Please report all spam threads, posts and suspicious members. We receive spam notifications and will take immediate action!
Page 2 of 2 FirstFirst 12
Results 11 to 19 of 19

Thread: PSU choice for my Gigabyte board: stability & low noise?




  1. #11
    Join Date
    Dec 2008
    Posts
    42

    Default Re: PSU choice for my Gigabyte board: stability & low noise?

    Ripple and crosstalk can have a detrimental effect on components, especially power filtering ciruitry. If you were to just run stock all the time you would probably never notice, but overvolting ang clocking brings out weaknesses in components.

  2. #12
    Join Date
    Apr 2008
    Posts
    942

    Default Re: PSU choice for my Gigabyte board: stability & low noise?

    Quote Originally Posted by pchaplo View Post
    Not audible noise - line noise!

    My new question is : does noise, ripple and voltage drop even within ATX outer range have any detrimental affect on components?
    I think you are getting into the realms of "how long is a piece of string".

    Breaking things down into basics, the PSU ATX standard, set's out guidelines from ripple, voltage tolerance to fault shutdown capabilities (including fire)
    So when you buy a PSU, or one is fitted by an decent OEM (in a finished PC), you know that if it meets those standards, anything PC you connect to it should operate correctly and with reliability.

    It would be a stupid standard if it allowed upto say 120mv of ripple on the 12volt line for PSU manufacturers and it was known that motherboard manufacturers decoupling capacitors failed quickly if they were introduced to anything near that level.

    Of course lower levels of ripple are favourable (that can't be denied). But when you consider motherboards and other peripherals have come a long way with the use of Solid type surface mount electrolytic capacitors, I would guess your PC would be in the recycling bin (been superceded), before you noticed any ripple accelerated faults.
    GA-P35C-DS3R Rev2.0 F11 bios, E8200 (@3.0Ghz), OCZ DDR3 PC3-10666 Reaper 4GB (@1200Mhz), Xonar D1, 8800GTS 512, Corsair HX520 (Single 12volt line, Max 40A), WDC 3200aaks/5000aaks in AHCI mode, Vista 64 Premium.

  3. #13
    Join Date
    Jan 2009
    Posts
    56

    Default Re: PSU choice for my Gigabyte board: stability & low noise?

    It sound like most people dont know. There are enough givens to answer the question. Look at the Corsair 750TX quantitative test results - is that amount of ripple at full load enough to damage anything? Yes or no. If the answer is no, then so be it. I wont waste time - but if the answer is yes, then there are alot of people saying that they have stable/clean power when they dont. Also, my first question was about RAM - look at the voltage drop for the 750TX - does that much drop affect anything like RAM or other such things? I dont know - do you really know? People tweak their RAM voltage by 0.1v to get it to match their RAM, dont they?

    ps: More like "does somewhat lower-quality rope really kill mountain climbers," given these specs for the rope. about ATX: its like car inspections - there are alot of cars on the road that pass, but I wouldnt want to be driving them for safety issues. In fact, a new SUV can roll easily - so the standards can be sloppy. Read "Unsafe at Any Speeed." Anyways back to the climbing rope. The rope is the Corsair 750TX. You have the length (the data for voltage drop and ripple as numerical values). It looks like a lot of ripple - does any engineer or pro tech here actually know?

  4. #14
    Join Date
    Jan 2009
    Posts
    56

    Default Re: PSU choice for my Gigabyte board: stability & low noise?

    here is the length of the string:
    The 3.3V rail is a little unstable, with a low result of just 3.17V -- still within the specs but lower than we like to see. The 5V rail is similar, dropping very close to the lower limit of what is allowed; at 4.79V it is very close to falling out of spec.
    The ripple is higher than we expected from such a unit, showing the first real flaw. 25mV on the 12V rail...

    source: AnandTech: Corsair TX750W Power Supply

    Are there any components that are affected by these specific & measured parameters? - thats what I am asking! Also, which components are more sensitive to these kinds of variations. I would like to hear from a tech person who has facts.

    Does anyone one have factual information? ATX standard is like like FDA drug approvals (get some viox !) or car inspections: being on the shelf or in the showroom, or hitting the road (or going to market) does not mean a vehicle is safe. Read "Unsafe at Any Speed."

    Even at full load, the 3.3v line drops about .15v - how do things that run on the 3.3v line respond to this? The 5v also has significant variance!

    For example - how does RAM work with .15v less? How much of the above drop reaches the RAM? I dont know - I am asking! What about other mobo components? -- what is on the 5v line? --what is on the 3.3v line? Tell me.

    Try it and measure your string!!!
    Last edited by pchaplo; 01-24-2009 at 12:28 PM.

  5. #15
    Join Date
    Dec 2008
    Posts
    42

    Default Re: PSU choice for my Gigabyte board: stability & low noise?

    You are forgetting the motherboard will help stabalize the RAM voltage, that's what capacitors are for. It's not the best Corsair branded unit at all, since it's a Channel Well not a Seasonic, but it's a great value. If ripple and line quality are your real concerns, you should go with a better built unit, but be prepared to pay for the quality. I would suggest Enermax, PC Power & Cooling, or Seasonic if you want the best.

    Again though, it goes back to how much you are going to stress the system, usually by overclocking and overvolting. If you run the system stock, the small varience in voltage won't matter much, but if you are pushing the parts to the envelope, it may reduce stability or cause premature failure.

    I myself don't skimp on the PSU and UPS, because I know I can migrate quality units accross several builds without issue. Plus I value quiet operating levels, and that usually costs more.

    But if you think about it, you will never load the 5v or 3.3v rails to anywhere near those levels. Most everything runs off the 12v rail now, except RAM and a few integrated chips (LAN, Sound codecs, SATA chips, etc...) and none of them get close to pulling enough to get those numbers.
    Last edited by vick1000; 01-24-2009 at 01:02 PM.

  6. #16
    Join Date
    Apr 2008
    Posts
    942

    Default Re: PSU choice for my Gigabyte board: stability & low noise?

    Quote Originally Posted by pchaplo View Post
    It sound like most people dont know. There are enough givens to answer the question. Look at the Corsair 750TX quantitative test results - is that amount of ripple at full load enough to damage anything? Yes or no. If the answer is no, then so be it. I wont waste time - but if the answer is yes, then there are alot of people saying that they have stable/clean power when they dont. Also, my first question was about RAM - look at the voltage drop for the 750TX - does that much drop affect anything like RAM or other such things? I dont know - do you really know? People tweak their RAM voltage by 0.1v to get it to match their RAM, dont they?

    ps: More like "does somewhat lower-quality rope really kill mountain climbers," given these specs for the rope. about ATX: its like car inspections - there are alot of cars on the road that pass, but I wouldnt want to be driving them for safety issues. In fact, a new SUV can roll easily - so the standards can be sloppy. Read "Unsafe at Any Speeed." Anyways back to the climbing rope. The rope is the Corsair 750TX. You have the length (the data for voltage drop and ripple as numerical values). It looks like a lot of ripple - does any engineer or pro tech here actually know?

    I think you are getting carried away with this....
    I started working in electronics 30 years ago and spent 15 years in the main as a bench engineer repairing Hi-Fi, video & tv

    I've tried to explain why the FACT that ther is an ATX standard - obviously to know avail. Equally you don't seem to be taking on board what other posters have said too.
    The other FACT is, non of us know if X make of motherboard caters better in the ripple stakes than Y make of motherboard. ALL we know is that they are to operate within a given standard.
    That means in the main, that you and anybody buying a PSU, avoids any make or model that exceeds ripple or voltage tolerance as set out by the ATX standard - that is the whole point of there being a Standard.

    look at the voltage drop for the 750TX - does that much drop affect anything like RAM or other such things? I dont know - do you really know? People tweak their RAM voltage by 0.1v to get it to match their RAM, dont they?
    NO

    Volt drop (if any) to RAM is caused by the board design vs RAM load.

    As per the Supply for the CPU, Ram and other components supplies are derived from one of the main PSU voltage lines.

    For instance the 12volt supply feeds (on modern boards) the CPU VRM. That's the Voltage Regulator (switch mode - phased) that supplies the CPU. Sometimes there can be problems if you are using a Supply Value that is close to the regulated supply you want to achieve.
    However in the CPU's case, 12 volts down to say 1.3volts Vcore regulation means that the 12volt supply can quite happily move up and down within specification (in fact it does - especially when you load up a 3dgame and theres current drawn from the graphics card).

    No one will be able to load up say use HWmonitor, or look in the PC's Bios Health Screen and see all the supply voltages "fixed" - they will always be moving +/- (within reason) every time they are polled. That though is due to variable loading through the board, rather than neccessarily the PSU varying. Although when the loading becomes that great the PSU will vary voltage in sympathy with that load.

    So back to your RAM question. Basically the RAM is like the CPU in regard to the way it's supply is derived from a higher value (than it needs) supply line.
    I'm not 100% sure where the RAM supply is derived from these days. However, if it's from either the 5 or 3.3volt, there is plenty of scope to regulate.
    For instance the ATX standard for those lines says +/- 5% are the limits.

    For example - how does RAM work with .15v less? How much of the above drop reaches the RAM? I dont know - I am asking! What about other mobo components? -- what is on the 5v line? --what is on the 3.3v line? Tell me.
    Anything that is Voltage tolerant will be Regulated, ie a voltage derived from a higher voltage supply line. That is why +/- 5% is allowed on the supply lines as per the standard. For example, anything that is voltage critical can still stay regulated and still at it's derived voltage regardless if the 5v line varies between 5.25volts or 4.75volts. Nearly all electronic designs are done this way, where a steady power supply is required. The voltage fluctuations you might see whilst in windows will be down to the loading on the board as different components draw current as and when they are invoked - Not likely because the PSU is fluctuating.

    Anyway, although I'd buy the TX750 anyday, to save you what seems like unwarranted worry, I suggest you look at alternatives.
    You might want to set yourself, up to a 3% tolerance on the main supply lines on your short list of PSU's.

    Edit...

    Does anyone one have factual information? ATX standard is like like FDA drug approvals (get some viox !) or car inspections: being on the shelf or in the showroom, or hitting the road (or going to market) does not mean a vehicle is safe. Read "Unsafe at Any Speed."
    Just going along with your Sarcasm...

    I take it you've heard of google? Well you'll find there are loads of publications on the ATX standard - which are there essentially for OEM's system builders and people like yourself if you care to read them.

    From Air Flow routes for cases to how a PSU should be self extinguishable - its there!
    Hopefully that source will be genuine enough for you and answer all your questions.
    Last edited by VorLonUK; 01-24-2009 at 09:15 PM.
    GA-P35C-DS3R Rev2.0 F11 bios, E8200 (@3.0Ghz), OCZ DDR3 PC3-10666 Reaper 4GB (@1200Mhz), Xonar D1, 8800GTS 512, Corsair HX520 (Single 12volt line, Max 40A), WDC 3200aaks/5000aaks in AHCI mode, Vista 64 Premium.

  7. #17
    Join Date
    Jan 2009
    Posts
    56

    Default Re: PSU choice for my Gigabyte board: stability & low noise?

    google? Lol right. Yes seconic manufactured units ARE superior. I've learned alot here - if you OC don't use the Corsair 750TX

    ps: all that and you don't even know what voltage rail feeds the ram? How?
    Last edited by pchaplo; 01-25-2009 at 01:02 PM.

  8. Default Re: PSU choice for my Gigabyte board: stability & low noise?

    Quote Originally Posted by pchaplo View Post
    I've learned alot here - if you OC don't use the Corsair 750TX
    That will depend on how many power consumption devices you have configured, I suppose.
    Yesterday, I was all day and by all day, I mean the 24h running CPU overcloked @ ~3400Mhz with CPU @ 1.28v, GPU slightly overcloked @ 780Mhz/950Mhz, 3 sticks of DDR3 RAM @ 1.6v, 1606Mhz. That's a reasonble OC'ing. I benchmarked and played several games (NFSU, FarCRy2, Crysis Warhead, CoDWaW, Fallout3, PES2009, GRID, GTA IV, Assassin's Creed). Also ran 3DMark Vantage and completed the default tests (scored P18074 btw). My ONLY concern was the temperatures (case open with desktop fan blowing in cool air LOL).
    I'm not saying Corsair TX750W is an extreme OC'ing PSU option nor am I selling anything but, if you have only the few necessary devices to run your PC and have a nice gaming time, that will do just fine. I'm talking by my own real time experience intensive gaming.


    JR
    System Components: Intel i7 920 C0 @ 4,0Ghz; Gigabyte EX58-UD5 (X58-EXTREME F12 BIOS); Sapphire HD5970 2GB OC Edition; Western Digital Blue 500GB; Seagate 7200.12 1000MB; 6GB Corsair 1333Mhz DDR3; DVD-RW LG; Corsair TX750W
    Watercooling System: EK-Supreme Skt 1366 WaterBlock + EK backplate Skt 1366; Kit EK-FB GAX58 WaterBlocks; Aqua-Computer AquagratiX 5970 WaterBlock; 360mm Thermaltake radiator; 360mm BI Stealth X-Flow radiator; Laing MCP355 pump; EK-DDC X-TOP rev.2 Plexi
    Danger Den Plexi E/S 1/2" reservoir; 4x A.C. Ryan UV 120mm; 3x Tacens Ventus PRO ICE 120mm; 2x NoiseBlocker Black Silent XE2 92mm; Liquid Cooling FESER Black; FESER tubing 3/8" ID and 5/8" OD; Enzotech Compression Fittings G 1/4

  9. #19
    Join Date
    Apr 2008
    Posts
    942

    Default Re: PSU choice for my Gigabyte board: stability & low noise?

    Quote Originally Posted by pchaplo View Post
    google? Lol right. Yes seconic manufactured units ARE superior. I've learned alot here - if you OC don't use the Corsair 750TX

    ps: all that and you don't even know what voltage rail feeds the ram? How?
    Because when I type on a post like this, I only reply on what I know to be fact. Because the source of where derived voltages has changed over the years, ie Athlon XP (earlier boards) Vcore was derived from the 5volt line, hence earlier PSU's had a very weak 12volt line and used to quote a "Combined Power" for the 3.3 & 5volt lines.

    In the scope of this post though it doesn't matter especially with the default ram voltages at 1.5 & 1.8volts (DDR3 & DDR2) and with a good percentage of variance (for overclocking etc), what rail the Ram supply is derived from.

    Just in case google doesn't work for you:-
    http://www.formfactors.org/developer...public_br2.pdf
    http://www.formfactors.org/developer...DG_rev_1_1.pdf

    You'll see on page 2 (link below) of this engineering document in regard to voltage stability for DDR2 & 3, that either the 3.3 or 5 volt rail are suggested.
    Also you'll note that DDR stability is important and measures to achieve that are incorporated into board design - rather than PSU design.
    http://www.wirelessnetdesignline.com/206902992
    Last edited by VorLonUK; 01-25-2009 at 08:20 PM.
    GA-P35C-DS3R Rev2.0 F11 bios, E8200 (@3.0Ghz), OCZ DDR3 PC3-10666 Reaper 4GB (@1200Mhz), Xonar D1, 8800GTS 512, Corsair HX520 (Single 12volt line, Max 40A), WDC 3200aaks/5000aaks in AHCI mode, Vista 64 Premium.

Page 2 of 2 FirstFirst 12

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •