close that you're still A-Bing or are you so baffled with the sound ? ;-)

(just curious about what your findings are)

 

 

@MarcelNL 

They are very different from each other. Both provide low noise floor but the Xilinx is a bit better in this regard. The Startech sounds a bit more musical and the Xilinx is more clean/precise so far. Lets give the card more time to break in. I did switch to my Planet SFPs to clean my pallet and because I did not like the 1321s with the Xilinx out of box. I will switch the Finisars in today as the sound of card is changing for the better as time goes on.

 

Note: I use a larger and brand new adapter for the Xilinx card so that too may require some settling in. 

 

Will give further updates as things progress.

 

[Exocer]

17 minutes ago, StreamFidelity said:

 

I think a riser cable is very unfavorable. I prefer a direct connection.

 

All eight PCIe lanes are used for data connection to the CPU, so a slot with a direct connection to the CPU is recommended.

 

RSS uses all 16 logical processors of my i9-9900K CPU. Activation of hyper-threading is required!

 

There are other recommendations:

 

- Interrupt moderation
Disable in Network Adapter’s Advanced Properties.

 

- Max frame size
Configure to maximum supported by network in Network Adapter's Advanced Properties.

 

- Receive Side Scaling (RSS)
RSS is enabled by default for best networking performance. It is best to set the maximum number of RSS queues to be equal to the maximum number of RSS CPUs.

 

Thanks! I keep my i9 9900k Hyper Threading activated at all times. My server runs on Euphony so I am not sure how to apply some of the recommendations but will look into it.

 

Yes, the more I think about it the more I realize this build deserves a full ATX board. That will be in the cards soon.

 

For now, it is using a riser cable. Perhaps the Startech is a better chose for those with Mini-ITX boards...maybe not. Time will tell. 

[Exocer]

I've switched the Finisar 1321's back in (from the Planet tech MGB-TLX) and today the card sounds A LOT better. The very first observation I didn't share is that the midrange seemed a bit recessed with the Xilinx out of the box and now they're coming in beautifully, like the Startech. I am really liking the sound today.

 

The plan is to leave the Xilinx/1321 combo in for a few weeks and switch back to the Startech.

 

I believe I have reached the point where the strengths of the Xilinx are apparent while at the same time giving me what I liked about the Startech. So this could be considered a win (so far).

[Exocer]

Update on Euphony OpenOnload capability:

 

Zeljko had some issues setting up OpenOnload on my system remotely. Also, losing connectivity and an inability to see errors makes it difficult for him to work. We will need to wait for him to obtain the card in-house to test/deploy the proper items and push it out to the rest of us. Will be looking forward to that :)

[Exocer]

@MarcelNLdid you purchase the "onload" version?

[Exocer]

On 1/28/2021 at 5:14 AM, seeteeyou said:

CN-0NPHCM-28298-74B-0435-A00 = PPID (Piece Part Identification) numbers from Dell

 

https://www.dell.com/support/kbdoc/en-hk/000139589/how-to-find-the-ppid-piece-part-identification-on-dell-poweredge-parts

 

And then there's something about SFKEY

 

https://forums.servethehome.com/index.php?threads/solarflare-sfn8522-20-shipped.25319/#post-234729

 

Solarflare Linux Utilities RPM (64bit) is linked below and there's a copy of SFKEY as mentioned above

 

https://support-nic.xilinx.com/wp/drivers

 

Apparently we could also get two different flavors (i.e. SFN8522 and SFN8522-Onload) from Dell

 

https://support-nic.xilinx.com/wp/oem

 

I also looked at page 13 of Solarflare Server Adapter User Guide under Solarflare SFN8522 for Dell

 

https://support-nic.xilinx.com/wp/oem?sd=SF-103837-CD-20&pe=SF-117316-DH-9

 

Now let's compare the differences here

 

https://www.provantage.com/solarflare-communications~880SOLA.htm

 

Flareon Ultra Dual-Port 10GBE Server I/O Adapter

Manufacturer Part#  SFN8522
UPC Code: 081619101050

https://www.provantage.com/solarflare-communications-sfn8522~7SOLA01M.htm

 

Flareon Ultra Dual-Port 10GBE Server I/O Adapter with LL Firmware

Manufacturer Part#  SFN8522-PLUS
UPC Code: 816191010519

https://www.provantage.com/solarflare-communications-sfn8522-plus~7SOLA01N.htm

 

Flareon Ultra 10GBE I/O Adapter with LL Firmware Onload Licenses

Manufacturer Part#  SFN8522-ONLOAD
UPC Code: 0816191010618

https://www.provantage.com/solarflare-communications-sfn8522-onload~7SOLA01T.htm

 

First one without LL (low-latency) firmware costs about $200 while the rest would be over $900, is it realistic to to expect that LL firmware could be flashed on any non-Onload cards (without that factory-enabled goodness) for free or what?

 

I guess that maybe it's a good idea to get an answer from Xilinx directly?

 

https://www.xilinx.com/news/press/2019/xilinx-to-acquire-solarflare.html

Just checked both of my Xilinx cards. It appears they both have Onload keys installed according to sfkey:

 

 

 

[Exocer]

7 minutes ago, seeteeyou said:

It might / might not be good enough, depending on what profile(s) we're looking for

 

https://www.xilinx.com/publications/product-briefs/SFN8041-product-brief.pdf

 

https://www.xilinx.com/publications/product-briefs/SFN8042-product-brief.pdf

 

https://forums.xilinx.com/t5/Ethernet/onload-reports-no-error-but-I-don-t-see-any-performance/m-p/1107552#M19026

 

https://forums.xilinx.com/t5/Ethernet/SFN8522-TCP-direct/m-p/1187109#M21575

 

828ns – A Legacy of Low Latency
https://technologyevangelist.co/2019/04/18/828ns-a-low-latency-legacy/

This is super helpful! I think I need to switch to the "Onload" card, as I am using the Scaleout Onload card.

[Exocer]

@StreamFidelityI would like to give you a huge thanks for introducing us to this fiber NIC.

 

Since initial testing, I have switched to a full ATX board, and have the Xilinx SFN8522 Onload sitting in a proper PCI-E 8X direct to CPU slot. I have also been able to move my Optane storage to a direct to CPU PCI-E slot. The board I am using has 3 16x sized PCI-E slots, with capability of PCI-E lane breakdown of 8x/4x/4x (last slot selectable between PCH or CPU direct). This was a noteworthy improvement in general. Everything sounds less "stressed" and fluid. Less restricted.

 

A more recent test was conducted, by switching the Startech PEX1000SP2 back into the system after having lived with the Xilinx for several weeks.

 

Observations:

• The Xilinx sounds a lot closer to local playback than the Startech.
• The sound stage was a bit smaller with the Startech, and overall resolution and presentation was a more obvious step-down compared to local playback.
• The Xilinx narrows the streaming/local playback gap much more so than the Startech.

Overall, very happy with the purchase and it is definitely here to stay. Heat may become an issue though, so I am researching methods to deal with that..unless you plan on mass-producing the HDPlex top plate! :)

 

Cheers

[Exocer]

1 hour ago, StreamFidelity said:

I actually offer the acrylic glass top plate. Unfortunately only within the EU, as shipping and customs regulations are too expensive and time-consuming.

Awww shucks 😁. Looks like I may need to relocate!

[Exocer]

2 minutes ago, John769 said:

Which is? (out of interest) 😐

Gigabyte Designare z390. Received a special offer I could not refuse :). Nothing new or cutting edge going on here.

[Exocer]

On 4/25/2021 at 11:29 AM, StreamFidelity said:

About the sound: Like the SFN8522 NIC, the sound is three-dimensional, more details can be heard. No sharpness and a deeply graduated bass. Of all these good features, with the X2522 Nic improves a bit. There are no worlds, but if you want the best ... 👍

Excellent report. Is there a specific license we should look for as well?

 

Thanks,

-Rob

[Exocer]

@StreamFidelity I tested with the Paul Hynes SR7T 19V 10A rail in the below post:

 

 

I encourage all Taiko DC-ATX owners to test this. There were certainly some positives in going this route even with the 19V input (using the SR7T). The best of both worlds would be to combine the positive traits I observed powering ATX and EPS with the Taiko with the increased slam/ease/dynamics/body you get with the SR7T going CPU direct. I have a feeling this is what the Unregulated supply will do, and, take the strengths even further. Will build the unregulated supply as soon as I can.

 

[Exocer]

4 hours ago, StreamFidelity said:

Audio PC: passive cooling of Solarflare network cards

 

Due to the high performance of the Solarflare NICs (Network Interface Card) of the 8 and x series, the chips installed there get very hot. They are already equipped with an oversized cooler compared to other cards, but the manufacturer expects a case fan with at least 200 rpm. The only solution so far has been generous ventilation holes on the lid or cover up the lid.

 

The Solarflare NICs have very low latencies and are therefore used, for example, in high-frequency trading on stock exchanges. In the audio sector, low latencies are also a great advantage. Taiko Audio and others noted outstanding sound, but it deteriorated as the heating increased. That is also my observation. That's why I developed passive cooling with heat pipes.

 

 

The HDPLEX 2nd Gen H5 Fanless Chassis with heat pipes offers excellent passive cooling options by simply dissipating the heat via the side wall. If you don't use a passively cooled graphics card, the left side wall is free. Due to the different slots on the motherboards, a flexible connection option for the heat pipes was required. The attachment in the middle enables the heat pipes to be shortened. Since it is a prototype, the production is still a bit rough.

 

 

The block at the top enables it to be attached to the Solarflare NIC with four precisely fitting holes.

 

 

Here you can see the heat sink of the XILINX Solarflare X2 Series Ethernet Adapter - XtremeScale X2522, which is not sufficient for passive cooling. The contact pressure of the heat sink on the chip is created by means of spring plug-in holders.

 

 

The spring clips must be carefully removed from the back of the circuit board. The bracket is pressed together with tweezers and the connection to the circuit board is loosened.

 

 

The bare solar flare chip comes out and it is cleaned with isopropyl alcohol. Then a new thermal paste or thermal pad comes on.

 

 

The new cooling block is carefully attached to the chip with the removed spring plug-in holders.

 

 

It must be ensured that the cooling block lies flat on the chip.

 

 

A thermal paste must be applied so that the heat is dissipated well via the side wall. Before doing this, you will try out how the heat pipes are positioned and marked.

 

 

Installed it looks like this:

 

 

With the acrylic glass lid closed, the prototype is already impressive.

 

 

performance

 

I was very curious to see if that would work. After about an hour the left side wall gets quite warm, which is a very good sign. The heat dissipation via the outer wall works! An endurance test now follows. I will also completely close the housing cover with the original.

 

Sound impression

 

Even after hours there is finally no audible drop in performance. The enormous clarity, the stage far out the loudspeakers and the unexcited, slag-free sound are retained.

Wonderful work! I am contemplating picking up the newer Solarflare card and will likely emulate what you've done here :) Thanks for sharing. It looks very well executed.

[Exocer]

FWIW, I find this aspect interesting although a bit vague:

 

Quote

 In order to implement audio performance that exceeds the limitations of general motherboards in the existing market, the sMB-Q370 has applied a large number of audio parts that have been verified by being used in SOtM products for a long time and is designed to minimize the impact on each element inside the board to reduce noise as much as possible. Although this explanation may seem very simple, it can be said that it is the culmination of the many technologies that SOtM has implemented so far. The very deep and in-depth accumulation of technologies has made this possible.

 

That being said, there has to be "some" sonic value in building a motherboard with parts known to sound good in audio-specific gear. If the price is within reach for me i'd be willing to take the plunge and report on whether any noteworthy gains are to be had with this board. Also worth noting, there are some well-known brands out there producing world-class sound quality with off the shelf boards, some (at least one) of which made an attempt to produce their own motherboard and decided the off the shelf board was better sounding.

 

Interested in seeing where this goes. The decision to stay several generations behind is a bit controversial, but may have been done intentionally? As someone who does not upsample I am not too bothered by their choice but it would have been nice to use a newer chipset for the increased PCI-E lanes, advancements in VRM, and future longevity.