stevv

two dead kill a watt for dummies after a year just plugged in to wall.  got them $15 each. 

saw these meters on the ltt show, anybody know what they are?

Well i only cut off everything above the capacity which are the same as in your screenshot. Some of the information im not getting. I guess you have a newer firmware or something. Don't think i really need that as the card is performing fine at the moment. Unless you say its really worth it, and i won't lose data because of an update. You have an 1880i or a newer version?

 

And the thing about the cables, i have no idea if that is true. I think i bought good ones, not really sure what brand they are though, doesn't say it on the cables themselves. I think its 3Ware, but i'm not sure. Which brand of cable is good? If it happens again i might look at some better cables.

I have the 1880i  with Firmware Version V1.52.  I've updated firmware a couple of time without issues to the arrays.  I wanted the TLER controls and 1024KB stripes.. Though I've didn't have an major issues even back in 2011 when I first got it, so not a critical update.

If you want to test the cable, you can swap the position of the cables-to-drives (after shutting down first).. that's what I did when my own drive was failing a couple of times... nice thing the card can handle drive reorder without a hitch.. hence this mess in the drive orders lol.

Oh man, I thought the first part of the build was a joke for "what not to do" in a server type build... and then it's all for a shroud... Interesting nevertheless :lol:
 
Except that the shroud would still work perfectly fine with a proper server motherboard that's compatible with that heatsink and backplate.. and even better since the ram won't get in the way of the cooling.
 
 

Socket 1155 and 1156 are physically the same.

 
Yeah, I use that same 1156 heatsink with on my supermicro board with 1155 chip in 1U and it's perfectly free of any obstructions top and bottom.  That intel board is definitely not designed with that heat sink in mind lol
 
Pictures... in case anybody else who might want a 1U unit... hopefully they would have less casualties lol.

Here i am again. Quick question which i don't really think needs its own topic... let me know if its better if i do make a new topic for it.

 

My 40TB array degraded 2 days ago. I've tested the drive that failed and it seems fine. The log showed it gave a timeout error. Is this drive actually dying? I've never had much of an idea of what the SMART stuff means but it seems fine

 

RAID-6-Degraded-Drive.png

 

I've got everything backed up so im in no way concerned, it's also a RAID-6 so another one can fail and all the data would still be fine. Should i prepare for this drive to fail more often or is something else wrong? Is it just some glitch or something that made the drive fail? When i looked before all the SMART stuff was just "-" and no numbers. I don't want to send in a perfectly fine drive for RMA

 

edit: i've got the drive back in and the array is rebuilding...

My last Red 4TB  errored ("read error" instead of "time out")  out 2-3 times but tested ok after "repair" and extended test using the wd diag tool

Finally took to RMA... last July and working fine since.

I got the same card you have.. but assuming, hardware is good... the drives don't always test out accurately.  Now I always give the drives a good exercise before adding to array.. like a few weeks of random tasks.

other note... did you truncate the screenshot of the drive information?  I get a lot more information for my 4TB drives.  Included screenshot .. hope that helps.  Interesting that your 5TB's are 5700RPM...

No backlight, no deal.

Same here.  

In the meantime, back to using the K800.  

Would be nice to have a secondary battery in case battery is low, and shuts off light first.

[snip]

• 2 x Crucial 16GB Kit (8GBx2) DDR3L 1600MT/s (PC3-12800) DR x8 ECC UDIMM 240-Pin

[snip]

 

CPU/MB/RAM

I decided to go with a full socket 1150 motherboard vs some of those Intel Atom setups that are popular. I found a combo deal on newegg which combined the Supermicro X10SL7-F-O and an Intel Xeon E3-1270V3 (Haswell). I added 16GB (2 x 8GB) of ECC RAM and I plan to increase that to 32GB shortly which is why I listed 32GB in the build list below. I decided to go with a bit more CPU power than I originally planned because I wanted to be able to give enough to ZFS and still use some extra for processing other work in the future. I'll be running some kind of Samba/NFS/CIFS to transport data to other systems in my house. I also plan to run some media server components once I've digitized movies into this NAS....

[snip]

Not sure if the crucial's you got would have the same issue, but for me the difference was the "TS" in kingston modules that caused minor issues.

Tests and runs fine but can get issues on specify memory intensive applications.  In my case it was a vbox VM and couldn't get it past 8GB (of 16 available).  Other apps and tests ran fine for over two years for two systems.  The third system, I got verified memory, and been running 32GB since mid last year 24/7 with a handful of VM's.  I run the X9's of the supermicro boards.

I just came to ask the same question. I think that gpu OUTPUTS high resolution so your SCREEN should downsample. So our screens dont support downsampling. if i understand this correctly in order for it to work gpu should RENDER 1440p resolution then DOWNSPAMPLE it to 1080p and only after OUTPUT 1080p (note: these resolutions was chosen for my setup i have 1080p screen and i want to render games at 1440p). I am sure that this is exactly how it works on new nvidia 970 and 980 cards. Please correct me if i am wrong, i am new to this.

P.S. I have r9 290 from asus

P.S.S. Sorry for my english, its not my native.

edit: just realized that even youtube can downsample unlike my gpu (if i choose 1440p or higher in youtube video)

The current nvidia's method, uses the hardware scaler on card (theoretically, any monitor should work, but it doesn't).  Nvidia's new "DSR" (dynamic super resolution) is supposed to do what your are referring to, but done on through software filters so it's more flexible and can generally result in better quality scaling but some performance hit (Although relatively little).  Most importanly, it's an officially supported feature, so hoping this will allow any monitor to work... especially those with no scalers (like my korean 1440p monitor) or maybe keep my 120hz@1080 while downsampling.  Reason to get a 980 even if you have a 720P monitor! lol

Somewhere (13 - 20 min) in the live stream explains is =).

http://www.pcper.com/news/Graphics-Cards/PCPer-Live-NVIDIA-Maxwell-GTX-980-GTX-970-Discussion-Tom-Petersen-QA

A. Rebuilt my sister's netbook on my bed.

B. Will be building my new rig on my bed.

C. Rebuilt my Dell on my bed.

D. Build ZEUS on my bed.

E. Rebuilt ZEUS on my bed.

 

See where I'm going with this?

Your bed is the floor?  With shaggy carpet.  B)

You look like you are over complicating it.... If you want to see if how much heat can this setup transfer out the case, you just need to know the heat properties of air, which are the Heat transfer coefficient of air and the thermal coefficient. Once obtained, you can say if X amount of CFM enters and leaves the radiator tower, and the heat transfer coefficient of the radiators is Y, then how much of it will be dissipated of the TDP the system is generating. This is the scientific way to do it, and it's much more easy to understand.

 

IMO, he just has to try with more powerful fans (he went with the low RPM version of the eloops). The Silverstone AP look like they have a ton of potential here, thanks to the wind tunnel they exhaust.

They be good, especially the ap123

Looking at these numbers, the AP123 would be borderline performing at best. 

HA!  I'll show you mine if you show me yours. 

 

snip* 

get cherry height/profile keycaps as they are slightly shorter than the OEM height/profile corsair ones. You wont be able to get the for your K95 though as it has non standard bottom row key spacing.

Thats one main difference between a scissor switch board which both the thinkpad and logitech are and a mechanical board. Another option might be some DSA profile caps since your used to them all being the same height. these can only be gotten from groupbuys on places like Geekhack.org, massdrop, and pimpmykeyboard.com (direct manufacturer). but they will once again not fit your K95.

Thanks, I'll have to check them out.

Right now, the K95 went to my brother, who is perfectly content with it and two finger typing =).  The blue backlight is a bit blinding.

The K800 is serving me nicely, but I might get a mechanical keypad for gaming and other "hot key" purposes in the future. pimpmykeyboard... nice  B)

You need to show the maths to back up a statement like that.

Well, you got me.. I'm pulling this out of my rear.  

The short: That's how much air your system of four fans would be pushing if it weren't stacked.  (33cfm@800rpm x4 ...  take away some from radiator fin resistance... ~100+ CFM ).  

This.  Where is this magic "100+ CFM" figure coming from?  There are far more important variables involved than CFM.  Ambient temp, radiator capacity, flow rate, static pressure, and system wattage all come into play to determine final temps.

 

Sorry to hear about the pump.  Been waiting for an update to this build for months, so I'm happy to see it getting close to finished.  I hope the new fans sort out the temperature issues.  Keep an eye on those coolant temps, it can cause more problems than a dead pump when you start seeing temps that high.  Good luck man!

Magic?   You can call out variables all day, but we're not talking about a setup with no data or examples or that we are doing very fine tuning.  Realistically, this system is clearly starved for cool air, counter rotation isn't going to do diddly.  One just need a ballpark number, and there's a decent amount of data out there, including those from the manufacturers, and not to mention similar systems.  Unless the machine is going to be in a closed closet or in a cooled data center... room temp is 25c, and medium pump speed.

Seriously, just look at the set up..  under 33 cfm of room temp cool air over a single intake, and overclocked cpu and 2 gpu?  That would certainly be a lot more efficient than the typical setup that most people have out there as well as what's the actual manufacturers are designing.   If so, that would be awesome.

disclaimer:  My "magic number" is just a approximate guess using examples available.  The best thing is just put better higher quality fans that can handle higher RPM when needed.  (if you worried about noise, just turn the speed down!). 

 

 

 

As mentioned before, treating it as simple systems.  heat source: water (yes, because we assume the target is for the system at max temp and at thermal equilibrium), conductor: radiator, heat sink: air. 

 

To get a ballpark requirement, 

-using data from thermochil HE 120.1

-room temp is 25c

-medium pump speed. 

 

For thermal resistance, or C/W (having the units °C/W):

 

C = [(Tinlet coolant + Toutlet coolant) ÷ 2] – Tinlet air      (all in °C)

 

and

 

W = m * Cg * (Toutlet – Tinlet)

where

  W = Watts

  m = mass flow rate, kg/sec

  Cg = specific heat capacity, for H20, 4186 J/(Kg°C)

 

This reduces to:

 

W = (lpm * 4186 * ΔT) ÷ 60

 

2: plug  and cross ref graphs 1,10,11.  (fans used are ~60 and 150CFM)

http://www.overclockers.com/an-evaluation-of-the-thermochill-radiators/

http://www.swiftech.com/Resources/White_Papers/Assessment%20of%20Radiator%20Performance.pdf

 

3: adjust for Running warmer air over the same area of radiator vs running room temp air through all radiators.  (temp in / temp out... log temp reduction in water, increase in air in each section))

 

http://www.techspot.com/community/topics/the-watercooling-guide-from-a-to-z.180876/

http://www.overclockers.com/radiator-heat-dissipation-testing/

 

 

 

What do you do about cable management. It always gets in the way. Has anyone got any ideas?

 Good o'l velcro.

 

Slight hiccough with the testing. But here's some temp results gathered so far for an i7 3930k (6cores/12threads) and 2 GTX 770s SLi:

 

Stock settings under Prime 95 for 30 minutes:

• idle 38-42°

• load 68-81°C

Overclocked to 4.5GHz under Prime 95 for 30 minutes:

• idle 49-52°

• load 78-91°

Overclocked to 4.5GHz under Unigine Heaven for 30 minutes:

• CPU 68-72°

• GPUs 51-52°

 

Overclocked, the CPU get's quite toasty. But stays under the throttling temperature (95°C), albeit by a couple degrees. Running at stock speeds (3.4 turbo) only saves 10°C at both idle and load.

 

Now onto the problem: the fans.

 

I chose 800rpm eLoops because they are as close to silent as you can get from a fan. They are silent because they barely spin. Because they barely spin, they hardly move any air.

Normally, when you finish a bench mark/stress test, the temps rapidly drop down 20-30° then then slowly level out. With the eLoops, the temps stayed stupidly high and only dropped very slowly. And when I tried to run a couple of graphics tests, things got warm. A little too warm. So warm, in fact, the fluid reached over 60°C. Now, I don't have any temp sensors in my fluid, so how could I possibly know that? Well, DDC pumps have an operating temperature of <60°C.

 

Yup. I fried my pump.

 

So this weekend, instead of getting some pictures of my finished build, I will be stripping down my whole system - partly because the pump is completely hidden underneath the inner part of the case - partly to swap out the eLoops for some Silverstone FQ121s, which look an awful lot like gentle typhoons…

 

For an overclocked 300+ watt of heat in the water system, you'll need to move 100+ CFM of "room temp" air to keep it under 80 degrees (or water around 40 degrees).  

I just made a new PC all thats left is for me to boot windows on it. I have a pny usb 8gb I have the ISO file on my usb i used windows 7 downloaded but whenever i plug in my usb it doesn't show up and wont let me boot i plugged in my mouse and went to board explorer and it shows up but when i plug in my flash drive it doesn't. I dont know what to do. Help?

You might need to make a UEFI boot usb, or somehow disable UEFI boot for win7 install.

Keycap comparison: Thinkpad (x301) vs Corsair K95 key.

I found the K95 keys too narrow, and my finger easily slip between keys when the center of my finger is actually still about a 1/4 from the edge of the key, and the wobbly switches doesn't help either.  Any thoughts?  I'm currently using a Logitech K800.

The Thinkpad 8 getting charged by portable (Anker® 2nd Gen Astro3) battery while watching youtube HD 1080.  Charged 3% in 10min.  The battery could charger over 2amps, but I didn't have a meter when I was testing it.

you mean like this one?  Thinkpad S1 Yoga.  I got the same specs i5 4200u / 8GB.  This is for my nephew in college.

Swapped the HDD for the new Samsung 850 pro.  

Quick rundown:

• Balance weight distribution.
• Solid (doesn't flex even holding at the corner)  
• The Matt Touch screen and bezels is a plus.  
• The keyboard flattens out when in tablet mode.  
• The touchpad physical "click" is an improvement over the last models. It's more like a big space bar with travel and a tad bit more resistance than a hard "snap" click.  
• The keyboard is also an improvement over a last gen "island".  It's solid (as in it doesn't wiggle, or "sound hollow" when typing. As an old school thinkpad keyboard lover, this a still not "perfect" but still the best since the new island keys.  
• Pen is useful (pressure sensitive), but design could be better (maybe an aftermarket one might work).  
• Speakers, are adequate.
• rotation detection is responsive (not buggy or laggy).
• I'm still testing out the sleep/battery drain, but my experience with the 4th gen chips, I expect under 10% / day. 
• Hibernation resume is fast... half a turn of the windows boot.  (might be because of the 850 pro , compared to the 850 EVO I had in a XPS 12).
• anything else?

This is compared to the DELL XPS I had , but sold.  The XPS 12 was an awesome laptop that would be a keeper too (it was and i7/8GB).  Oh yeah... will miss that awesome woven carbon fibre top and bottom.  I think the best feeling surface of all laptops.  Grippy and decently fingerprint resistant. 

Got a 4TB RED down!  I revived it, but lasted another month then crapped out. 

RIP lil guy... 

They replaced with this... a bit disappointing!  They even removed any traces of the word "RED"  :rolleyes:  

some notes:

-did an advanced shipment with a debit card(hold without charge) and bought return label with WD's service (cheaper!).  Get drive faster, test, and save time and $ for box and return shipping.

The copy picture and pastes as text looks very useful.

Currently I use:

PrintScreen key, or Windows Snippet (automatically in clipboard). => paste irfanview, edit => save to google drive (automatically in google photo's).  ==> view in google photo, resize window, copy picture link.  The shared image will be that size.

Nice write up! Is the definition of  "PC Components" in this thread limited to internal components?  

The screen and other external components come to mind, here are my thoughts from my experience.  

• I keep both 70% and 99% isopropyl alcohol around.  
• I find the 70% is better for surfaces that like screens, keyboards, mouse, laptops, phones... etc, as the water content helps dissolve other solids like food and dirt better.  Since these are external components, and more often then not, would need cleaning.  So long as there's, some common sense... not drowning area or using soaked and dripping rag/towel.  
• Isopropyl alcohol + microfiber cloth doesn't stink up like Windex after a few uses.
• For some stubborn thermal compounds, I use the combo, letting 70% sit for a minute, wipe off and 99% to finish.
• Other internal components, only soft dry toothbrush and compressed air (with no moisture).  Sometimes just very lightly moistening it with 99% alcohol.
• Nail polish remover great for smoothing out plastic after the dremel and file lol.
• Cleaner solvent for "sticker/adhesive" goop? 

People who buy the cheapest ($300-$400) "PC" for themselves or their kids....  then buys a $1000+ mac a few years later, saying they are better and faster.

edit:  Same with phones 

APC BR1500 + external battery add-on. Total 865W limit.

Server, desktop and network. Idle around 250w.

Summer, A/C and vacuum cleaner doesn't mix well lol.

Gave me power for led light during hurricane sandy too !

completed new build!!! its called "baby blue machine"

yeah... in that case..  cyan sound better lol

I had been looking around how to make these headphones more comfortable and came across a site that has velour/velvet earpads made specifically for them. Although the price is fairly steep, I've heard great things about them. Anybody have experience with them?

http://www.soundprofessionals.com/cgi-bin/gold/category.cgi?item=AT-135409020-VELVET

 

Thanks,

Bacon

I have them and they are a exact fit from the originals.  It's worthwhile if the originals are uncomfortable.

http://linustechtips.com/main/topic/137702-buying-audio-technica-ath-m50/?p=1830061