A novel way to massively improve the SQ of computer audio streaming

[Nenon]

I would not use water cooling in my system. 

I don't like the 'wet' sound! Just kidding. Any type of vibration is an enemy and reduces the performance of your clock. That's the main reason I do passive cooling, not because of the noise of the fan, which is much quieter than my A/C. Water cooling is not vibration free.

[Nenon]

On 8/1/2019 at 3:48 PM, austinpop said:

Sorry if I caused any confusion. Here are the HW specs of the custom server:

• ASRock H370M-ITX/ac motherboard
• Intel Core i7-8700T CPU
• Kingston HyperX Fury DDR4/2400 2x4GB RAM
• HDPlex H3 case
• Intel M10 Optane 32GB SSD
• JCAT Net Card Femto

 

@austinpop Have you considered sending that ASRock motherboard to SOTM to replace the clocks?

[Nenon]

8 hours ago, Ricardo007 said:

A lot of praise has been given to lush usb cable by eminent members of this forum.

I wonder if any experience has yet been gathered on ethernet and I2S versions of it:

 

https://stordiau.com/collections/cables-and-more-snakes-oil/products/ethist-2-ethernet-cable

 

https://stordiau.com/collections/cables-and-more-snakes-oil/products/hdmi-2-i2s-cable

 

Looks exciting for extra tweaking hdmi i2s with ground box like entreq

I am planning to try their i2S cable and compare with a decent HDMI cable and a DIY cable from a really-really good 24AWG 5N solid silver I have been using for years.

 

I am also planning to compare two AMD-based motherboards - an Asus gaming motherboard (ASUS ROG Strix X470-I) with a AsROCK server motherboard (X470D4U) with ECC RAM. Keep you posted on that as well. 

 

[Nenon]

On 8/10/2019 at 7:21 AM, seeteeyou said:

 

Recently Marcin went from previous generation of Apacer 75.B93H7.G000B to their latest wide temp. model

 

https://www.apacer.com/News/Detail/DDR4High-PerformanceWide-TempMemory

http://jplay.eu/forum/index.php?/topic/2786-ram-adventures-extremely-interesting-results/page-2#entry49239

 

Mostly likely B93H0 should be a typo, it's good for 0 °C ≤ TC ≤ +85°C

 

http://eflash.apacerus.com/spec/DRAM/75.B93H7.G000B.pdf#page=12

http://eflash.apacerus.com/spec/DRAM/75.B93H7.G000B DDR4 ECC DIMM 2400-17 512x8 4GB SA-E Anti-Sulfuration HF.pdf#page=12

 

Even that would cost $130 for 4GB, imagine how much we're gonna pay for the new one

 

https://www.mouser.com/ProductDetail/Apacer/75B93H7G000B?qs=0lQeLiL1qya80yN9UFKvQw%3D%3D

 

This is what Marcin was talking about and it's announced back in April

 

https://industrial.apacer.com/en-ww/DRAM/DDR4-Wide-Temp.-UDIMM

 

Maybe that's the next best thing when compared to what we're getting from Taiko Audio? Though there's always something even better that could go all the way up to +125°C

 

https://www.micron.com/products/dram/dram-for-automotive

 

Not that we'll reach that kinda temperature by any means, it's just a matter of stringent quality that's gotta require (much) better materials in the first place.

 

Do you know the model number of the new Apacer memory? 

 

[Nenon]

On 9/4/2019 at 1:20 PM, seeteeyou said:

There might be yet another important point that we're missing, Innuos made that decision to put everything inside the same box in order to minimize the degradation of clock cables. In Roy's experiences, even average clocks inside the same chassis that were powered by average PSU could work pretty much as good as something else that's much more expensive

 

https://www.head-fi.org/threads/review-comparison-of-5-high-end-digital-music-servers-aurender-n10-cad-cat-server-totaldac-d1-server-auralic-aries-audiophile-vortex-box.787020/page-78#post-14117821

https://www.head-fi.org/threads/review-comparison-of-5-high-end-digital-music-servers-aurender-n10-cad-cat-server-totaldac-d1-server-auralic-aries-audiophile-vortex-box.787020/page-78#post-14117902

 

For the PSU side, he also talked about how internal hardwiring could trump external DC connectors

 

https://audiophilestyle.com/forums/topic/29458-ultracap™-lps-1-listening-impressions-thread/page/4/#comment-589391

https://audiophilestyle.com/forums/topic/29458-ultracap™-lps-1-listening-impressions-thread/page/7/#comment-595070

 

More on the conductivity

 

https://audiophilestyle.com/forums/topic/37373-the-paul-hynes-sr7/page/22/#comment-943926

https://audiophilestyle.com/forums/topic/29458-ultracap™-lps-1-listening-impressions-thread/page/13/#comment-611953

 

Material IACS (International Annealed Copper Standard)

 

https://audiophilestyle.com/forums/topic/30376-a-novel-way-to-massively-improve-the-sq-of-computer-audio-streaming/page/49/#comment-645953

https://www.head-fi.org/threads/chord-electronics-dave.766517/page-396#post-13065507

 

Currently we do have plenty of both RCA jacks and plugs with pure copper contacts but they might not be exactly the "right" choices in terms of DC power. OTOH, Furutech made some jacks while Silversonic made some plugs as follows

 

http://www.furutech.com/2013/01/27/1153/

http://www.furutech.com/2013/01/27/1157/

https://silversonic.com/products/connectors/ultimate-xlr/

 

Speaking of connectors with pure copper contacts, could @austinpop please kindly ask the member who made that DC cable (with Oyaide FTVS-910) about this company in Japan?

 

http://www.to-pura.com/new item.html

 

They seemed to be able to manufacture some (pure copper) male connectors for headphones but the female ones are only good for adapters, therefore not meant for the jacks that are mounted on a chassis. However, they also listed something about custom orders on their site and it would be great to find out what the minimum order is. Thanks.

 

Perhaps offtopic, but just a brief comment on this:

The Jaeger connectors Paul Hynes uses in his upgraded SR7 are really good! I use them too... with the Mundorf SGW115 wire and JSSG360 in my LPS. 

https://www.sealconusa.com/images/connectors/pdf/standard-rapid-series.pdf

 

Look at the 50A / < 0,0007 Ohms "533 760 006" for example.

[Nenon]

15 minutes ago, Superdad said:

 

But don't forget that there is about 20x more resistance in 2 feet of 18AWG wire. 

 

Of course! More like 8x more resistance for the 15 AWG Silver but still a valid point. 

But also, a regular 2.1 mm DC barrel jack plug has 29x more resistance (reference: Switchcraft DC connector specs).

 

[Nenon]

5 hours ago, seeteeyou said:

https://audiophilestyle.com/forums/topic/34613-optane-ssd-as-boot-drive-sq-nirvana/page/10/?tab=comments#comment-825445

 

https://audiophilestyle.com/forums/topic/30376-a-novel-way-to-massively-improve-the-sq-of-computer-audio-streaming/page/462/?tab=comments#comment-900522

 

For those of us who aren't able to try Optane without any PCI-Express or M.2 slots available, both ElecGear NV-C01 and ElecGear NV-i9 are supposed to work with multiple Optane models when they're connected via USB

https://www.amazon.com/NVMe-PCIe-Gen2-Enclosure-PCI/dp/B07GNZSN3R
https://www.amazon.com/NVMe-PCIe-Gen2-Enclosure-PCI/dp/B07MLJ7FRD

UK ones

https://www.amazon.co.uk/USB-NVMe-PCIe-Enclosure-PCI/dp/B07M9LJG53
https://www.amazon.co.uk/USB-NVMe-PCIe-Enclosure-PCI-M-2/dp/B07GPB1VR4

In addition, Optane drives could be powered externally if our USB ports (e.g. tX-USBultra or tX-USBexp) were accepting external power.

However, 600P / 660P / 760P ain't even Optane to begin with. No M.2 models for 900P whatsoever while the only 905P is 22110 instead of 2280. Finally we've got 800P but that's limited to 118GB.

 

Putting an Optane card in such enclosure and connecting via USB defeats the purpose of using (low latency) Optane. One might be better off with some of the other USB storage methods discussed here in that case. 

 

Although, it may still sound good, I would argue that all the listening tests done with Optane connected to the M.2 slot with direct and low latency access to the CPU would need to be re-evaluated for USB-based connectivity. 

 

In other words, the quote in your post was NOT intended for USB-based usage of Optane. 

5 hours ago, seeteeyou said:

However, interestingly, music on NVMe Optane SSD was almost as good as in RAM.

 

[Nenon]

1 hour ago, LJONESATL said:

Has anyone tried out this type of I2S link and compared it to a similar quality USB connection (e.g., SOtM, JCAT, etc.)?  If there's a unique thread for a discussion along those lines, please point me towards it.

 

Yes, and I like the I2S better than USB... especially if you get the OCXO clock and a good external LPS. Give it a try if you have a chance. 

Just keep in mind that the I2S card does not work on Intel Core CPUs.

[Nenon]

6 hours ago, LJONESATL said:

That's surprising to hear.  Does it only work on AMD CPU's then? 

 

It works on Intel Atom CPUs. Not sure about other models - I haven't tried it. 

[Nenon]

To me, getting I2s from the PCIe slot on the computer is the best way to extract digital audio from the source to my DAC. Given that most DACs don't reclock the I2s signal, it is important to have a good clock at the source. This is where the PinkFaun Ultra OCXO makes a good audible difference. Once you extract the digital bits from your computer, the main thing we need to worry about it is preserving that clock signal over cables to your DAC. Luckily, most I2s DACs use Low-voltage differential signaling (LVDS), and that helps a lot. I guess the signal would degrade quite a lot if LDVS was not used.

 

I have compared various USB cards, reclockers, etc. But to me I2s done right sounds better than USB. And I've done a lot of tests, mostly with really good and expensive USB cables, many of the top power supplies on the market, really good DC cables, connectors, etc. Plus I am a purist and prefer less digital conversions. Why do PCIe - USB then use a USB cable to a recklocker, then another USB cable to the DAC's USB input, to do PCIe to I2s if you can simplify this and do PCIe - I2s. This is a controversial topic, and I don't want to start any useless debates. I am just sharing what works best for me in my system. Every design is a compromise, and my I2s is not an exception. It is a compromise as well. But there is something in the (well done) I2s signal that you don't get from the USB no matter how much effort you put in there. The sound is always more clear, and it fills out the room much better. But the thing I like the most is that the area between my speakers becomes much more dense, filled with sound, with very clear instrument separation, while not losing the depth and width of the soundstage (which is goes beyond the walls in my room). I've only experienced that with I2s. It's closer to the sound of my best recorded vinyls. 

 

I have posted about this previously, but I also compared the Intel i9-9900K / ASRock Z390 Phantom Gaming ITX motherboard to various AMD-based boards. For USB out of the motherboard, I prefered the ASRock / Intel combination. I did not bother testing with various USB cards as I was not that interested. At that point I decided to go I2s / Ultra OCXO, so I put more effort on comparing AMD-based motherboards. For the past month I have been working on a comparison as I mentioned in one of my previous posts:

On 8/15/2019 at 1:17 PM, Nenon said:

I am also planning to compare two AMD-based motherboards - an Asus gaming motherboard (ASUS ROG Strix X470-I) with a AsROCK server motherboard (X470D4U) with ECC RAM. Keep you posted on that as well. 

This is an interesting comparison with an Apacer industrial ECC RAM. I am wrapping up my findings and will let you know soon.

 

@Energy - you would not be able to get the PinkFaun I2s bridge work on your i7 NUC. The card does not work with Intel Core i7 for some unknown reason. I believe the manufacturer of the PCIe to I2s chip (CMedia CM8888) is not interested to do further development (but that's only rumors I have heard). For that same reason the PinkFaun I2s bridge does not support native DSD. If you decide to use that I2s card, your DSD files would have to be converted to PCM. Even with the convert to PCM option, they sound amazingly good in my system... but did I say that "every design is a compromise" in my second paragraph above?

 

Let me use the opportunity to criticise my current design now. What have I done? Let's summarize. I used an I2s card to avoid the conversion to USB and then to I2s. And because of that, I can't play native DSD files, and I have to convert DSD to PCM :). What just happened to the purist in me :)?!? Well, my DAC does not support native DSD at the moment, so I can live with that. The end result is still better than any USB output I have tried (in my system!). And I don't have that many DSD files anyway. 

 

BTW, PinkFaun USB cards don't have that problem. They work on every CPU and support DSD (assuming your DAC does). You can also add a Ultra OCXO clock. It's a good alternative to JCAT, SOTM, Matrix, etc. I have tried to compare the USB cards I've had access to, and they all have their own sound signature... just like anything else (i.e. motherboards, RAM, power supplies, etc.). 

 

The last thing I wanted to mention to @Energy is this M.2 to PCIe adaptor:

https://www.ebay.com/itm/25cm-M-2-Key-M-to-PCIe-x1-Riser-Card-Cable-for-Bitcoin-mining-R41SR-WK/253454407364

Since you have a cable company, this might be of interest to you. I was able desolder the ribbon cable and solder my favorite 5N solid silver wire and shield it with JSSG360 shield. If there is any audible difference with that adaptor vs. connecting my JCAT Net card directly to the PCIe slot, I was not able to hear it. But I did not really spend much time on that A/B test.

 

 

[Nenon]

1 hour ago, Superdad said:

 

Hi Nenon:

Nice post.

While I too enjoy using the I2S/LVDS input (on the Holo Spring L3 currently), what you say about most DACs not reclocking the I2S is not correct.  DACs such as the PS Audio units, most ESS Sabre DACs, and most others do not rely solely on the audio clock of the external I2S. The Holo Spring and I am sure a couple of others are the exception (and certainly this accounts for the large delta I hear when feeding it a well-clocked I2S).

Cheers,

--Alex C.

I stand corrected. Too late to edit my post, so let me say that my experience is limited to DACs that do not reclock the I2s signal. 

 

I did have a PS Audio DirectStream DAC and was aware about the I2s recklocking but for some reason I thought that was the exception, not the rule. A little research confirmed that more DACs recklock the I2s that don't. I also owned a Denafrips Terminator for almost a year, but never tried with good quality I2s. I wish I did...

 

50 minutes ago, Superdad said:

Yes, still the heavily modified (courtesy of @scan80269) SU-1

So you are still using USB and then converting to I2s. My post was about eliminating USB altogether and going PCIe->I2s / LVDS -> DAC. This is when the magic happens in my opinion.

 

But that brings up another really good point. There is no universal solution. And it's essential to understand the design of our DAC before building a server. If I was to bring back the PS Audio DAC, which BTW I have no intention to do, knowing that it is a pure DSD DAC, I would try upsampling my music to DSD first. It's a completely different approach than the PinkFaun I2s I have been talking about and most likely would have to be done with USB. It may or may not sound better than I2s. And then, even with the next firmware release, things can change completely... there is no universal formula here. 

[Nenon]

1 hour ago, LJONESATL said:

I just hope my idea doesn't go down in flames if for some reason their I2S bridge can't be used within this current new build

Sorry to be the bearer of bad news, but the Pink Faun I2S bridge would not work with your Intel I9-9900 CPU. I thought I made that clear a couple of posts ago. That's the main reason I started looking at the new AMD Ryzen CPUs. 

 

1 hour ago, Chopin75 said:

There is also a review of a pinkfaun server that uses its usb, i2s and spdif/aes bridges, each seems to have certain advantage, i will see if i can find that review when i get home

Probably this one - https://www.hifi-advice.com/blog/review/digital-reviews/music-server-reviews/pink-faun-2-16x-1-3/ 

 

[Nenon]

2 minutes ago, Energy said:

So you're saying it won't work on Intel i7 CPU's in general?

Correct. 

 

3 minutes ago, Energy said:

If that IS the case, that would mean it's not a driver issue so it has nothing to do with the operating system and more related to the CMedia CM8888 chip or CPU type as you say.

Correct.

 

3 minutes ago, Energy said:

If one does manage to get it to work on perhaps an AMD system, does it have issues upsampling PCM to DSD?

The PinkFaun I2s bridge is working perfectly fine on my AMD-based system and sounding amazing. However, it can only output PCM. You cannot upsample PCM to DSD with this card. 

 

25 minutes ago, Energy said:

And yes, I was going to use a M.2 to PCI-E like the one you suggested and shorten the cable using better quality wires with JSSG360^3 shields. Mylar aluminum foil (15mm) will go spirally down certain data wires to further reduce noise once I can figure out which ones should optimally be grouped up or separated. The teflon and JSSG360 of course would go over them all after completion. 

If this is not a trade secret (since you have a cable company), can you please share some pictures with the DIY community? 

[Nenon]

Now that everyone's attention is on the EtherREGEN thread ( @Superdad announced the first listening impressions and final price) I can report my latest findings.

 

About a month ago I decided to compare two AMD-based motherboards - a server motherboard - AsROCK X470D4U - and a gaming motherboard - ASUS ROG Strix X470-I. Not only that, but I also wanted to find out if using ECC memory makes a difference. Here is my original post:

On 8/15/2019 at 1:17 PM, Nenon said:

I am also planning to compare two AMD-based motherboards - an Asus gaming motherboard (ASUS ROG Strix X470-I) with a AsROCK server motherboard (X470D4U) with ECC RAM. Keep you posted on that as well. 

 

I was using both motherboards in the same configuration:

- AMD Ryzen 7 3700X CPU

- Sean Jacobs DC3 4-rail LPS (ATX - 12V, 5V,3.3V; EPS - 12V; all separate rails)

- PinkFaun I2S Bridge with OCXO clock

- Euphony / Stylus one box solution

 

I started with the ASUS ROG Strix X470-I motherboard and 2 x 8GB DDR4-3200 CL16 Non-ECC RAM. I had it for 3 weeks in my system. Just enough to get used to the sound. At that time I did a quick comparison between the AMD/ASUS and an AsROCK / Intel i9-9900K. Here is my post about that comparison:

On 8/7/2019 at 7:27 PM, Nenon said:

 

After spending a couple of weeks with a AMD Ryzen 7  3700X / ASUS ROG Strix X470-I and comparing with the Intel i9-9900K / AsROCK Z390 Phantom Gaming-ITX/ac, here is what I think:

- If I am using the USB out of the motherboard, I prefer the Intel / AsROCK. It sounds better to my ears in my system than the AMD / ASUS. 

- I was introduced to the Pink Faun I2S card and their ultraOCXO clock. Unfortunately I could not get that card to work on the Intel server, but it worked on the AMD. I prefer by a huge margin the sound of the AMD with the PinkFaun I2S.

- I compared the PinkFaun I2S with two other I had with me - JCAT USB card and Audiobyte Hydra (USB to I2S output), all fed by the same external LPS. I liked the PinkFaun a lot better. I wish I had more USB cards to compare with. It would have been nice to compare with a PinkFaun USB with ultraOCXO clock, but I did not have one.

- One thing I learned from all those tests is the faster you process the signal, the better it sounds (assuming you have a good LPS of course). More cores, more cache, higher speed - the more the better. Looking forward to try an AMD Ryzen 9 3900X when it gets available at normal prices. 

 

There is something really good about not using USB at all. I guess the simplicity of getting the signal out of the (low latency) PCIe slot and converting it to I2S with a really low noise clock pays out. Guitar strings and piano has never sounded so good out of my digital rig. Actually, I always thought they sounded really really good with my USB output. But my reference analog rig did a good job showing me they can sound better. With the I2S they get super close to my analog rig, perhaps even better in some cases.

 

If I was using USB, I would prefer the Intel / AsROCK. For me I2S without any USB conversions just sounds better, so I am going the AMD route. I will probably try a Ryzen 9 3900X whenever it becomes available.

Now the bad news. First, the PinkFaun I2S does not work on Intel. Second, it does not support DSD. Third, it only supports up to 192 kHz. For those of you with native DSD DACs who would like to upsample, that may not be for you. For me what I hear sounds a lot better, and I have made my choice. For now at least :). 

 

BTW, since I don't need the Intel i9-9900K / AsROCK Z390 Phantom Gaming-ITX/ac, PM me if you are looking for one. 

 

 

Then I switched to the AsROCK X470D4U, keeping the same Non-ECC RAM. I did not like it from day 1, but I kept listening to it for almost 3 weeks. To be honest, I could not wait to go back to the ASUS. Each motherboard had its own sound signature, and I certainly prefered the gaming ASUS motherboard over the server AsROCK. There was no enjoyment with the server AsROCK. I could listen to background music, but that was it. 

 

But what about the ECC memory? Per the ASUS ROG Strix X470-I specs, it does NOT support ECC memory. The AsROCK X470D4U does. 

I did some research and decided to try the industrial wide temperature Apacer ECC RAM (https://industrial.apacer.com/en-ww/DRAM/DDR4-Wide-Temp.-ECC-UDIMM ). The idea was not that I needed usability from -40°C to +85°C, but that this type of RAM was built with better quality parts and may sound better. Plus, I've heard from a few places that ECC RAM sounds better. The Taiko Extreme server uses ECC RAM for example. JCAT is also a big fan of the Apacer ECC RAM. 

 

I installed the Apacer RAM on the AsROCK motherboard and heard an improvement immediately. I still think I prefered the ASUS motherboard, but they got close. I wish I could install the Apacer ECC RAM on the ASUS motherboard. Well, I had the suspicious that although the ASUS ROG Strix X470-I motherboard does not officially support ECC memory, it may actually work. So I gave it a try. AND IT WORKED! I was able to boot the ASUS ROG Strix X470-I with the Apacer ECC memory. And it sounded really really good. I could not stop listening music all night that day. And I am still enjoying it :).

 

I still have the Intel i9-9900K / AsROCK Z390 Phantom Gaming-ITX/ac. So I decided to do another quick comparison between the AMD/ASUS and Intel/AsROCK gaming motherboards. Unfortunately, the PinkFaun I2S card does not work with my Intel CPU, so I had to switch to USB. I used the motherboard USB output to compare. That's the same comparison I did in the beginning of August (see above). The only difference was that this time I used the ECC RAM on the AMD/ASUS. And while the soundstage collapsed, because I was using the motherboard USB output, not the PinkFaun with ultra OCXO, I do believe I liked the AMD better this time. 

 

That leads me to a conclusion that the RAM, a component I always neglected in the past, is very important. That industrial Apacer ECC RAM is really good. I don't know if it sounds so much better because it is good quality or because it is ECC. My guess is that both contribute to what I am hearing. 

 

To summarize for those who did not want to read the entire long post:

- The gaming ASUS ROG Strix X470-I motherboard sounded a lot better than the server AsROCK X470D4U motherboard (to my ears, in my system).

- The industrial Apacer ECC RAM made a big difference. 

- It just happened that although not officially supported, the ASUS ROG Strix X470-I motherboard works fine with the Apacer ECC RAM, so that's what I ended up using. 

 

Note: I am assuming that everyone reading this post understands the difference between ECC and non-ECC memory.

Intel purposely does not add support for ECC memory to their consumer grade CPUs such as the Core i7 and Core i9, so they cannot be used with ECC memory. That option is reserved for Intel's server-based CPUs (i.e. Intel Xeon), which typically require server-grade motherboards. So Intel-based gaming motherboards typically only work with consumer CPUs and don't support ECC memory.

AMD on the other hand does not have such restrictions. This is why we can use ECC memory on some consumer grade AMD CPUs, such as the AMD Ryzen 7 3700X in my case.

[Nenon]

@rando

One of the main reasons to try a server motherboard was the ECC memory support.

To my surprise the ASUS gaming motherboard just works perfectly fine with ECC memory. It's also smaller, cheaper, and sounds better. 

 

The second reason was reading some posts where people claimed that server motherboards in general sounded better than gaming motherboards. But let's not generalize...

 

And the third reason was that the AsROCK X470D4U had IPMI interface, so no need to have a video card.

Those Ryzen CPUs require an external video card. That actually is quite inconvenient, but I think at the end, it might be a benefit for audio. The video chip on some of my Intel-based motherboards was one of the noisiest chips (it generated a lot of EMI; also treating it with EMI absorbers helped a little). So maybe it is a good thing to use a CPU and motherboard with no video support after all. 
 

BTW, I wonder how much the IPMI chip was influencing the sound on the AsROCK X470D4U motherboard. I never bothered disabling it, because the BIOS reset was really hard to reach under my LPS regulators.

[Nenon]

33 minutes ago, austinpop said:

 

Thank you. These are outstanding findings. Not only do you confirm the claims about the superior SQ of ECC RAM that Emile, Marcin, et al have been making, the fact that it works in the better sounding ASUS mobo is very exciting.

 

The only shame is that X470D4U looked perfect from a PCIe perspective, since it has 3 PCIe slots without risers, which would be perfect to house a JCAT Net Femto for Ethernet, and JCAT USB Femto for USB. As far as I can see the ASUS only has 1 PCIe slot, which seems more typical of mini-ITX mobos.

 

Actually the PCIe slots was another reason I picked the X470D4U motherboard. It was fitting perfectly fine in a Streacom FC9 chassis, but only two of the three PCIe openings in the chassis were possible to use. One of the PCIe slots did not match an opening. 

Also, the motherboard had a layout designed for server-based airflow, and it was not very easy to fit the copper cooling pipes. The CPU socket was shifted to the left and that required longer pipes than the standard provided by Streacom. Also, the RAM slots were standing between the CPU socket and the heatsink, and I had to use a cooler riser to raise the copper pipes above the RAM. There was also a lot of stability issues until a BIOS update was released a couple of weeks ago. To be honest, it was a hassle to work with this motherboard. In a way, I am glad it did not sound as good as the ASUS. 

 

I am starting to think that those mini-ITX boards may sound so good because of the shorter thick PCB traces on the board. But that's just a speculation in my head. 

 

The ASUS motherboard has only one PCIe slot, which I use for the PinkFaun card. It has two M.2 slots - one on the bottom where I install my Optane card and one on the top where I install one of these for my JCAT Net card - https://www.ebay.com/itm/25cm-M-2-Key-M-to-PCIe-x1-Riser-Card-Cable-for-Bitcoin-mining-R41SR-WK/253454407364. And I ended up replacing that ribbon cable with solid silver wire and JSSG360. As I mentioned in one of my previous posts, if that riser with the silver cable degrades the sound of my JCAT Net card, I have not been able to hear the difference. 

 

But yes, I am happy I tried the ECC memory, and I really enjoy the result.

 

[Nenon]

1 hour ago, Gavin1977 said:

Thanks Seeteeyou - so what is more important ECC, Wide temperature or Industrial grade?

 

We don't know. 

Wide temperature = better quality parts used.

ECC = error correction.

We have reports from people saying that non-ECC wide temp memory made a positive difference. We also have reports that ECC memory sounds better. Probably both make a difference, but that's just a guess. 

 

Theoretically, I can buy some non-ECC wide temperature Apacer memory and compare in my system. But honestly I am not that interested to spend the extra money and time for that test. Maybe someone else can do it, but it would require an ECC-compatible computer (i.e. AMD, Xeon, etc.). Can't test on a NUC7i7DNBE as it does not support ECC. 

 

I would not completely change my system just for ECC memory support. All those happy NUC users who want to try different memory can just try the non-ECC industrial wide temperature range Apacer (or another brand) memory. And for those building a new computer now may consider going the AMD route.

 

A few people reached out to me asking about the Apacer memory. Mouser is one of the places in the US that sells it with a hefty premium. If we can get enough people, we might be able to do a group buy direct at significantly lower price. Not sure I really want to organize that, but if you are interested ping me on PM. 

[Nenon]

36 minutes ago, Dev said:

The Apacer memory is something Marcin from Jplay has been suggesting for a very long time and definitely for good reasons.

Just to add to that - it’s ECC memory what Marcin is actually suggesting. 

[Nenon]

1 hour ago, Superdad said:

 

Hi:

I am genuinely curious about this RAM stuff. To be honest I do not recall exactly what modules are installed in my AL/NUC. But I do think they were chosen for very low latency, something like CAS 9 or better. Yet the memory you linked to has much longer latency—CAS 17. 

So if CAS is no longer an important parameter in selecting RAM, please discuss what is.

THANKS!

 

It's not that simple. Latency does make a difference, but it's only one of the factors. If all other factors are equal, the lower the latency, the better. But the quality of the memory matters even more. Those wide range temperature Apacer RAM modules are built with better parts. And they sound better. From my experience the quality of the RAM module is more important than the latency factor. And then there is the ECC RAM, which does not work on most NUCs but might be even another step up. 

 

 

[Nenon]

5 hours ago, TheAttorney said:

So, in my circumstance, I'm now convinced that....

      1 x 8GB stock sounds better than 2 x  8GB stock

 

I once forgot to connect the external power on my JCAT NIC. And I immediate heard significant degradation of the sound quality. I was not streaming anything, but playing locally cached files on my Optane card with Euphony / Stylus. I think the only accountable difference was that the JCAT NIC was drawing power from the motherboard. And the difference in sound quality was quite noticeable. 

My theory is that one memory module draws less power from the motherboard than two, and that's why there are various reports that a single memory module sounds better. 

[Nenon]

If ECC memory sounds better, which is a big IF (I haven't confirmed that, because it is not that easy to test, but other people have), I don't think it would be because of the actual detection of a bit error. ECC adds more complexity and the interaction between CPU, cache, memory, chipset, etc. changes. Parts of the CPU cache is utilized differently for example. So again, it's not that simple, and as many things in (digital) audio, it might be difficult to explain.

 

The aerospace industry uses many error correction methods, including ECC memory for any critical applications. You would never see an overclocking gaming motherboard with non-ECC memory doing the autopilot of a commercial airplane. In fact, at 30,000 feet you have much higher chances of neutron strike impact on Si devices, which is one of the main causes of such errors. 

 

Lastly, for those interested, here is an interesting thread about various memory experiments:

http://jplay.eu/forum/index.php?/topic/2786-ram-adventures-extremely-interesting-results/

 

[Nenon]

6 hours ago, RickyV said:

🥴 stock is now 0 at SOSelectronic.com, I bought the last one. I hope they restock soon. 👍

 

3 hours ago, elan120 said:

Aside from SOS, they can also be purchased from Mouser with a higher price.

 

I just talked to my friend (his company uses Apacer products). He said he could get this memory from their supplier and can provide it for $75. If anyone in the US is trying to find it, send me a PM, and I will give you his contact. 

[Nenon]

12 minutes ago, seeteeyou said:

Hardwiring could be done with those DIY kits, how about going nuts with 6AWG silver wires for their exceptional level of resistance?

 

http://www.tempoelectric.com/cables_DIY.htm

https://www.riogrande.com/product/999-Fine-Silver-Round-Wire-6-Ga-Dead-Soft/105306

 

And then here's 0AWG

 

https://www.riogrande.com/product/Sterling-Silver-Round-Wire-0-Ga-Dead-Soft/100654

 

UP-OCC ones are more like 12AWG

 

https://goods.ruten.com.tw/item/show?21821076194416

https://audiosensibility.com/blog/accessories/parts/wire-and-cable/#!/Neotech-12AWG-OCC-Silver-Teflon-Insulated-Solid-Core-per-foot/p/132810030/

 

I will take the thinner 16AWG UP-OCC wire over the thicker jewelry silver 6AWG wire any day! Those jewelry silver wires are what put the "bright" label on anything silver.

I have tried both (not those exact models but very similar) in various applications (signal and power) and the results have always been the same. But as I have mentioned before, my favorite for LPS wiring is the Mundorf Silver/Gold SGW115. YMMV. 

[Nenon]

12 hours ago, Volfram said:

Dear Nenon, how is the DC3 connected to ATX and EPS contacts?

Do you have schematics?

I suppose you dont have all 24 + 8 connected since some are not used? -12V for example.

(See number 14 in attached picture)

 

A 20 PIN connector with 3.3V, 5V, and 12V. And an 8 PIN EPS with a 12V rail.

Modern motherboards don't use the -12V or -5V. Just googled Innuos images, and on this one you can clearly see what PINs they wire up on the 20 PIN connector.

 

Don't worry about PIN16 (PS-ON), you can skip that one but should connect the rest of the wires on the 20 PIN connector + the EPS connector. That worked on all motherboards I have tried so far ...

 

[Nenon]

4 hours ago, Volfram said:

@Nenon, To understand the schematics I compared it to a ATX 20 pin layout I found on wiki(see attached picture)

I had to flip the wiki layout to match the innous, can you confirm this is correct?

Whats the voltage rating for blue cable in port 14 "PS_ON#" and gray in port 8 "PWR_OK"?

 

You don't need to connect PIN 14 "PS_ON".

PIN 1 is +3.3VDC.

PIN 3 is Ground.

PIN 9, PIN 8, and PIN 4. is +5VDC. 

PIN 10 is +12VDC.

That's how I connect it. We are going offtopic here, though. Ping me on PM if you have more questions.