Jump to content
IGNORED

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


Message added by The Computer Audiophile

Important and useful information about this thread

Posting guidelines

History and index of useful posts

Most important: please realize this thread is about bleeding edge experimentation and discovery. No one has The Answer™. If you are not into tweaking, just know that you can have a musically satisfying system without doing any of the nutty things we do here.

Recommended Posts

http://www.myhifishop.de/Devices/AQ-SWITCH-SE-Audiophile-High-End-Network-Switch-audio-LAN-isolat::87.html?language=en

Quote

- The internal power supply circuits are now carefully optimized with additional Ultra-Fast capacitors.

 

Holy smoke, are we running Yottabit Ethernet with this kinda speed now?

 

Ultrasound Beyond the Speed of Light - Acoustics.org
http://acoustics.org/pressroom/httpdocs/150th/Mobley.html

 

piSZLAQ.jpg

 

Seriously, this particular one doesn't seem to be small enough to fit inside a switch

 

https://cdme.osu.edu/sites/cdme.osu.edu/files/uploads/0006a_capacitorbank.pdf

 

Now I wonder if those Ultra-Fast capacitors were any good if they're powering something else.

Link to comment
  • 3 weeks later...

Not sure if there were any Classic ThinkPad (with 7-row keyboards) fans here, apparently a group of enthusiasts in China were "upgrading" some old motherboards with Intel i5 or i7 processors as follows

 

https://www.facebook.com/lcdfans

 

Here are some articles and discussions

 

https://hn.svelte.technology/item/15274644

https://geoff.greer.fm/2017/07/16/thinkpad-x62/

https://www.notebookchat.com/index.php?topic=65985.0

https://www.laptopmag.com/articles/lenovo-thinkpad-x62-mod

https://www.notebookcheck.net/X62-Laptop-Review.211598.0.html

https://www.reddit.com/r/thinkpad/comments/5w4zpw/t70_specs_and_3rd_batch_x62_orders_available/

 

Then I checked prices and how many orders were required for each batch of production, they're asking 3999 RMB a piece (about $600) for 100 pieces of "ThinkPad X310" and 2999 RMB each (about $450) for 50 pieces of "ThinkPad X210" respectively.

 

Here are some links to whole bunch of pictures

 

https://weidian.com/item.html?itemID=2208761094

https://forum.51nb.com/thread-1812742-1-1.html

 

https://weidian.com/item.html?itemID=2205138453

https://forum.51nb.com/thread-1811153-1-1.html

https://forum.51nb.com/thread-1811697-1-1.html

https://forum.51nb.com/thread-1813129-1-1.html

https://forum.51nb.com/thread-1814305-1-1.html

 

Basically they're legit and the minimum number of orders seemed to be quite reasonable for such a niche product. Last time @romaz approached DFI in Taiwan and they're quoting $1,500 to get one piece of custom-made motherboard off the ground. Now we're talking about 40% of that ($600) @ 100 pieces or even 30% of that ($450) @ 50 pieces for an older model, not to mention that Celeron N3160 should be even cheaper than those i5 or i7 processors.

 

Do you guys think we'll ever reach that kinda number so that we could make our "dream" music server motherboards a reality? It will be designed exclusively for sCLK-EX so basically there will be no system clocks whatsoever. And then we could use LT3045 for all regulators that require less than 1.0A while the rest could be supplied by an external PSU separately. 4 lanes of PCIe x1 are connected to the CPU of each Celeron SoC and we could have up to 4 PCIe x1 slots available, that would match 4 taps from each sCLK-EX and 4 clocks from 4 pieces of tX-USBexp perfectly. Obviously we won't have on-board SATA and Ethernet etc. to minimize the size of the motherboard in order to shorten the length of clock cables as much as possible.

Link to comment
20 minutes ago, LTG2010 said:

plus the expense of a second tx. Any thoughts appreciated.

 

If you don't wanna invest in a second tx, that's really up to you and end of the story.

 

20 minutes ago, LTG2010 said:

but you also have a weak link, the USB socket on the HDD enclosure


Someone in Japan actually modified the system clock of this SATA to USB adapter and that particular move turned out to be even more effective than any other tweaks

 

http://blog.jspcaudio.net/?eid=276

 

However, almost all HDD docks and enclosures etc. have DC inputs so basically all power regulators inside them pretty much suck. That's why we might wanna get something like this instead

 

http://www.addonics.com/products/adsau3.php

https://www.conrad.com/ce/en/product/975128/GBIC----1x-USB-30-port-B-1x-SATA-socket-2-pin--------28554C127

 

ReKbief.jpg

 

SOtM SATA Filter II is still good to go, and then we could still supply power directly to that 15-pin SATA power connector without any power regulators whatsoever. Even the system clock could be modified by SOtM easily as it's located next to the USB connector.

Link to comment
3 hours ago, lmitche said:

Hi seeteeyou,

 

I'm powering a 3.5 inch hdd inside a Startech USB 3.1 gen 2 enclosure and my experience is different than you describe here.  The enclosure has a 12 volt input as you explain.  However the 5 volt power source for the drive appears to come from the USB vbus. Enhancing the quality of this vbus power actually enhances drive performance where I can measure variations from 165 to 212 mbs transfers rates. SQ improves as well.

 

Hi lmitche,

 

Many thanks for sharing you tips regarding vbus power since I didn't expect that would affect both USB DACs and HDDs, the most interesting part would be both quantitative and qualitative changes were observed. We learn something new every day.

 

Regarding the current of HDD docks and enclosures, now I realize the former ones are definitely good for both 2.5" and 3.5" drives but the latter ones might be a different story?

 

We need Type-C connectors on a server to reach 1.5A / 3A / 5A maximum current. Let's say we're talking about the legendary tX-USBexp here, we should be limited to 0.9A since Type-B connectors aren't really meant for anything that requires 1A or above. We could still get away with 3.5" HDDs because even 12TB ones would require only 0.4A from 5 volt power source

 

https://images.anandtech.com/doci/11842/wd_gold_12tb_bg.jpg

 

However, some "monster" drives such as this 2.5" SSD from Crucial would actually require 2A

 

https://www.vortez.net/articles_file/29793_crucial-bx100-1tb-bottom.jpg

 

1A for Samsung 850 EVO

 

https://images-na.ssl-images-amazon.com/images/I/91gOObm6rFL.jpg

 

0.5A for another one from China

 

https://ae01.alicdn.com/kf/HTB1CUvZKpXXXXa.XVXXq6xXFXXXG/F9-1TB-KingFast-2-5-SATA-SSD-For-Dell-HP-Thinkpad-Lenovo-ASUS-Acer-Sony-Toshiba.jpg

 

Fortunately we don't really need any 1TB SSDs since HDDs are so much cheaper while we're also getting a much higher capacity. However, the USB vbus of HDD docks alone (i.e. inevitably some power should be drawn from 12V DC input after going through a regulator) might not be good enough for some hungry 2.5" SSDs if a server weren't having USB Type-C.

 

I really was paying too much attention to 5.25" optical drives in the first place since I just kept thinking about how Addonics would manage to provide compatibility for both HDDs and optical drives in the first place. I checked a bunch of BD and DVD drives, so far 5 volt power source requirements seemed to be ranging anywhere from 0.8A to 2.0A

 

http://www.ixbt.com/optical/dvd-01-03/pioneer117-label.jpg

https://ssli.ebayimg.com/images/g/sNkAAOSwhQhY0y0a/s-l1600.jpg

 

Some USB DACs don't really need vbus at all when they're getting power from an internal PSU instead. However, sometimes they might wanna keep the connection alive by means of computer "handshake" and that won't happen if vbus weren't present. Maybe the performance of vbus still matters even if only computer "handshake" were involved? Now I wonder if similar cases were applicable for both USB DACs and HDDs.

Link to comment

The whole is greater than the sum of its parts.

 

While improving an individual system clock might not be necessarily such a big deal (who knows?) when we're merely focusing on a single piece of component, IMHO it's still a good idea to consider how a mediocre system clock might end up being the weakest link that could potentially "ruin the day" for a server. As long as we've got something that's sticking out like a sore thumb, it's still somewhat tricky to estimate how much damage is done to the overall performance.

 

Maybe that's why in theory the performance of a built-in system clock wouldn't even make any differences for stuff like Ethernet devices, in practice we still have to figure out the opportunity cost of NOT replacing that with sCLK-EX. Some of us were dedicated enough to modify every single system clock with throughout the entire chain, it's considered a 100% sCLK-EX rig since they're leaving no stone unturned. As a result, we've gotta give credits to BOTH presence of superior system clocks AND absence of inferior ones.

 

Like what Roy mentioned before, this game should be all about minimizing the overall amount of noise (in terms of system clocks and power etc.) since it's gonna up the damage after going through so many components. Even the manufacturers of computer parts themselves might not be aware of how they're contributing to a complete system because plenty of them wouldn't really consider PCs to be toys for we audiophiles in the first place.

Link to comment

Intel Atom C3000 Series

https://ark.intel.com/products/codename/63508/Denverton

 

Do you guys know how to interpret these block diagrams? Some PCIe lanes should be linked to HSIO but not sure if they're actually connected to the CPU directly

 

http://linuxgizmos.com/linux-ready-module-features-atom-c3000-and-4x-10gbe-ports/

https://www.kontron.com/products/boards-and-standard-form-factors/com-express/com-express-basic/come-bdv7.html

 

EqvrL57.jpgdr2Wlbp.png

And then we've got something similar from DFI (i.e. @romaz's favorite brand) that's gonna accept 5VSB / 12V power

 

https://www.dfi-itox.com/products/product.php?productId=10091

http://linuxgizmos.com/type-7-module-runs-yocto-linux-on-16-core-atom-c3000-soc/

 

pt5l9hB.jpg

 

Plenty of PCIe lanes are available and we could actually pick the right option as shown below

 

https://b2b.rosch-computer.de/shop/ceb-dv970.html

http://www.dfi.com.tw/Upload/Product/Documents/DFI-DV970-COM-Express-Basic-DataSheet.pdf#page=2

Link to comment
8 hours ago, mozes said:

Two more significant boosts to SQ

So I went with scenario 1 and worked on optimizing it. The first significant boost in SQ I got is when I removed the enclosure of the 2.5inch 2TB HDD which was powered by USB3.0 from the tX-USP-exp (LPS-1). I went with the recommendation of @seeteeyou and ordered the SATA to USB3.0 adapter from Addonics. This enabled me to power the HDD from a TP-Link power bank and obviously I switched off the Vbus power on the tX-USBexp. This was a great improvement, more than switching between any of the above 4 scenarios. Thank you @seeteeyou, I am grateful! this is the link if someone is interested.

 

http://www.addonics.com/products/adsau3.php

 

I'm so glad that's working great for you, I also briefly mentioned the affordability and quality of LT3045 back in October 2017 here

 

https://www.computeraudiophile.com/forums/topic/30376-a-novel-way-to-massively-improve-the-sq-of-computer-audio-streaming/?page=152&tab=comments#comment-724439

 

Regarding USB 3.0 cables, we've got quite a few choices as listed below

 

http://www.curiouscables.com/buy.html

https://ifi-audio.com/portfolio-view/gemini3-0/

https://ifi-audio.com/portfolio-view/mercury3-0/

https://www.startech.com/Cables/USB-3.0/A-B/

http://www.audioquest.com/usb3-digital-audio/diamond

http://www.synergisticresearch.com/cables/galileo/galileodigital/digital/

https://store.wireworldcable.com/products/starlight-7-usb-3-0-audio-cables

https://www.amazon.com/Cable-connectors-Blue-Jeans-brand/dp/B004YF8KG8

https://www.amazon.com/Cable-connectors-Blue-Jeans-brand/dp/B004Z1VRZC

 

For those of us who might be interested in a very different kinda trifecta, SOtM could modify three clocks in a row like this

  1. One clock inside the 2.5" SATA drive
  2. One clock inside the USB-to-SATA adapter
  3. One clock inside tX-USBexp

Most likely we're gonna need something like Intel X25-E according to @romaz because of its low current draw, the clock is sitting between the SATA connectors and the SSD controller as shown below

 

https://www.pcper.com/reviews/Storage/Intel-SSD-730-480GB-Full-Review-Overclocked-Data-Center-SSD-Enthusiast

https://www.pcper.com/reviews/Storage/Intel-Enterprise-SSD-Roundup/Internals-Testing-Methodology-and-System-Setup

Mujqxg4.jpg

 

Some of us might be still interested in an enclosure for 2.5" drives with 5V DC power input, maybe it's possible to get the clock modified?

 

Inateck FE2006
https://www.amazon.com/Inateck-Enclosure-Support-Optimized-FE2006/dp/B00OFJ1UK6

 

OWC Mercury Elite Pro mini MEPMU3ES
https://www.amazon.com/OWC-Mercury-Bus-Powered-Portable-Enclosure/dp/B00740BJQ8

 

StarTech S2510SM12U33
https://www.amazon.com/StarTech-com-External-Hard-Drive-Enclosure/dp/B00GYR75GK

 

Century CSS25EU3BK6G
https://www.amazon.co.jp/センチュリー-シンプルBOX2-5-USB3-0-SATA6G-CSS25EU3BK6G/dp/B00OUAJV2O

 

3MTKqPK.jpg

 

BTW, I did find some USB-to-SATA adapters that should be good for more than one 2.5" drive

 

http://www.satadrives.com/usb20espcbfo.html

http://www.cooldrives.com/index.php/usb20espcbfo.html

https://www.amazon.com/Oodelay-eSATA-Port-Multiplier-Controller/dp/B00PZ7347E

https://www.amazon.com/Oodelay-eSATA-6Gbps-Multiplier-Controller/dp/B00T22JUT4

https://www.amazon.com/eSATA-SATAIII-controllers-driver-less-retain/dp/B00U9EIRMA

 

If we're looking for a REALLY low current draw, it's hard to be CF and SD cards as I mentioned before

 

https://www.computeraudiophile.com/forums/topic/30376-a-novel-way-to-massively-improve-the-sq-of-computer-audio-streaming/?page=217&tab=comments#comment-757303

 

Apparently we could also ask SOtM to modify the clocks inside card readers

 

https://www.kogan.com/usermanuals/usb-30-universal-memory-card-reader/

http://goughlui.com/2014/02/17/review-teardown-kogan-usb-3-0-universal-memory-card-reader/

i2iHCK6.jpg

 

https://www.amazon.com/USB-Gen1-All-UHS-II-Multi/dp/B01443P08K

http://goughlui.com/2017/06/01/quick-review-teardown-transcend-rdf9k-uhs-ii-usb-3-13-0-card-reader/

bH6HeW8.jpg

Link to comment
On 11/6/2017 at 12:52 AM, mozes said:

In this case, does it make sense to consider Pico ITX boards?

 

Recently I was reading quite a bit about something called MIOe from Advantech

 

http://www2.advantech.com/mio_extension/

http://www2.advantech.com/embcore/promotions/mio_brochure_0604.pdf

http://www2.advantech.com/embcore/promotions/whitepaper/Design_Benefits_of_the_MIO_Extension_Solution.pdf

 

All MIOe modules should cost less than 100 bucks

 

https://www.mouser.com/Advantech/Embedded-Solutions/Interface-Modules/_/N-5g1o?P=1yv2wobZ1yv2wde

http://download.advantech.com.tw/ProductFile/PIS/MIOE-PWR1/Product - Datasheet/MIO Extension Modules_DS(12.19.13)20131224135914.pdf

 

Industrial Pico-ITX boards MIO-2360 with Celeron N3350 could be purchased for roughly $225

 

https://ipc2u.com/catalog/mio-2360n-s1a1e

https://my.advantech.com/Product/Model_Detail.aspx?model_no=MIO-2360

http://linuxgizmos.com/advantechs-first-five-apollo-lake-sbcs-include-mini-itx-pico-itx-models/

http://advdownload.advantech.com/productfile/Downloadfile2/1-1E4FC8Q/MIO-2360_User_Manual_ed.1_FINAL.pdf#page=13

Quote

4 PCI-Express x1 Lanes
Lane 1 &2: MIOe connector
Lane 3: Full-size Mini PCIe connector
Lane 4: Intel i210 GbE controller

 

http://advdownload.advantech.com/productfile/Downloadfile2/1-1E4FC8Q/MIO-2360_User_Manual_ed.1_FINAL.pdf#page=21

Quote

 

Interface
SMBus, 2 x USB3.0, LPC, 2 x PCIe x1, line out, DisplayPort/HDMI*, +5 Vsb/+12 Vsb power, Power On, Reset, SATA*.

 

Total peripheral power supply output
5V @ 3A for CPU board and MI/O Extension module totally, 12V @ 2A for MI/O Extension module.

 

 

I could count three (top side = 1 and bottom side = 2) clocks when I was looking at page 22

 

http://advdownload.advantech.com/productfile/Downloadfile2/1-1E4FC8Q/MIO-2360_User_Manual_ed.1_FINAL.pdf#page=22

 

There's another $80 evaluation board MIOE-DB2100 that's designed for a different model MIO-3260 with Celeron N2930

 

https://www.mouser.com/ProductDetail/Advantech/MIOE-DB2100-00A1E/?qs=OCoHSBuAoyt9N2SSzkMOzg==

http://www.advantech.com/products/mi{{--o_extension_evaluation_board/mioe-db2100/mod_f7f2d694-863b-4bc2-b67b-00b31359f4f4

http://download.advantech.com.tw/ProductFile/Downloadfile1/1-X5X9S6/MIOe-DB2100_User_Manual_Ed.1-FINAL.pdf

 

We could power MIO-2360 and the expansion board separately so that could be a bonus, not to mention that both of them don't come with 5.5-mm DC power connectors (i.e. high impedance) by default.

 

Now the question really is how much is a custom-made expansion board that's compliant with their specifications

 

https://www.mouser.com/pdfdocs/AdvantechMIOExtensionSpecification_V10.pdf

Link to comment
5 hours ago, LTG2010 said:

Its a celeron j900 10W motherboard, theres a  DC input to the motherboard, one to the hard drive and one to the 'Pico' type dc - dc connector which seems to power the optical drive. Those are Mundorf Capacitors on the regulator board- very expensive, it apparently ran @ romaz set up pretty close, so it would be very interesting to hear @ austinpop's evaluation.

 

That's quite an impressive achievement without any sCLK-EX / Paul Hynes SR7 goodness whatsoever, maybe we could actually learn a thing or two from their magic sauce PSU?

VJRNx8A.jpg

 

http://www.monoandstereo.com/2017/05/new-innuos-zenith-se-music-server.html

Quote

 

With a new Triple-Linear PSU designed by Sean Jacobs, custom-wound toroidal transformer and a raft of improvements on EMI shielding and vibration isolation

  • New oversize, custom-wound audiophile toroidal transformer with 3 independent outputs for each Linear Power Supply
  • New Triple-Linear power supply designed by Sean Jacobs using high-end Mundorf caps
  • Three anti-vibration feet with asymmetrical positioning to better isolate vibration without a ecting dynamics
  • Audiophile-grade silver-coated and shielded cabling for internal connections, hand-soldered directly to circuits to lower impedance from connectors.

 

 

Here's Supermicro X10SBA-L with Celeron J1900

 

http://www.supermicro.com/products/motherboard/celeron/x10/x10sba-l.cfm

 

Here's a different PSU from CHC

 

http://www.custom-hifi-cables.co.uk/home/power-supplies/dc3-power-supply

Quote

 

£520.00
3 to 25VDC (fixed)
1.5A max. current
& 1m output cable

 

Upgrade to Mundorf
capacitor arrays:
add £180

 

High Current Boost
Module (up to 5A):
add £150

 

Extra separately-
regulated outputs:
add £120

 

BTW, here's yet another OCC silver PSU

 

https://item.taobao.com/item.htm?id=534199855459

 

Finished products with Mundorf caps

 

https://item.taobao.com/item.htm?id=530916796794

https://item.taobao.com/item.htm?id=562470157045

 

So yeah, we've got something better than the ones from HDPlex and Paul Pang etc.

 

http://www.chiphell.com/thread-1378630-1-1.html

https://item.taobao.com/item.htm?id=531993464123

https://item.taobao.com/item.htm?id=538554836020

 

EFarsC2.jpg7FRwgvJ.jpg

Link to comment
7 minutes ago, Cornan said:

 

Very interesting linear power supply indeed! I will most definately add this to  my opt's except sPS-500 & SR4. Thanks for sharing! ? I found it here as well  https://www.yoycart.com/Product/41032527596/

BTW is it possible to somehow translate the whole tabao site?

 

So much would be lost in translation when machine translation could be a disaster of some sort.

 

That yoyocart stuff really ripped everyone off because it's a significant cheaper silver-plated copper version, actually stuff like that would only cost 2,780 RMB or just under 430 bucks

 

https://item.taobao.com/item.htm?id=41032527596

 

Actually Kenneth Lau also sourced his OCC transformers from Simon Tuned, maybe it's a better bet to go that route instead?

 

https://www.computeraudiophile.com/forums/topic/18477-handcrafted-audiophile-linear-power-supply-by-kenneth-lau/

 

Anyways, maybe at least part of the "magic" might be coming from those insane capacitors even though the output impedance might not be all that impressive when compared to the numbers we're getting from Paul Hynes.

 

Both Simon Tuned and Kenneth Lau didn't seem to provide any kinda numbers whatsoever, and then most likely we wouldn't expect anything from CHC as well.

 

Since Vinnie Rossi stopped making MINI PURE-DC-4EVR already, our PSUs of choice would be left with LPS-1 / Paul Hynes / SOtM etc. in addition to these FINEMET core transformers from Japan

 

https://samizuacoustics.com/about/

https://www.ebay.com/sch/fulken/m.html

http://noguchi-trans.co.jp/digitalcatalog/main/fs801.pdf

Link to comment

Just an FYI - we could multiply some numbers by 6 as shown in the picture of Triple-Linear PSU designed by Sean Jacobs

 

http://www.mundorf.com/en/?category=hifi&menu=caps_audio&content=mcap_supreme_evo_silvergold_oil

Quote

Since 2015, the MCap® SUPREME EVO SilverGold has been our undisputedtop capacitor model. It combines all our technical highlights, such as SilverGold metallisation, SUPREME and EVOLUTION winding technology and over 30 years of experience in the manufacturing of audiophile masterpieces into a unique capacitor type of equally unique musicality: finely nuanced and lively shining acoustic colours, which combine into a breathtaking, beautiful, live-like music experience with the utmost precision, presence and three-dimensional plasticity.

 

Mundorf MCap Supreme Silver Gold Capacitors ($38.28 to $233.88)

https://www.hificollective.co.uk/catalog/-c-61_68_190.html

http://www.partsconnexion.com/capacitor_film_mundorf_sg.html

 

Mundorf MCap Supreme Silver Gold Oil Capacitors ($45.48 to $245.88)

https://www.hificollective.co.uk/catalog/-c-61_68_313.html

http://www.partsconnexion.com/capacitor_film_mundorf_sgo.html

 

Mundorf MCap Supreme EVO Silver Gold Oil Capacitors ($21.48 to $359.88)

https://www.hificollective.co.uk/catalog/mundorf-mcap-supreme-evo-silver-gold-oil-capacitors.html

http://www.partsconnexion.com/capacitor_film_mundorf_evo_sgo.html

 

We just dunno what caps Innous opted for powering their ZENith SE.

Link to comment
20 minutes ago, RickyV said:

Wrong caps

 

Really? I was going by this one mentioned below

 

http://www.corsair.com/en-us/blog/2013/december/power-supply-capacitor-q-and-a

Quote

Other capacitors used in computer power supplies are “metalized polypropylene” capacitors, or “film capacitors”. These are generally used for EMI filtration on the AC input of a power supply.

 

Could music servers actually go anywhere close to several hundred amperes?

 

https://www.hificollective.co.uk/components/mcap_mlytic_hc.html

Quote

These MLytic® HC electrolytic capacitors are designed for continuous current of up to several hundred amperes.

 

Link to comment

I've been looking for DDR3L memory with lowest current draw out there, so far Samsung seemed to be the winner with every single number that's kept under 1,000 mA

 

M471B5674QH0 : 2GB (256Mx64) Module

http://www.samsung.com/global/business/semiconductor/file/product/135V_DDR3_4Gb_Qdie_UnbufferedSODIMM_Rev121.pdf#page=28

 

Samsung 2GB PC3-12800 DDR3-1600MHz Non-ECC CL11 204pin Memory M471B5674QH0-YK0
https://www.amazon.com/Samsung-PC3-12800-DDR3-1600MHz-Non-ECC-M471B5674QH0-YK0/dp/B00TUBT2KY

 

Samsung 2GB 1Rx16 PC3L-12800S Laptop Memory Ram M471B5674QH0-YK0 Tested
http://www.ebay.com/itm/Samsung-2GB-1Rx16-PC3L-12800S-Laptop-Memory-Ram-M471B5674QH0-YK0-Tested-/253324257430

 

I checked everything on Micron and Samsung etc. but pretty much all of them would exceed 1,000 mA. Of course our winner is compatible with Bay Trail (e.g. Celeron J1900 inside ZENith SE) as well as Braswell (e.g. Celeron N3060 / N3160 etc.) that could be found in many motherboards

 

https://ark.intel.com/products/78867/Intel-Celeron-Processor-J1900-2M-Cache-up-to-2_42-GHz

https://ark.intel.com/products/91832/Intel-Celeron-Processor-N3060-2M-Cache-up-to-2_48-GHz

https://ark.intel.com/products/91831/Intel-Celeron-Processor-N3160-2M-Cache-up-to-2_24-GHz

 

BTW, let's see if we're able to bypass all regulators and power DDR3L memory directly with LT3045 since we're keeping everything under 1,000 mA

 

http://www.ebay.com/itm/LT3045-A-Ultralow-noise-0-8-Vrms-LDO-linear-regulator-0V-15V-1A-Fixed-out/253298389543

 

MoBo Memory Power Bypass

http://www.tirnahifi.org/forum/viewtopic.php?f=15&t=3267

http://www.aktives-hoeren.de/viewtopic.php?f=30&t=6015

Link to comment
33 minutes ago, beautiful music said:

But Zenith SE raised a curiosity what is the specification of it's linear power supply.

 

Does it surpasses the Paul Hynes SR7 performance!!!!!!!

 

ZENith SE should be limited to 100 units and they're down to last 10 or so already.

 

It's gonna be a stretch but one of us with ZENith SE plus either SR4 or SR7 might give this a try. Turn off ZENith SE and remove the cover. Disconnect the SATA power cable from the Samsung SSD inside ZENith SE and connect that to another SSD inside one of those spaghetti solutions. Turn ZENith back on and pray that nothing bad would happen. Now get back to the spaghetti solution and play a few familiar tracks. Then power everything off and then play the same tracks again with either SR4 or SR7 afterwards.

 

Granted we're still comparing a different DC power cable from SR4 / SR7 to SSD versus another cable from that CHC Triple-Linear LSU, though it's better than nothing.

 

Or someone in UK could also try that DC3 linear PSU from CHC for 30 days and see if that were really MUCH better than LPS-1 etc.

Link to comment
5 hours ago, jean-michel6 said:

Thank you .... Very interesting reading as i am in the process to develop a new server using this super micro board .

 

You're very welcomed and that sounds exciting.

 

BTW, here's a little tip from Roy and maybe you'll find it helpful

 

https://www.computeraudiophile.com/forums/topic/30376-a-novel-way-to-massively-improve-the-sq-of-computer-audio-streaming/?page=50&tab=comments#comment-647362

Quote

One other area I got to test was 2GB vs 4GB of RAM.  While the difference wasn't huge, 2GB for sure sounds better and I will see if I can get 2GB to work out.

 

He also compared 2GB memory modules to 8GB ones below

 

https://www.computeraudiophile.com/forums/topic/30376-a-novel-way-to-massively-improve-the-sq-of-computer-audio-streaming/?page=47&tab=comments#comment-644453

Quote

2GB ECC DDR3 SDRAM UDIMM (1.5V)
Active standby current - 279mA
Operating burst read current - 675mA
Operating burst write current - 711mA
Burst refresh current - 1611mA

 

If you're able to run 2GB of RAM at 1,066 MHz, definitely get HMT425S6CFR6A from SK Hynix (it's like $9.95 a piece on eBay now) and you're gonna get much lower (maybe it's pretty much the lowest?) current draw to further reduce the noise

 

https://www.skhynix.com/products.view.do?vseq=1740&cseq=75

Quote

2GB, 256M x 64 SO-DIMM: HMT425S6CFR6A

Active standby current - 112mA
Operating burst read current - 368mA
Operating burst write current - 380mA
Burst refresh current - 520mA

 

Link to comment
3 hours ago, jean-michel6 said:

- minimize the ram size.

-.....

I am now using one 2gb memory module on the X10SBA. I am using crucial memory modules. 

I will look in the memory you recommend as lowest current draw. From my reading I got the info that Samsung memory modules were very energy efficient.

 

The ones from Micron should be a joke because the current draw seemed to be nothing to write home about. Samsung ain't too bad the and lowest numbers of M471B5674QH0 are kept under 500mA. SK Hynix should be fantastic and HMT425S6CFR6A would give us numbers that are belowe 400mA. I just kept looking but Crucial might not be willing to publish any numbers or something.

 

It ain't an exhaustive list by any means, just giving everyone a glance at what kinda DDR3L memory modules we could get on the market right now

 

http://www.samsung.com/global/business/semiconductor/file/product/135V_DDR3_4Gb_Qdie_UnbufferedSODIMM_Rev121.pdf#page=28

Quote

M471B5674QH0 : 2GB (256Mx64) Module

Active standby current - 110mA
Operating burst read current - 465mA
Operating burst write current - 440mA

Burst refresh current - 570mA

 

http://www.samsung.com/global/business/semiconductor/file/2011/product/2011/9/2/412764ds_ddr3_2gb_d-die_based_sodimm_rev14.pdf#page=20

Quote

M471B5773DH0 : 2GB (256Mx64) Module

Active standby current - 110mA
Operating burst read current - 520mA
Operating burst write current - 560mA
Burst refresh current - 880mA

 

http://www.samsung.com/global/business/semiconductor/file/product/ds_ddr3_2gb_c-die_based_1_35v_sodimm_rev131-0.pdf#page=23

Quote

M471B5773CHS : 2GB (256Mx64) Module

Active standby current - 360mA
Operating burst read current - 720mA
Operating burst write current - 800mA
Burst refresh current - 1320mA

 

https://www.skhynix.com/products.view.do?vseq=1740&cseq=75

Quote

2GB, 256M x 64 SO-DIMM: HMT425S6CFR6A

Active standby current - 112mA
Operating burst read current - 368mA
Operating burst write current - 380mA

Burst refresh current - 520mA

 

https://www.skhynix.com/products.view.do?vseq=1742&cseq=75

Quote

2GB, 256M x 64 SO-DIMM: HMT425S6CFR6C

Active standby current - 124mA
Operating burst read current - 392mA
Operating burst write current - 424mA
Burst refresh current - 520mA

 

https://www.skhynix.com/products.view.do?vseq=1946&cseq=75

Quote

2GB, 256M x 64 SO-DIMM: HMA425S6AFR6N

Active standby current - 136mA
Operating burst read current - 586mA
Operating burst write current - 626mA
Burst refresh current - 728mA

 

https://www.skhynix.com/eolproducts.view.do?pronm=DDR3+SDRAM&srnm=HMT325S6BFR8A&rk=20&rc=module

Quote

2GB, 256M x 64 SO-DIMM: HMT325S6BFR8A

Active standby current - 200mA
Operating burst read current - 560mA
Operating burst write current - 560mA
Burst refresh current - 1120mA

 

https://www.skhynix.com/eolproducts.view.do?pronm=DDR3+SDRAM&srnm=HMT325S6CFR8A&rk=20&rc=module

Quote

2GB, 256M x 64 SO-DIMM: HMT325S6CFR8A

Active standby current - 160mA
Operating burst read current - 520mA
Operating burst write current - 520mA
Burst refresh current - 880mA

 

https://www.skhynix.com/eolproducts.view.do?pronm=DDR3+SDRAM&srnm=HMT325S6EFR8A&rk=20&rc=module

Quote

2GB, 256M x 64 SO-DIMM: HMT325S6EFR8A

Active standby current - 144mA
Operating burst read current - 424mA
Operating burst write current - 440mA
Burst refresh current - 1280mA

 

https://www.skhynix.com/eolproducts.view.do?pronm=DDR3+SDRAM&srnm=HMT425S6AFR6A&rk=20&rc=module

Quote

2GB, 256M x 64 SO-DIMM: HMT425S6AFR6A

Active standby current - 104mA
Operating burst read current - 364mA
Operating burst write current - 400mA
Burst refresh current - 800mA

 

https://www.skhynix.com/eolproducts.view.do?pronm=DDR3+SDRAM&srnm=HMT425S6MFR6A&rk=20&rc=module

Quote

2GB, 256M x 64 SO-DIMM: HMT425S6MFR6A

Active standby current - 120mA
Operating burst read current - 500mA
Operating burst write current - 500mA
Burst refresh current - 560mA

 

https://www.skhynix.com/eolproducts.view.do?pronm=DDR3+SDRAM&srnm=HMT351S6BFR8A&rk=20&rc=module

Quote

2GB, 256M x 64 SO-DIMM: HMT325S6BFR8A

Active standby current - 200mA
Operating burst read current - 560mA
Operating burst write current - 560mA
Burst refresh current - 1120mA

 

https://www.skhynix.com/eolproducts.view.do?pronm=DDR3+SDRAM&srnm=HMT351S6CFR8A&rk=20&rc=module

Quote

2GB, 256M x 64 SO-DIMM: HMT325S6CFR8A

Active standby current - 160mA
Operating burst read current - 520mA
Operating burst write current - 520mA
Burst refresh current - 880mA

 

https://www.skhynix.com/eolproducts.view.do?pronm=DDR3+SDRAM&srnm=HMT351S6EFR8A&rk=20&rc=module

Quote

2GB, 256M x 64 SO-DIMM: HMT325S6EFR8A

Active standby current - 144mA
Operating burst read current - 424mA
Operating burst write current - 440mA
Burst refresh current - 1280mA

 

https://www.skhynix.com/eolproducts.view.do?pronm=DDR3+SDRAM&srnm=HMT451S6AFR6A&rk=20&rc=module

Quote

2GB, 256M x 64 SO-DIMM: HMT425S6AFR6A

Active standby current - 104mA
Operating burst read current - 360mA
Operating burst write current - 400mA
Burst refresh current - 800mA

 

https://www.skhynix.com/eolproducts.view.do?pronm=DDR3+SDRAM&srnm=HMT451S6MFR8A&rk=20&rc=module

Quote

2GB, 256M x 64 SO-DIMM: HMT425S6MFR6A

Active standby current - 120mA
Operating burst read current - 500mA
Operating burst write current - 500mA
Burst refresh current - 560mA

 

https://www.skhynix.com/eolproducts.view.do?pronm=DDR3+SDRAM&srnm=HMT312S6DFR6A&rk=20&rc=module

Quote

1GB, 128M x 64 SO-DIMM: HMT312S6DFR6A

Active standby current - 100mA
Operating burst read current - 420mA
Operating burst write current - 400mA
Burst refresh current - 400mA

 

And then we've got some "monsters" from a few companies, unfortunately they're just boasting "low power consumption" without bothering to back that claim up with a datasheet of some sort

 

http://industrial.adata.com/us/product/317

https://www.transcend-info.com/Embedded/Products/No-589

http://www.advantech.com/products/memory_module/sqr-sd3i/mod_c55cd924-ea83-43e3-962a-04ef161594e1

Link to comment
3 hours ago, jean-michel6 said:

There is also an important parameter to consider I think is the latency ?

 

It's more likely to be a big deal for gamers according to this post

 

https://www.computeraudiophile.com/forums/topic/13948-article-computer-audiophile-pocket-server-caps-v3-lagoon/?page=5&tab=comments#comment-192067

 

Actually Roy was throttling the speed of both CPU and RAM to 800 MHz so they're synchronized to the same frequency. And then we could also tweak the PLL with ClockGen

 

http://www.sisoftware.eu/faq/pll-ic-clock-generator/

http://www.softpedia.com/get/Tweak/CPU-Tweak/ClockGen.shtml

 

Maybe we could dial something down a notch for the sake of lower power consumption?

 

BTW, someone is gonna share something to slim Windows 10 down with NTLite soon

 

https://www.ntlite.com/community/index.php?threads/w7-and-w10-presets-downloads.135/

https://www.ntlite.com/community/index.php?threads/w7-and-w10-presets-questions.154/

 

And then there's also another one here

 

https://www.ntlite.com/forum/discussion/1364/tutorial-for-creating-a-700-mb-windows-10-iso-and-install-in-a-vm

Quote

Windows 10 LTSB 1607 (32bit) NTliteTiny2 - 1.2 GB (no swap)

 

In that case, Windows 10 could be installed on SD cards with the lowest power consumption (28mA read / 55mA write) ever

 

https://www.mouser.com/access/?pn=922-603989

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...