Results 1 to 23 of 23

Thread: X299 VRM Disaster

  1. #1
    Join XS BOINC Team StyM's Avatar
    Join Date
    Mar 2006
    Location
    Tropics
    Posts
    16,492

    X299 VRM Disaster


  2. #2
    Xtreme Cruncher
    Join Date
    Nov 2008
    Location
    NE Ohio, USA
    Posts
    1,601
    24/7 Cruncher #1
    Crosshair VII Hero, Ryzen 3900X, 4.0 GHz @ 1.225v, Fractal Design Celcius S36 AIO, 4x8GB GSKILL 3600MHz C15, 2x MSI 1660ti Armor OC
    Samsung 970 EVO 250GB NVMe, 2x10TB WD Red RAID1, BR Burner, Win 10 Pro, Enthoo Luxe TG, SeaSonic Focus+ Platinum 850W

    24/7 Cruncher #2
    GA-P67A-UD4-B3 BIOS F8 mod, 2600k (L051B138) @ 4.5 GHz, 1.260v full load, Arctic Liquid 120, (Boots Win7 @ 5.6 GHz per Massman binning)
    Samsung Green 2x4GB @2133 C10, 2x MSI 1660ti, Vertex 4 128GB, 2x3TB WD Red RAID1, BR Burner, Win7 Ult x64, Enthoo Pro, HX750

    24/7 Cruncher #3 ... Crucial M225 64GB SSD Donated to Endurance Testing (Died at 968 TB of writes...no that is not a typo!)
    GA-EP45T-UD3LR BIOS F10 modded, Q6600 G0 VID 1.212 (L731B536), 3.6 GHz 9x400 @ 1.312v full load, Zerotherm Zen FZ120
    OCZ 2x2GB DDR3-1600MHz C7, Gigabyte 7950 @1200/1250, Crucial MX100 128GB, 2x1TB WD Red RAID1, Win7 Ent x64, Centurion 590, XFX PRO650W

    Music System
    SB Server->SB Touch w/Android Tablet as a remote->S/PDIF->Denon AVR-X3300W->JBL Studio Series Floorstanding Speakers


  3. #3
    One-Eyed Killing Machine
    Join Date
    Sep 2006
    Location
    Inside a pot
    Posts
    6,334
    There's a good chance Der8aur kind of screwed up on this one.

    JonnyGuru pointed out that the PSU he's using has severe issues including bad quality and thin wires that suffer a whole lot of voltage drop.
    Coding 24/7... Limited forums/PMs time.

    -Justice isn't blind, Justice is ashamed.

    Many thanks to: Sue Wu, Yiwen Lin, Steven Kuo, Crystal Chen, Vivian Lien, Joe Chan, Sascha Krohn, Joe James, Dan Snyder, Amy Deng, Jack Peterson, Hank Peng, Mafalda Cogliani, Olivia Lee, Marta Piccoli, Mike Clements, Alex Ruedinger, Oliver Baltuch, Korinna Dieck, Steffen Eisentein, Francois Piednoel, Tanja Markovic, Cyril Pelupessy (R.I.P. ), Juan J. Guerrero

  4. #4
    Registered User
    Join Date
    Oct 2005
    Location
    Austria
    Posts
    68
    Well, the VRM parts were quite toasty on many X79 and X99 boards too, when properly overclocked, so nothing new here.

    However, he is right that the VRM coolers are a joke nowadays, as they are only there to look nice or fancy but fail to meet the basic criteria of a heatsink, which is a large surface area.
    We laughed at the heatpipe circus stuff 10 years ago, but they would definitely do a better job than those useless chunks of metal currently used.

    The VRM component quality doesn't seem to be particularly great either, when I look at the squeaking VRM noises in this video:
    https://youtu.be/15g_XLRc3QA
    Sounds a bit like an old matrix printer mixed with a FAX machine and this was just with a 7740X, not a 7900X.
    Power Rig: Core i7-5930K, ASRock X99 Extreme6/3.1, 16GB G.Skill DDR4-2400, Asus Strix GTX980 OC
    Time Sink: Core i7-5775C, ASRock Z97E-ITX/ac, 16GB AMD DDR3-2133, Silverstone PT-09 w/ 120W Power Brick
    HTPC: Athlon 5350, ASRock AM1H-ITX, 4GB DDR3, Supermicro SC-101i

  5. #5
    I am Xtreme zanzabar's Avatar
    Join Date
    Jul 2007
    Location
    SF bay area, CA
    Posts
    15,857
    Quote Originally Posted by BenchZowner View Post
    There's a good chance Der8aur kind of screwed up on this one.

    JonnyGuru pointed out that the PSU he's using has severe issues including bad quality and thin wires that suffer a whole lot of voltage drop.
    does that matter in this case? if anything it should help the argument here if he has poor quality cables on the psu as you cannot guarantee things will work over spec. he was running at stock and skylake X when complaining, what happens with the big parts that require rack or liquid cooling.
    5930k, R5E, samsung 8GBx4 d-die, vega 56, wd gold 8TB, wd 4TB red, 2TB raid1 wd blue 5400
    samsung 840 evo 500GB, HP EX 1TB NVME , CM690II, swiftech h220, corsair 750hxi

  6. #6
    One-Eyed Killing Machine
    Join Date
    Sep 2006
    Location
    Inside a pot
    Posts
    6,334
    It does, because with lower input voltage the VRMs draw more current.

    Anyway, apart from limiting the stable 24/7 air oc ( if ), temps aren't a problem warranty-wise, the mosfets are rated at at least 125C ( normally 150C+, with high end models coming in at up to 240C! )
    Coding 24/7... Limited forums/PMs time.

    -Justice isn't blind, Justice is ashamed.

    Many thanks to: Sue Wu, Yiwen Lin, Steven Kuo, Crystal Chen, Vivian Lien, Joe Chan, Sascha Krohn, Joe James, Dan Snyder, Amy Deng, Jack Peterson, Hank Peng, Mafalda Cogliani, Olivia Lee, Marta Piccoli, Mike Clements, Alex Ruedinger, Oliver Baltuch, Korinna Dieck, Steffen Eisentein, Francois Piednoel, Tanja Markovic, Cyril Pelupessy (R.I.P. ), Juan J. Guerrero

  7. #7
    Brilliant Idiot
    Join Date
    Jan 2005
    Location
    Hell on Earth
    Posts
    11,085
    Quote Originally Posted by BenchZowner View Post
    There's a good chance Der8aur kind of screwed up on this one.

    JonnyGuru pointed out that the PSU he's using has severe issues including bad quality and thin wires that suffer a whole lot of voltage drop.
    Ok maybe that explains part of it...except he pulled vrm sinks off and got good temps after doing so.

    That sort of eliminates psu no?
    Last edited by chew*; 06-30-2017 at 04:40 AM.
    heatware chew*
    I've got no strings to hold me down.
    To make me fret, or make me frown.
    I had strings but now I'm free.
    There are no strings on me

  8. #8
    One-Eyed Killing Machine
    Join Date
    Sep 2006
    Location
    Inside a pot
    Posts
    6,334
    Quote Originally Posted by chew* View Post
    Ok maybe that explains part of it...except he pulled vrm sinks off and got good temps after doing so.

    That sort of eliminates psu no?
    Nope, it's like graphics card overclocking and its VRM/memory chips.
    Remember the old models where the stock cooler would "split" in two or more parts ?
    Usually if you removed the "backplate" which was supposedly cooling the VRM & memory chips you'd get lower temps.

    Those terrible heatsinks ( allow me to call them... HeatStinks ) are just gathering heat and are terrible at disposing it.

    It's like taking a heatspreader to the VRM & memory and the GPU... it'll transfer heat between components and as time goes by it will raise the temps of all components.

    What would've been way more effective is something we used to do in the past if you recall for VRM & Memory ICs, small copper heatsinks, each piece for each IC.

    In any case, without the VRM Heatsink and with the same fan blowing them, with another PSU with proper voltages, it would've been even cooler.
    Coding 24/7... Limited forums/PMs time.

    -Justice isn't blind, Justice is ashamed.

    Many thanks to: Sue Wu, Yiwen Lin, Steven Kuo, Crystal Chen, Vivian Lien, Joe Chan, Sascha Krohn, Joe James, Dan Snyder, Amy Deng, Jack Peterson, Hank Peng, Mafalda Cogliani, Olivia Lee, Marta Piccoli, Mike Clements, Alex Ruedinger, Oliver Baltuch, Korinna Dieck, Steffen Eisentein, Francois Piednoel, Tanja Markovic, Cyril Pelupessy (R.I.P. ), Juan J. Guerrero

  9. #9
    Brilliant Idiot
    Join Date
    Jan 2005
    Location
    Hell on Earth
    Posts
    11,085
    I am guessing he was using seasonic?

    I have tried 3 dinosaurs on my setups. TT toughpower 1200w (cwt) pc power and cooling ( should have been retired long ago ) and Antec TPQ 1200w.

    All same results. Heatstinks suck..
    heatware chew*
    I've got no strings to hold me down.
    To make me fret, or make me frown.
    I had strings but now I'm free.
    There are no strings on me

  10. #10
    One-Eyed Killing Machine
    Join Date
    Sep 2006
    Location
    Inside a pot
    Posts
    6,334
    Superflower modular with the crystal clear connectors [ don't know why anyone would use that PSU ]
    Coding 24/7... Limited forums/PMs time.

    -Justice isn't blind, Justice is ashamed.

    Many thanks to: Sue Wu, Yiwen Lin, Steven Kuo, Crystal Chen, Vivian Lien, Joe Chan, Sascha Krohn, Joe James, Dan Snyder, Amy Deng, Jack Peterson, Hank Peng, Mafalda Cogliani, Olivia Lee, Marta Piccoli, Mike Clements, Alex Ruedinger, Oliver Baltuch, Korinna Dieck, Steffen Eisentein, Francois Piednoel, Tanja Markovic, Cyril Pelupessy (R.I.P. ), Juan J. Guerrero

  11. #11
    Brilliant Idiot
    Join Date
    Jan 2005
    Location
    Hell on Earth
    Posts
    11,085
    Same reason people would beat there head in PI 32m on a known slow board instead of just using the clearly faster board.

    They do not want to take a wizz in there sponsors wheaties.
    Last edited by chew*; 06-30-2017 at 08:25 AM.
    heatware chew*
    I've got no strings to hold me down.
    To make me fret, or make me frown.
    I had strings but now I'm free.
    There are no strings on me

  12. #12
    One-Eyed Killing Machine
    Join Date
    Sep 2006
    Location
    Inside a pot
    Posts
    6,334
    Sad but true.
    Coding 24/7... Limited forums/PMs time.

    -Justice isn't blind, Justice is ashamed.

    Many thanks to: Sue Wu, Yiwen Lin, Steven Kuo, Crystal Chen, Vivian Lien, Joe Chan, Sascha Krohn, Joe James, Dan Snyder, Amy Deng, Jack Peterson, Hank Peng, Mafalda Cogliani, Olivia Lee, Marta Piccoli, Mike Clements, Alex Ruedinger, Oliver Baltuch, Korinna Dieck, Steffen Eisentein, Francois Piednoel, Tanja Markovic, Cyril Pelupessy (R.I.P. ), Juan J. Guerrero

  13. #13
    I am Xtreme
    Join Date
    Dec 2002
    Posts
    5,951
    Quote Originally Posted by BenchZowner View Post
    It does, because with lower input voltage the VRMs draw more current.

    Anyway, apart from limiting the stable 24/7 air oc ( if ), temps aren't a problem warranty-wise, the mosfets are rated at at least 125C ( normally 150C+, with high end models coming in at up to 240C! )
    More amps at less volts is the same current as more volts as less amps. Volts * amps = watts. Never heard of a power supply affecting temps before. Certainly clean power into the VRM will get better power into CPU, better the VRM design the less that matters.

    Anyone here been around long enough to remember the plastic "air channel" shroud on the Abit IC7-G? Had lots of tiny little heatsinks in my toolbox then. HiPro's modding guide on that board got me into OC.

    Wasn't that annoying to mod the hell out of the board when it wasn't pushing 400 bux.

  14. #14
    One-Eyed Killing Machine
    Join Date
    Sep 2006
    Location
    Inside a pot
    Posts
    6,334
    Quote Originally Posted by Revv23 View Post
    More amps at less volts is the same current as more volts as less amps. Volts * amps = watts. Never heard of a power supply affecting temps before. Certainly clean power into the VRM will get better power into CPU, better the VRM design the less that matters.

    Anyone here been around long enough to remember the plastic "air channel" shroud on the Abit IC7-G? Had lots of tiny little heatsinks in my toolbox then. HiPro's modding guide on that board got me into OC.

    Wasn't that annoying to mod the hell out of the board when it wasn't pushing 400 bux.
    You are missing the point.

    1) Lower driving voltage = mosfets operate at sub-optimal efficiency ( hint: worse efficiency = not as clean output, and more power gone to waste = heat )
    2) A mosfet gets hotter when it provides more Ampere compared to the same output but at a higher voltage with lower Ampere
    Coding 24/7... Limited forums/PMs time.

    -Justice isn't blind, Justice is ashamed.

    Many thanks to: Sue Wu, Yiwen Lin, Steven Kuo, Crystal Chen, Vivian Lien, Joe Chan, Sascha Krohn, Joe James, Dan Snyder, Amy Deng, Jack Peterson, Hank Peng, Mafalda Cogliani, Olivia Lee, Marta Piccoli, Mike Clements, Alex Ruedinger, Oliver Baltuch, Korinna Dieck, Steffen Eisentein, Francois Piednoel, Tanja Markovic, Cyril Pelupessy (R.I.P. ), Juan J. Guerrero

  15. #15
    I am Xtreme zanzabar's Avatar
    Join Date
    Jul 2007
    Location
    SF bay area, CA
    Posts
    15,857
    Quote Originally Posted by BenchZowner View Post
    You are missing the point.

    1) Lower driving voltage = mosfets operate at sub-optimal efficiency ( hint: worse efficiency = not as clean output, and more power gone to waste = heat )
    2) A mosfet gets hotter when it provides more Ampere compared to the same output but at a higher voltage with lower Ampere
    isnt that why they should have an 8+4 or 8+8 connector on the board. especially ones that are intended to push chips that are over 160W stock with turbo.
    5930k, R5E, samsung 8GBx4 d-die, vega 56, wd gold 8TB, wd 4TB red, 2TB raid1 wd blue 5400
    samsung 840 evo 500GB, HP EX 1TB NVME , CM690II, swiftech h220, corsair 750hxi

  16. #16
    One-Eyed Killing Machine
    Join Date
    Sep 2006
    Location
    Inside a pot
    Posts
    6,334
    Nope, it's ot required, unless the PSU manufacturer gets cheap and puts low quality connectors on their PSUs like SuperFlower did.
    That voltage drop is ridiculous.

    Haven't seen any decent and pricey PSU suffer such a terrible drop even during hot room tests
    Coding 24/7... Limited forums/PMs time.

    -Justice isn't blind, Justice is ashamed.

    Many thanks to: Sue Wu, Yiwen Lin, Steven Kuo, Crystal Chen, Vivian Lien, Joe Chan, Sascha Krohn, Joe James, Dan Snyder, Amy Deng, Jack Peterson, Hank Peng, Mafalda Cogliani, Olivia Lee, Marta Piccoli, Mike Clements, Alex Ruedinger, Oliver Baltuch, Korinna Dieck, Steffen Eisentein, Francois Piednoel, Tanja Markovic, Cyril Pelupessy (R.I.P. ), Juan J. Guerrero

  17. #17
    I am Xtreme
    Join Date
    Dec 2002
    Posts
    5,951
    Quote Originally Posted by BenchZowner View Post
    You are missing the point.

    1) Lower driving voltage = mosfets operate at sub-optimal efficiency ( hint: worse efficiency = not as clean output, and more power gone to waste = heat )
    2) A mosfet gets hotter when it provides more Ampere compared to the same output but at a higher voltage with lower Ampere
    Hey BZ, can you tell me where you got this info? This is the first time I've ever heard anything like this. By this logic I should go to Europe and run on 220 to get a cooler running PSU.

    Edit - just realized I'm being somewhat combative, not meaning to have much respect for you, I just have never heard this. Certainly dirty power (droop, ripple,etc) can cause inneficiency... But major increase in temp (to the point of VRM throttling) because of 12v rail being low? (But still within spec)
    Last edited by Revv23; 07-13-2017 at 05:56 AM.

  18. #18
    One-Eyed Killing Machine
    Join Date
    Sep 2006
    Location
    Inside a pot
    Posts
    6,334
    Even the first sentence you wrote is correct.

    220V will net your PSU BETTER EFFICIENCY ( which also includes less heat output ).
    If you want to look further into such things you should start reading something like an electrical engineering 101 book.
    Coding 24/7... Limited forums/PMs time.

    -Justice isn't blind, Justice is ashamed.

    Many thanks to: Sue Wu, Yiwen Lin, Steven Kuo, Crystal Chen, Vivian Lien, Joe Chan, Sascha Krohn, Joe James, Dan Snyder, Amy Deng, Jack Peterson, Hank Peng, Mafalda Cogliani, Olivia Lee, Marta Piccoli, Mike Clements, Alex Ruedinger, Oliver Baltuch, Korinna Dieck, Steffen Eisentein, Francois Piednoel, Tanja Markovic, Cyril Pelupessy (R.I.P. ), Juan J. Guerrero

  19. #19
    I am Xtreme zanzabar's Avatar
    Join Date
    Jul 2007
    Location
    SF bay area, CA
    Posts
    15,857
    Quote Originally Posted by Revv23 View Post
    Hey BZ, can you tell me where you got this info? This is the first time I've ever heard anything like this. By this logic I should go to Europe and run on 220 to get a cooler running PSU.

    Edit - just realized I'm being somewhat combative, not meaning to have much respect for you, I just have never heard this. Certainly dirty power (droop, ripple,etc) can cause inneficiency... But major increase in temp (to the point of VRM throttling) because of 12v rail being low? (But still within spec)
    more input amps means more heat, if you look at the efficiency curves on a PSU they always have one for for 110-120 and one for 220-230 like this
    *
    5930k, R5E, samsung 8GBx4 d-die, vega 56, wd gold 8TB, wd 4TB red, 2TB raid1 wd blue 5400
    samsung 840 evo 500GB, HP EX 1TB NVME , CM690II, swiftech h220, corsair 750hxi

  20. #20
    I am Xtreme
    Join Date
    Dec 2002
    Posts
    5,951
    I feel like we are talking about different things here. I am asking how an in spec PSU is causing the VRM to throttle. Maybe I will pick up the EE book so I can find the section about tolorances

  21. #21
    One-Eyed Killing Machine
    Join Date
    Sep 2006
    Location
    Inside a pot
    Posts
    6,334
    Actually they aren't throttling.

    Throttling would be dropping the output and the logic circuit downclocking the CPU.

    They are simply "overheating", in quotes because they aren't really overheating, FETs are rated at 150C++ anyway and and under 125C which is the crappiest spec of FETs I've ever seen on a mobo from a major brand ( Asus, Gigabyte, ASRock, MSI, eVGA, etc ) is far from "problematic".

    You may not want more heat in your case, but having mosfets running at 100C for example is far from creating any sort of issues.

    Just like every single chip/circuit out there, FETs have their own optimal input and operating specifications, going outside those thresholds the efficiency goes down and as efficiency goes down heat output gets higher.
    I can't seem to find a way to explain it in easier and clearer layman's terms.
    Coding 24/7... Limited forums/PMs time.

    -Justice isn't blind, Justice is ashamed.

    Many thanks to: Sue Wu, Yiwen Lin, Steven Kuo, Crystal Chen, Vivian Lien, Joe Chan, Sascha Krohn, Joe James, Dan Snyder, Amy Deng, Jack Peterson, Hank Peng, Mafalda Cogliani, Olivia Lee, Marta Piccoli, Mike Clements, Alex Ruedinger, Oliver Baltuch, Korinna Dieck, Steffen Eisentein, Francois Piednoel, Tanja Markovic, Cyril Pelupessy (R.I.P. ), Juan J. Guerrero

  22. #22
    I am Xtreme zanzabar's Avatar
    Join Date
    Jul 2007
    Location
    SF bay area, CA
    Posts
    15,857
    Quote Originally Posted by BenchZowner View Post
    Actually they aren't throttling.

    Throttling would be dropping the output and the logic circuit downclocking the CPU.

    They are simply "overheating", in quotes because they aren't really overheating, FETs are rated at 150C++ anyway and and under 125C which is the crappiest spec of FETs I've ever seen on a mobo from a major brand ( Asus, Gigabyte, ASRock, MSI, eVGA, etc ) is far from "problematic".

    You may not want more heat in your case, but having mosfets running at 100C for example is far from creating any sort of issues.

    Just like every single chip/circuit out there, FETs have their own optimal input and operating specifications, going outside those thresholds the efficiency goes down and as efficiency goes down heat output gets higher.
    I can't seem to find a way to explain it in easier and clearer layman's terms.
    they were throttling since the boards are exceeding their temp rating. the temps are not from the parts them selves but from the pcb close to it. if you read 120c+ from the board it is likely 20-50c over that on the part with those. that is why boards will throttle a bit when the vrm goes over 100-120 on the probe.

    we also are going to have parts that can draw 40-60% more power so this could be a huge platform problem.
    5930k, R5E, samsung 8GBx4 d-die, vega 56, wd gold 8TB, wd 4TB red, 2TB raid1 wd blue 5400
    samsung 840 evo 500GB, HP EX 1TB NVME , CM690II, swiftech h220, corsair 750hxi

  23. #23
    Registered User
    Join Date
    Nov 2010
    Location
    derrrr
    Posts
    15
    Quote Originally Posted by BenchZowner View Post
    Even the first sentence you wrote is correct.

    220V will net your PSU BETTER EFFICIENCY ( which also includes less heat output ).
    If you want to look further into such things you should start reading something like an electrical engineering 101 book.
    Lol i logged in the first time in like 5 years to say LOLOLOLOLOLOLOL

    You said it well... EE 101

    Rds(on) of mosfet generate heat

    If your CPU need 300 watt

    And you give it 11.4v instead of 12.2v(Stepping down to vcore)

    It will draw more amps at 11.4v to reach 300 watt

    Heat is Amps * Amps * resistance

    Yes, AMPS SQUARED

    Every millivolt matter

    Regarts

Bookmarks

Bookmarks

Posting Permissions

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