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

Z370 / Z390 VRM Discussion Thread - Page 4

post #31 of 1407
Anyone know if the VRM on the Asus Strix Z370 ITX is upgraded from the Z270 version? Considering the higher power draw of Z370, I certainly hope it is but need to know specifics. If anyone can share its VRM details, thanks.
post #32 of 1407
Quote:
Originally Posted by AlphaC View Post

IMHO the Taichi is the same price and comes with wifi + audio AMP and supposedly better VRM components (if it's the Fairchild FDPC5030SG ones as on Fatal1ty pro)
Back of the board has doublers.

Also this confirms the Killer SLI is bad. It doesn't even have USB 3.1 gen2 , dual BIOS, or ALC1220, so for the $10 price difference to the Extreme4 it is really bad.

The VRM is at least $20 upgrade , audio is minimal (maybe $10) , USB 3.1 probably $15-20 based on add-in card pricing.

The Taichi is using Fairchild FDPC5030SG parts while the Extreme4 looks to use Sinopower SM7341EH.

FAIRCHILD http://www.onsemi.com/PowerSolutions/product.do?id=FDPC5030SG
Channel 1
RDS(on) = 5.0mΩ @ V_GS = 10V , I_D = 17A
RDS(on) = 6.5mΩ @ V_GS = 4.5V , I_D = 14A
2ns rise time (10ns max), 2ns fall time (10ns max)
Channel 2
2.4mΩ @ V_GS = 10V , I_D = 25A
3.0mΩ @ V_GS = 4.5V , I_D = 22A
4ns rise time (10ns max) , 3ns fall time (10ns max)

SINOPOWER SM7341EH www.sinopowersemi.com/temp/SM7341EHKP_datasheet.pdf
Channel 1
30V/24A,
RDS(ON) = 3.9mΩ (max.) @ V_GS = 10V
RDS(ON) = 6.5mΩ (max.) @ V_GS = 4.5V
9.6ns rise time, 19ns fall time
Channel 2
30V/44A,
RDS(ON) = 1.2mΩ (max.) @ V_GS =10V <--- the marketing number comes from here
RDS(ON) = 2mΩ (max.) @ V_GS =4.5V
14ns rise time, 25ns fall time

asdkj1740 , the ubiq parts might be what are used on MSI GPUs.
QN3101 high side , QN3107 low side seems to be what is on Zotac's GTX 1080 Ti amp
asrock has been cheaping out on audio part in recent generation..
post #33 of 1407
Quote:
Originally Posted by asdkj1740 View Post

exactly, i have no idea why on the back of the taichi pcb there are 12 little ic which seem to be all drivers...

https://www.intersil.com/content/dam/intersil/documents/isl6/isl6596.pdf
should be this model..
it is just a driver, not a doubler, i guess...
although it states that it can driver two dual n channel mosfet, there are two pin (UGATE (Pin 1) & LGATE (Pin 5)) for high side and low side respectively.

the question then is, where the f are those doublers! there is no way for that pwm controller to drive true 8+4 phases.

datasheet says "Drives Two N-Channel MOSFETs"


See

https://www.allaboutcircuits.com/textbook/reference/chpt-9/transistors-insulated-gate-field-effect-igfet-or-mosfet/
Edited by AlphaC - 10/6/17 at 1:39pm
post #34 of 1407
Quote:
Originally Posted by br0da View Post

@αC: Impressive post, awesome!
I'd guess it's a 8+4+1+1 since I can't find any other converters for VCCIO and VCCSA but that wouldn't fit the doubler sheme. headscratch.gif

Edit: I'm going to update this post and the hardwareluxx list as soon as I can but atm I'm quite busy and there's a huge amount of info that needs to be added.

Edit2:
Due to mosfet count per phase I'd expect this to run in 4+1 mode with one doubler used for VCCGT (can be seen in the picture near the inductor).
I'd agree with asdkj1740's interpretation of the cheap MSI design.

Edit3: I'm drowning in Reviews!
http://www.coolpc.com.tw/phpBB2/viewtopic.php?f=70&t=243608&sid=5953aa46427649d74420df2cd02cb957
-> Intersil drivers on the backside of the Taichi:

I still don't get how this design works. headscratch.gif

http://www.coolpc.com.tw/phpBB2/viewtopic.php?f=70&t=243595&sid=5953aa46427649d74420df2cd02cb957
http://www.coolpc.com.tw/phpBB2/viewtopic.php?f=70&t=243584&sid=5953aa46427649d74420df2cd02cb957
https://www.chiphell.com/article-18765-1.html
https://www.overclockers.ua/motherboard/asus-rog-strix-z370-e-gaming/
So Asus Z370 Strix E/F have worse VRM than X370-F? I mean, i have 10 IR3555s and this uses classic HS+LS config with the SiRa fets.
post #35 of 1407
Here's the spec sheet for the mosfets on the Gigabyte Aorus Gaming 7: https://www.intersil.com/content/dam/Intersil/documents/isl9/isl99227-27b.pdf

Efficiency looks impressive but the rise/fall times seem a bit slow? I'm not sure which specs are important.
post #36 of 1407
Quote:
Originally Posted by HKPolice View Post

Here's the spec sheet for the mosfets on the Gigabyte Aorus Gaming 7: https://www.intersil.com/content/dam/Intersil/documents/isl9/isl99227-27b.pdf

Efficiency looks impressive but the rise/fall times seem a bit slow? I'm not sure which specs are important.
The times are related to efficiency, so you just look on the graph and you're gtg.
Tr/Tf are used in HS switching heat loss calculation which goes like this: Psw= Vin * 0.5 * Iout * (Tr+Tf)s * Fsw
The times dont really matter on LowSide
post #37 of 1407
Thread Starter 
Quote:
Originally Posted by AlphaC View Post

The Taichi is using Fairchild FDPC5030SG parts while the Extreme4 looks to use Sinopower SM7341EH.

What is the guess about the Extreme4 FETs based on? smile.gif
Quote:
Originally Posted by asdkj1740 View Post

it is just a driver, not a doubler, i guess...
[...]
the question then is, where the f are those doublers! there is no way for that pwm controller to drive true 8+4 phases.

Yes these are just 12 drivers.
My guess is that the controller runs in 4+2 mode, ASRock is using each PWM signal for two phases.
The last two phases are for VCCIO and VCCSA.

Having @Nickshih here for clearing things up would be great.
Quote:
Originally Posted by VRMfreak View Post

So Asus Z370 Strix E/F have worse VRM than X370-F?

Yes you are right about that.
Quote:
Originally Posted by VRMfreak View Post

The times dont really matter on LowSide

For lowside it matters as much as for highside.
Edited by br0da - 10/7/17 at 6:43am
post #38 of 1407
Quote:
Originally Posted by br0da View Post

For lowside it matters as much as for highside.
But when lowside has most of its losses as resistive (conductive), then the Tr/Tf have less impact on the overall efficiency compared to highside (???????)
Edited by VRMfreak - 10/7/17 at 6:42am
post #39 of 1407
Thread Starter 
Okay of course the impact of switching losses on the overall efficiency is lower if the conductive losses are ways too high. biggrin.gif
But still each MOSFET in a buck converter is getting switched two times each period no matter if it's used as a control or sync FET.
post #40 of 1407
Quote:
Originally Posted by VRMfreak View Post

So Asus Z370 Strix E/F have worse VRM than X370-F? I mean, i have 10 IR3555s and this uses classic HS+LS config with the SiRa fets.

Aww I was really hoping to maybe get one of the z370e or f

How bad are the vrms for over clocking on it. Is it just they will use more power?
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel Motherboards
Overclock.net › Forums › Intel › Intel Motherboards › Z370 / Z390 VRM Discussion Thread