Building a DIY Music Server

[MarcelNL]

2 minutes ago, Gavin1977 said:

When did this all kick off:

 

https://taikoaudio.com/shop/

 

New network switch and other stuff!

 

Even the DIY chassis is now back in stock in the DIY section.

 

 

A good while ago....over at WBF you cab keep up with developments by Taiko.

For now all that stuff is exclusively for Extreme owners.

[Gavin1977]

On 1/11/2023 at 8:11 PM, Exocer said:

You just reminded me, there were a few I did not try yet! Will report back as well.

 

These would be between my individual ER and Server.

I’ve been hearing good things about this idea too… but nothing concrete to share yet.

[rico25]

32 minutes ago, MarcelNL said:

A good while ago....over at WBF you cab keep up with developments by Taiko.

For now all that stuff is exclusively for Extreme owners.

First the switch was only for Extreme owners, but for 2 or 3 days now, the switch, network card and upcoming router can be purchased by everyone :)

[Superdad]

4 hours ago, rico25 said:

the switch, network card and upcoming router can be purchased by everyone :)

Everyone that has a spare $5K+ that is... 

[Nenon]

10 hours ago, rico25 said:

First the switch was only for Extreme owners, but for 2 or 3 days now, the switch, network card and upcoming router can be purchased by everyone :)

 

Emile has confirmed that the NIC and the switch will be available to non-Extreme users with the second batch. The network card is supported on Windows and Linux. 

 

The upcoming Taiko router and PCIe DAC/DDC will also be available to non-Extreme users. However, at the moment the PCIe DAC/DDC has drivers for Windows only. It should not be an issue to develop Linux drivers but there is no commitment for that yet.

 

Also, the Taiko USB card will be made available to non-Extreme users later this year. The USB card should work natively on Linux but that has not been tested yet. I can test it at some point. Obviously, it works fine on Windows. 

[Tatomek7]

4 minutes ago, Nenon said:

Also, the Taiko USB card will be made available to non-Extreme users later this year.

Do you think, dedicated Taiko drivers will be available as well?

[matthias]

26 minutes ago, Nenon said:

The upcoming Taiko router and PCIe DAC/DDC will also be available to non-Extreme users. However, at the moment the PCIe DAC/DDC has drivers for Windows only. It should not be an issue to develop Linux drivers but there is no commitment for that yet.

 

Nenon,

maybe you can convince Emile to offer then drivers for both Linux and Mac. I think this DAC card is very interesting and deserves a broader market penetration. A further advantage for Taiko might then be a better revenue for the R&D costs.

 

Thank you

 

Matt

[MarcelNL]

Lets bundle it with a port of xdms ;-)

[Gavin1977]

7 hours ago, Nenon said:

 

Emile has confirmed that the NIC and the switch will be available to non-Extreme users with the second batch. The network card is supported on Windows and Linux. 

 

The upcoming Taiko router and PCIe DAC/DDC will also be available to non-Extreme users. However, at the moment the PCIe DAC/DDC has drivers for Windows only. It should not be an issue to develop Linux drivers but there is no commitment for that yet.

 

Also, the Taiko USB card will be made available to non-Extreme users later this year. The USB card should work natively on Linux but that has not been tested yet. I can test it at some point. Obviously, it works fine on Windows. 

 

TACDA and TACDD for Taiko Audio CPU Direct to Analogue and Taiko Audio CPU Direct to Digital.

 

Uses Rohm BD34301EKV DAC... I should check out the What's Best Forum more often.

 

Analogue out AND toslink... so easy to compare with your existing DAC.

 

Summary document:  https://www.whatsbestforum.com/threads/taiko-audio-sgm-extreme-the-crème-de-la-crème.27433/post-845628

 

Wish I could go to Munich in May...

[matthias]

50 minutes ago, Gavin1977 said:

Wish I could go to Munich in May...

 

I will be there, VERY curious about TACDA.......😃

 

Matt

[littlej0e]

On 9/20/2022 at 1:28 PM, di-fi said:

APACER Industrial M.2 2280 SSD 120GB SATA 3.1 TLC -40…85°C @ 68 EURO (3 left)

 

FLASH Type : TLC BiCS3

= pSLC (as far as I understand, but doublecheck)

My understanding is the pSLC needs 1/3 of the total disk space to work optimal. So at 120GB you will have 40GB to install OS, should be plenty.

 

https://www.soselectronic.com/products/apacer/a52-255hha-00104-320333

 

PM me if you want to mail SOS direct (I have no financial interest other than that I bought all my APACER from them for a good price, shipped to Canada).

You rule! Just saved me hundreds and hundreds of dollars. Very much appreciated! 

[MarcelNL]

21 minutes ago, littlej0e said:

You rule! Just saved me hundreds and hundreds of dollars. Very much appreciated! 

Are you sure you can operate a regular TLC SSD in pseudo SLC?

Imo it's not that pseudo SLC needs 1/3 of the available space, you just have 1/3 left when operating in pseudo SLC.

 

 

 

[littlej0e]

Great question. I have no idea.
 

I only purchased Apacer RAM from SOS electronics. Paid less than half what it would have cost through other avenues. 

[MarcelNL]

RAM is random access memory, aka the sticks that provide volatile memory in your computer.

 

SsD is Solid State Disc, which is memory that does not suffer from Alzheimer....

Pseudo SLC is a mode in which you use another SSD storage format way less efficuent but with a sonic advantage

[littlej0e]

1 hour ago, MarcelNL said:

RAM is random access memory, aka the sticks that provide volatile memory in your computer.

 

SsD is Solid State Disc, which is memory that does not suffer from Alzheimer....

Pseudo SLC is a mode in which you use another SSD storage format way less efficuent but with a sonic advantage

Understood. In that case, I can't confirm TLC can indeed be used in pseudo SLC mode. Also not sure about the penalties you speak of, but I haven't noticed any restrictions using my 16G RAM sticks in a 10/10/12 config (10G dedicated to RAM, 10G to ZRAM/a.k.a. ramroot/a.k.a swap, 12G free). 
 

I assume the fake Taiko (Faiko?!?) server I'm building will be more of the same. Obviously, 48G RAM will be supreme overkill, so I'll play around with the configs and see what sounds best. Given the direct PCIE bus access for both storage and boot drives, I do wonder if loading to RAM will actually sound better in this hardware configuration. 

[MarcelNL]

12 minutes ago, littlej0e said:

Understood. In that case, I can't confirm TLC can indeed be used in pseudo SLC mode. Also not sure about the penalties you speak of, but I haven't noticed any restrictions using my 16G RAM sticks in a 10/10/12 config (10G dedicated to RAM, 10G to ZRAM/a.k.a. ramroot/a.k.a swap, 12G free). 
 

I assume the fake Taiko (Faiko?!?) server I'm building will be more of the same. Obviously, 48G RAM will be supreme overkill, so I'll play around with the configs and see what sounds best. Given the direct PCIE bus access for both storage and boot drives, I do wonder if loading to RAM will actually sound better in this hardware configuration. 

I'm afraid we're miscommunicating...

 

RAM memory is neither TLC nor SLC afaik...

 

Loading to RAM likely sounds better since all the processing is done upfront

[di-fi]

3 hours ago, littlej0e said:

I only purchased Apacer RAM from SOS electronics.

Hi @MarcelNL

The confusion is probably about my suggestion to buy Apacer (all types RAM and NVMe) at SOS electronics and that's what he did.

 

The other part ( @littlej0e quote) was the discussion you and me and others had here around 09/20/2022 about alternatives for pSLC type memory like the Revelation SSD and a way to power it externally (anyone yet?). 

 

A manufacturer quickly suggested here Apacer pSLC (in NVMe format) was excellent after he tested it but without providing how and what it improved soundwise and neither did he share the exact Apacer type he used :-(. Not very useful.

I think we all agree that Apacer RAM usually is excellent, and some will buy it cheaper ;-)

[di-fi]

3 hours ago, MarcelNL said:

Imo it's not that pseudo SLC needs 1/3 of the available space, you just have 1/3 left when operating in pseudo SLC.

That's correct, I think so too today. 

Capacity in pSLC Mode changes to 320GB pSLC instead of 960GB in TLC mode.

 

Google:

Operating Modes:

▪ pSLC Cache (Fastest) – 100%

▪ pSLC Enabled (Slowest!) – 30% • Widely varying performance (>10x)

 

pSLC stands for pseudo single-level cell. pSLC flash uses MLC NAND flash components, but the entirety of each cell is not used for storage, thus limiting the amount of data that can stored on each cell.

 

https://www.delkin.com/blog/comparing-pslc-and-slc/#:~:text=pSLC stands for pseudo single,and SLC are very different.

[Gavin1977]

Could try these new SLC SSD:  https://www.mouser.co.uk/ProductDetail/Swissbit/SFSA160GM2AK1TA-I-8C-11P-STD?qs=MyNHzdoqoQLUA8SQn5NJLg%3D%3D

 

Or, go for network boot to RAM and avoid an SSD all together (although I guess if you rambooted, e.g. like what Euphony does, then the SSD will be largely inactive anyway).

[MarcelNL]

2 minutes ago, Gavin1977 said:

Could try these new SLC SSD:  https://www.mouser.co.uk/ProductDetail/Swissbit/SFSA160GM2AK1TA-I-8C-11P-STD?qs=MyNHzdoqoQLUA8SQn5NJLg%3D%3D

 

Or, go for network boot to RAM and avoid an SSD all together (although I guess if you rambooted, e.g. like what Euphony does, then the SSD will be largely inactive anyway).

Interesting find!

 

Though it's pseudo SLC it's one of the few!

 

Booting to Ram works, yet in my system booting to Ram from the pSLC femto drive sounds better than from regular SSD.

for local files it'd be nice to have a pSLC a bit bigger than this one.

[Gavin1977]

1 minute ago, MarcelNL said:

booting to Ram from the pSLC femto drive sounds better than from regular SSD.

Interesting....

 

The Swissbit go upto 360GB..., or just go 900p optane...

[Nenon]

On 1/11/2023 at 12:21 PM, rico25 said:

Hi Nenon,

Doesn't change from ryzen 7 3700X, the ULPS without load is 32V :)

 

On 1/11/2023 at 12:59 PM, Gavin1977 said:

Yeah just checked - 33v out unloaded, v3 power supply seems to be working fine.

 

CPU is 8086k.  Next step is to start removing connections, or take back just to MB and build back up from there.  I’ve built a few PCs in my time, but this behaviour is odd.  Anyway error is now consistently A2, so should be possible to overcome.

 

No idea why this single component swap has caused such issue - everything else remains unchanged…

 

On 1/11/2023 at 3:16 PM, Dev said:

well, I also have the same issue. I found that when the ulps powered on from cold, the motherboard doesn't start although the motherboard is receiving the power. Give it couple of mins and it boots fine. Alternately, if you don't want to wait, one could short the ps_on jumper on the Taiko and the motherboard will start right away. Since I run the sever 24x7, I haven't been able to convince myself its a concern but when I am out for couple of days, I would typically shut down everything and that is where I would be experiencing it. Anyway, glad to know that the issue is with others as well. In one conversation with @SoulAnalogue, he seemed to think it could be related to high freq impedance and suggested me to add 0.1uf bypass but I already had it and it didn't solve the issue. It appears to me that its possibly some kind of handshake between the mobo and Taiko that might be causing it. It would be helpful if @JayM Taiko  can comment on it.

 

Discussed with Emile this morning. The root cause of this is the ULPS not being to deliver all the current the motherboard needs at cold boot. This is most likely caused by the 5A Hammond choke. You need a bigger choke and perhaps more capacitance (although if you use the Taiko cap, capacitance is probably fine but that also depends on the choke you pick). Higher voltage also helps as long as it does not exceed the Taiko DC-DC ATX specs. 

 

I was pretty sure that was the issue, but when I asked what CPUs you are using and @rico25 said Ryzen 7 3700X, that confused me. My guess is there are some peripherals/components in addition the the Ryzen that draw power. Maybe a GPU that draws less power is all @rico25 needs?

 

When the motherboard powers on it draws too much too power. That saturates the ULPS (5A Hammond choke in this case), and the voltage sags. When the voltage sags, the Taiko DC-DC ATX turns off to protect itself. That usually happens so quickly that's nearly impossible to see with a multimeter connected to the ULPS. Then you have to wait till everything resets...

 

Why that did not happen with the HDPlex DC-DC ATX for @Gavin1977? Most likely the protection circuit on the HDPlex is not as sensitive as the one on the Taiko DC-DC ATX. I managed to fry the first prototype of the DC-DC ATX Taiko sent out to test (it was fried about 2 hours after Fedex delivered it), and that forced Emile to adjust the protection to make the unit more durable / DIY friendly. 

 

So, the solution is to use a more powerful choke (and potentially add more capacitance). I have an old post about some options here: 

On 9/13/2021 at 10:20 AM, Nenon said:

For those who want to play with the Hammond 195J10 or the even bigger Hammond 195K12 choke, here is what you need to do:

• 22000uF - Hammond 195J10 - 22000uF - Hammond 195J10 - 220000uF; or
• 22000uF - Hammond 195K12 - 22000uF - Hammond 195K12 - 330000uF

 

However, when it comes to higher current chokes, there might be some better options than the Hammond. 

@MarcelNL  has some locally made chokes that were reported to be exceptionally good. I haven't tried them, but that would be my first choice to try, and I may do in the near future.

@Exocer had a great feedback with a Mundorf MCoil ZeroOhm inductor, but I am not sure what the max rating on this would be. These were designed for speaker crossovers but Mundorf tested them for @Exocer and confirmed they are good as a 5A replacement. I just don't know if they are capable to go beyond 5A safely. 

[Exocer]

@Rovo checked with Mundorf and they have validated the VN300-10 as a suitable alternative to the higher current handling Hammond chokes.

[di-fi]

On 1/11/2023 at 2:39 PM, Superdad said:

 

 

Major mea culpa alert! Send me eggs and I will apply to my face. 

 

What I wrote was correct about RJ45 SFP modules containing an Ethernet PHY, tiny cheap clock, and cheap voltage regs. (Avoid those!)

 

But what I said about Direct Attach Cables (DAC) was 100% wrong.  For some reason (ignorance and mind being elsewhere that day and not paying attention to words on the screen) I assumed that these DA Cables used 4 pair and were signaling Ethernet protocol--just as the RJ45 SFPs do.

Now I see (and many of the deeper tech details of it were quickly explained to me by @JohnSwenson when I ask him about it) that Direct Attach Cables use shielded twinaxial cable--just sending LVDS (low-voltage differential signaling) in fiber-mode format (a format close to SGMII, sometimes called 1000Base-X or Fibermode).  [The PCB traces at the back of all SFP/SFP+ cages in Ethernet equipment are these differential pairs going direct to an Ethernet switch chip, transceiver, or other protocol converter.]

 

Thus use of a Direct Attach Cable does--just like optical SFP/SFP+ transceivers--skip the need for an internal copper>fiber-mode PHY (as in the RJ45 SFPs).  And yes, unlike an optical SFP transceiver, there is no laser/optical>LVDS converter circuit inside--a DAC really is just twin-axial wire connected to basically blank SFP PCB. (That is why they are so inexpensive; I saw some for just $10) [This is true for the Passive versions of DAC, the Active versions of DAC--for runs longer than 7 meters--have LVDS "booster" chips to compensate for the attenuation.]

 

So indeed these DAC cables reduce the power requirements of the equipment they are connecting. How significant that is to performance and ground-plane noise is unclear. The quality of the wire used is important to keep BEC (bit error rate) low. High BEC will cause extra error-correction activity in the chips, something that runs counter to the goal of maintaining low ground-plane noise in gear. One can see that the wire gauge of a DAC gets progressively heavier--even just as the length goes from 1 meter to the 7 meter max for a Passive DAC. LVDS is great, but we are also taking about very high signaling rates, so attenuation must stay very low at high frequency and shielding should be good.

 

All that said--and my apologies again for getting it wrong the first go-round:

One should keep in mind that unlike a fiber-optic SFP/SFP+ connection, these Direct Attach Cables do not offer any galvanic isolation between the components.  So they will not interrupt the path of leakage currents or other "noise" that may be present from the upstream gear.

 

[To add that isolation back, one could design a box with high-speed digital isolators, followed by ultra-low-jitter "reclocking" flip-flops, fed by a low-phase noise clock (1.25GHz for Gigabit Ethernet), all supplied with great power networks. But that's basically a form of our EtherREGEN! (Gen2 as that's the one that will be putting Gigabit SGMII across the active-differential isolation "moat"; the first generation was just 100Mbps across the moat.)]

 

Still, a passive Direct Attach Cable may have some interesting applications and I intend to try a 7 meter DAC between two EtherREGENs in my own music system to hear the result. I presently use single-mode Finisar SFPs and fiber between them.

 

Cheers,

--Alex C.

 

Hi Alex (🍳)🥚🥚🥚,

 

So replacing a fibre optic cable with a Copper Twinax SFP Cable (or DAC cable) between ER and a Fiber network card in the server could improve sound due to less power needed resulting in the lowest possible noise?

 

I noticed some DAC cables are made for a specific brand but there are also general DAC cables. Anything to avoid for the ER?
You use 7m because the ER is next to the router I suppose? My ER is next to my server so 1M should do.

 

But I will have to give up both selected SFP (Finisar FTLF1318) modules on each side.

 

Quote

Taiko's recommendation for the lowest possible noise:
• RJ45 from router to switch as it still uses less power than dual fiber SFP modules (for which you need a SFP RJ45 module
• SFP DAC twinax copper cable from switch to (Extreme)

 

Also when using A and B side in ER there is still galvanic isolation between ER and server.
 

Please let us know your results.

Thanks,

Paul 

[Gavin1977]

Hi @Nenon - that is very timely feedback. 

 

So I'm using the HDPlex 800w DC-DC, with the same original silver cables, for troubleshooting, since this worked with the Nenon V3 and these components perfectly fine for the past year (my problems started and have remained since the Taiko ATX was inserted into my system).

 

Components in use: Asus Z390 Apex, 8086k, Apacer ram, Optane H10.

 

Step 1: Change GPU

Since bios indicated A2 and white light (GPU fault).  I thought this would cure my issue.  So I just tried a different GPU... Sapphire Radeon HD6450 1GB GDDR3 Low profile (<30w).  Cheap as chips off eBay and it's the one Pink Faun use in their server... its a good choice and much easier to work with.  Anyway...  I insert this and A2 bios error persists - but now with green LED light (SATA) rather than white (GPU). 

 

Step 2: HD6450 + switch out SSD

So I switch out the m2 SSD... Optane H10 for a H20 I have.  A2 error remains. green LED light.

 

Step 3: HD6450, plus new Optane H20 AND switch to powering HDPlex DC-DC using the HDPlex SMPS I used to use.

A2 error remains. green LED light.  CMOS reset doesn't help.  Switching to alternative bios doesn't help.  Now if I'd fried the choke by pulling too much current then you'd have thought that switching to an SMPS would have fixed it.

 

This build always ran pretty cool, so I can't believe it'd be motherboard VRM related (failed cap for example)...

 

I've built many PCs in my time, but this one has me stumped.  No logical reason why the issue occured in the first place, or why is persists.

 

I can understand your logic on power supply, I do have the Taiko Rectifier and cap ready to be installed, and was planning to upgrade the Hammond... but obviously can't progress with that currently.

 

Next step remove DC-DC and try with regular 500w ATX power supply (which I'd have to buy, as I sold my Corsair SFX a long time ago)...?

 

Could a failed/saturated choke cause excessive ripple to enter DC-DC converter and cause damage to both DC-DC converter and motherboard...