Overclock.net › Forums › Intel › Intel Motherboards › Z370 / Z390 VRM Discussion Thread
New Posts  All Forums:Forum Nav:

Z370 / Z390 VRM Discussion Thread - Page 28

post #271 of 1132
Im weird and like RGB, asrock not including an addressable header is a real factor for me lol. Its hard to believe gaming 5 has lower quality power delivery than extreme 4, it retails for 40 dollars more.
Steve
(9 items)
 
HTPC
(9 items)
 
 
CPUMotherboardGraphicsRAM
AMD Ryzen 7 1700 3.9@1.376v Asrock Killer x370 SLI/AC EVGA GTX 1060 SSC 6gb Gskill ripjaws 3200/2933 14-16-16-36 
Hard DriveCoolingMonitorPower
Seagate firecuda 1tb Be Quiet! Dark Rock Pro 3 Dell S2417DG EVGA 650 GQ 
Case
NZXT H440 Matte Black 
CPUMotherboardGraphicsRAM
Intel Pentium G4560 Asrock H110M-ITX PNY GT 1030 4GB Team Dark 2400  
Hard DriveMonitorKeyboardCase
PNY CS900 240gb SSD Vizio Smartcast E55 Logitech K400 Inwin BP655.300 
Audio
Vizio 3.1 Soundbar 
  hide details  
Reply
Steve
(9 items)
 
HTPC
(9 items)
 
 
CPUMotherboardGraphicsRAM
AMD Ryzen 7 1700 3.9@1.376v Asrock Killer x370 SLI/AC EVGA GTX 1060 SSC 6gb Gskill ripjaws 3200/2933 14-16-16-36 
Hard DriveCoolingMonitorPower
Seagate firecuda 1tb Be Quiet! Dark Rock Pro 3 Dell S2417DG EVGA 650 GQ 
Case
NZXT H440 Matte Black 
CPUMotherboardGraphicsRAM
Intel Pentium G4560 Asrock H110M-ITX PNY GT 1030 4GB Team Dark 2400  
Hard DriveMonitorKeyboardCase
PNY CS900 240gb SSD Vizio Smartcast E55 Logitech K400 Inwin BP655.300 
Audio
Vizio 3.1 Soundbar 
  hide details  
Reply
post #272 of 1132
Quote:
Originally Posted by Techhog View Post

So I can't go above 150W with the Strix E, right?

How can one extrapolate the maximum wattage provided by board 'x' at a certain vrm temperature from the initial table ? or am i missing something ?
post #273 of 1132
Any new info about VRMs used in gigabyte gaming 5 ?
post #274 of 1132
Quote:
Originally Posted by awesomegamer919 View Post

You certainly could, make sure the VRM heatsink has some airflow though.

Quote:
Originally Posted by AlphaC View Post

Depending on mosfet cooling I'd say it's more like 220W... figure <20A per phase (see the limit of SiRA14dp at 25°C ambient) with ~ 1.3-1.4V, ~88% efficiency.

To expand on http://www.overclock.net/t/1638955/z370-z390-vrm-discussion-thread/70#post_26383261

At 150A , 300kHz f_sw, 1.4V I would expect the lowside Sira12dp to reach about 90°C and that is without accounting for the ~ +20% RDs(on) adjustment for temperatures around 75°C. Wouldn't recommend no airflow.

If we use 100°C as the target max VRM temp and assuming 1 sq in of 2oz copper PCB that is. In reality it is less than 1 sq in.

No airflow? You're looking max 100A and 1.3V (about 130W) using the steady state thermal resistance of 70°C/W before things hit 100°C.

Thanks guys. My VRM temp was hitting 80C at 150W (if I was reading the correct temp), so I guess I'm safe now since I'll never even be hitting that much again unless I delid. Though, that also means that I'll still have headroom if I delid! Hm, decisions...
Quote:
Originally Posted by catacavaco View Post

How can one extrapolate the maximum wattage provided by board 'x' at a certain vrm temperature from the initial table ? or am i missing something ?

You can't unless you understand it. I was just asking since I thought I saw someone here say that. Turns out it's a much higher ceiling; higher than anyone wish standard cooling would ever need in fact.
post #275 of 1132
Quote:
Originally Posted by winter2 View Post

Any new info about VRMs used in gigabyte gaming 5 ?

Latest info I have without having a board in hand: 4+4, Intersil ISL95866, ISL6625Ax2, 1H1L
post #276 of 1132
Can anyone confirm gigabyte is having high VRM temps? Or was that 124c a one off result?

Also how much do heat pipes lower temps, id like to see a comparison between extreme 4 and k6 (they look like same heatsinks, k6 having a heatpipe).

Im sooooo annoyed asrock boards are not sold by amazon yet, i have 20% off coupon that ends on the 23rd lol.

Edit: Also does anyone else find the m.2 heatsink on the strix boards ugly? Looks out of place somehow.
Edited by Scotty99 - 10/19/17 at 2:58pm
Steve
(9 items)
 
HTPC
(9 items)
 
 
CPUMotherboardGraphicsRAM
AMD Ryzen 7 1700 3.9@1.376v Asrock Killer x370 SLI/AC EVGA GTX 1060 SSC 6gb Gskill ripjaws 3200/2933 14-16-16-36 
Hard DriveCoolingMonitorPower
Seagate firecuda 1tb Be Quiet! Dark Rock Pro 3 Dell S2417DG EVGA 650 GQ 
Case
NZXT H440 Matte Black 
CPUMotherboardGraphicsRAM
Intel Pentium G4560 Asrock H110M-ITX PNY GT 1030 4GB Team Dark 2400  
Hard DriveMonitorKeyboardCase
PNY CS900 240gb SSD Vizio Smartcast E55 Logitech K400 Inwin BP655.300 
Audio
Vizio 3.1 Soundbar 
  hide details  
Reply
Steve
(9 items)
 
HTPC
(9 items)
 
 
CPUMotherboardGraphicsRAM
AMD Ryzen 7 1700 3.9@1.376v Asrock Killer x370 SLI/AC EVGA GTX 1060 SSC 6gb Gskill ripjaws 3200/2933 14-16-16-36 
Hard DriveCoolingMonitorPower
Seagate firecuda 1tb Be Quiet! Dark Rock Pro 3 Dell S2417DG EVGA 650 GQ 
Case
NZXT H440 Matte Black 
CPUMotherboardGraphicsRAM
Intel Pentium G4560 Asrock H110M-ITX PNY GT 1030 4GB Team Dark 2400  
Hard DriveMonitorKeyboardCase
PNY CS900 240gb SSD Vizio Smartcast E55 Logitech K400 Inwin BP655.300 
Audio
Vizio 3.1 Soundbar 
  hide details  
Reply
post #277 of 1132
@ Techhog:

The weird part is that the Asus Z370-A prime looks to have a better heatsink in terms of fin area than the STRIX ones.

Gigabyte had been using SiRA12dp for a while now on Intel mainstream boards , with OnSemi 4C10N + 4C06N recently. That is why I am skeptical that STRIX boards are worthwhile for VRM alone. The package design and datasheet specs are not a massive improvement over the Z370-A.

At least one review has 200W delta power consumption for the i7-8700k CPU when pushing the limits. But as mentioned earlier , so long as Asus is not using 4 phases with two MOSFETs per phase like the Gigabyte boards with four CPU phases I would think you'd be fine when applying airflow.

edit: also the Z370-A prime has a power on button , so it's better for a bench test setup before installing it in a case

@ Scotty:

that heatpipe only transfers heat between the two heatsinks. If you look at the heatsink it doesn't have the heatsink contacting the base. As such all it does is more or less double the effective fin area since you have two heatsinks (of similar size) instead of one.

The Fatal1ty k6 is worth the extra $10 USD on Newegg due to the Debug LED IMO. The power/reset are not as essential.

See also www.ti.com/lit/an/slva462/slva462.pdf
www.altera.com/en_US/pdfs/literature/an/an185.pdf
https://www.sparkfun.com/tutorials/314





If you look at the above, you reduce R_sA (the thermal resistance of heatsink to air since you get more fin area)

For people that missed the post in AM4 thread re: power dissipation
Quote:
Originally Posted by catacavaco View Post

How can one extrapolate the maximum wattage provided by board 'x' at a certain vrm temperature from the initial table ? or am i missing something ?

see
https://www.fairchildsemi.com/application-notes/AN/AN-6005.pdf
http://www.ti.com/lit/an/slyt664/slyt664.pdf
http://ww1.microchip.com/downloads/en/AppNotes/01471A.pdf
http://rohmfs.rohm.com/en/products/databook/applinote/ic/power/switching_regulator/power_loss_appli-e.pdf
http://www.electronicdesign.com/power/fundamentals-buck-converter-efficiency

I typically use worst case V_GS = 4.5V if possible. V_GS = 10V results in fewer losses.

Most mosfet datasheets are based off 1 sq.in 2 oz copper PCB

Thermal resistance junction to ambient = no heatsink
For a heatsinked mosfet , R_thetaJC Thermal resistance junction to case (not including heatsink) + R_CS Thermal resistance case to heatsink (usually the thermal tape) + R_SA Thermal resistance of heatsink to ambient

----
Quote:
Originally Posted by GBT-MatthewH View Post

Latest info I have without having a board in hand: 4+4, Intersil ISL95866, ISL6625Ax2, 1H1L

In plain English I think that means
4 CPU phases doubled to 8 + 2 uncore/SOC/whatever other phases doubled to 4

PWM = ISL95866 https://www.intersil.com/en/products/power-management/computing-power-vrm-imvp/multiphase-controllers/ISL95866.html
Quote:
Dual output controller
Voltage regulator A: 4-, 3-, 2-, or 1-phase designs with two +12V integrated gate drivers
Voltage regulator B: 3-, 2-, or 1-phase designs with one +12V integrated gate driver

TWO Dual output driver ISL6625A per PWM phase? https://www.intersil.com/en/products/power-management/mosfet-drivers/synchronous-drivers-for-multiphase-pwm/ISL6625A.html

presumably "1H1L" = 1 highside mosfet + lowside mosfet of unknown quality

The VRM looks identical to the Z270X Gaming 5 now that I look at it

Isn't it more likely to be 4 doubled to 8 and another 3 phases for the iGPU / etc?

On the bright side Steven put the Gigabyte z370 Gaming 7 fears to rest , he obtained 47°C VRM temps while using 1.3V and 4.9GHz (albeit with 120mm fan over VRM) in Handbrake... which is pretty much what we would expect
https://www.tweaktown.com/reviews/8377/gigabyte-z370-aorus-gaming-7-motherboard-review/index10.html
Edited by AlphaC - 10/19/17 at 5:25pm
Workstation stuff
(407 photos)
SpecViewperf 12.0.1
(179 photos)
 
Reply
Workstation stuff
(407 photos)
SpecViewperf 12.0.1
(179 photos)
 
Reply
post #278 of 1132
Quote:
Originally Posted by AlphaC View Post

@ Techhog:

The weird part is that the Asus Z370-A prime looks to have a better heatsink in terms of fin area than the STRIX ones.

Gigabyte had been using SiRA12dp for a while now on Intel mainstream boards , with OnSemi 4C10N + 4C06N recently. That is why I am skeptical that STRIX boards are worthwhile for VRM alone. The package design and datasheet specs are not a massive improvement over the Z370-A.

At least one review has 200W delta power consumption for the i7-8700k CPU when pushing the limits. But as mentioned earlier , so long as Asus is not using 4 phases with two MOSFETs per phase like the Gigabyte boards with four CPU phases I would think you'd be fine when applying airflow.

Have you seen a side view of the heatsink? It might balance out. Either way I'm strongly considering buying the Maximus X Code when that releases and selling this board to my friend just so he can feel secure that he doesn't end up with a board that's not well supported by its manufacturer, but he's not gonna want to spend $200+ on just a board.
post #279 of 1132
Quote:
Originally Posted by AlphaC View Post

It's very good. Better than the TI NexFET Powerblocks actually.

The ISL99227B has more thermal dissipation due to the exposure of the junction to the heatsink , plus they won electronic innovation awards in 2016. They're actually a bit better than International Rectifier IR3555. They have up to 2MHz switching frequency vs 1.5MHz on NexFETs and 1MHz on IR powerstages.

Generally speaking : Powerstage (integrated driver+high side+low side) such as ISL99227B / IR3555 > Powerblock (high+low side) such as the TI NexFET 87350 , Fairchild dual-N Powertrench, or the Sinopower > separate Powerpak / 2 Low RDS(on) mosfets for high side such as Onsemi 4C09N + 4C06N or Vishay Sira19+Sira12


Texas Instruments diagram


http://www.ti.com/lsds/ti/power-management/power-mosfet-module-overview.page

Right now it's looking like for VRM (subject to change if some boards are using 2 low side mosfets aka "doubled low side"):
TOP = Gigabyte Z370 SOC Force (unreleased, found on hwbot records) , Gigabyte Z370 Gaming 7 due to iSL99227B (60A rating) ... top audio,top overkill VRM , ... likely Asus Z370 Maximus X Extreme too
Upper tier (LN2, custom water) = Asrock Z370 Fatal1ty Gaming Pro i7 (whatever the full name is, the expensive one) , Asrock Z370 Taichi , using Fairchild Dual-N (low+high side) when heatsinked properly with heatpipe (limited to about 35A per mosfet) , Asus ROG APEX (a record holder on LN2) /Formula/Hero boards using TI NexFETs (per marketing page) , MSI Z370 Godlike (until more info it goes here, but it's poor price/perf regardless)
Midrange <$200 (All-in-one watercooling/dual tower air cooling)= Asrock Z370 Extreme4 / K6 using Sinopower 2-in-1 / "DSM"/ low+high side
Low Midrange , should be $160ish or less (Asus z270-A level VRM)= Asus Z370E/F /G , Asus Z370-A , Gigabyte Z370 Gaming 5 (here or higher), MSI Z370 Pro carbon , MSI z370 M5
Entry (sub 150W probably / 120mm fan cooler level, should be $120 or less if VRM is the only concern) = Asrock Z370 Killer SLI (cut down features is another issue on top of regular Sinopowers) , MSI z370 SLI PLUS (uPi controller + Ubiq mosfets), MSI Z370 Tomahawk (uPi controller + Ubiq mosfets ; ALC892) , MSI Z370 Krait (uPi controller + Ubiq mosfets)
Not suitable for really overclocking unless it's not an i7 (at best MCE/ multi-core turbo only) = Asus TUF Pro Gaming / z370 STRIX H, Gigabyte Z370 Ultra Gaming , Gigabyte Z370 Gaming 3 & K3, Gigabyte Z370 XP SLI, MSI Z370 Gaming Plus, etc
<4 phases , or partially Unheatsinked (don't buy) :Asus Z370-P / TUF Z370 Plus, MSI Z370-A Pro , MSI z370 PC Pro, Asrock Z370 Pro4 , Gigabyte Z370 HD3, Gigabyte Z370M D3H , etc

Oct 18,2017 Made more legible:

Is this top still valid? best motherboards to buy (regardless of the price) should still be Gigabyte Z370 Gaming 7 & Asus Z370 Maximus X Extreme ?
post #280 of 1132
Quote:
Originally Posted by Scotty99 View Post

Can anyone confirm gigabyte is having high VRM temps? Or was that 124c a one off result?

Also how much do heat pipes lower temps, id like to see a comparison between extreme 4 and k6 (they look like same heatsinks, k6 having a heatpipe).

Im sooooo annoyed asrock boards are not sold by amazon yet, i have 20% off coupon that ends on the 23rd lol.

Edit: Also does anyone else find the m.2 heatsink on the strix boards ugly? Looks out of place somehow.

Gigabyte gaming 5 - 8700k 4.9Ghz - 1.3V - Stress test = temperature around 60s reported by motherboard VRM sensor
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel Motherboards
Overclock.net › Forums › Intel › Intel Motherboards › Z370 / Z390 VRM Discussion Thread