ASRock.com Homepage
Forum Home Forum Home > Technical Support > AMD Motherboards
  New Posts New Posts RSS Feed - Fatal1ty X370 Gaming K4 can't get it stable
  FAQ FAQ  Forum Search Search  Events   Register Register  Login Login

Fatal1ty X370 Gaming K4 can't get it stable
 Post Reply Post Reply Page  12>
Author
Message Reverse Sort Order
  Topic Search Topic Search  Topic Options Topic Options
justqw View Drop Down
Newbie
Newbie


Joined: 06 May 2017
Status: Offline
Points: 26 		Post Options Post Options   Thanks (0) Thanks(0)   Quote justqw Quote  Post ReplyReply Direct Link To This Post Topic: Fatal1ty X370 Gaming K4 can't get it stable
    Posted: 09 May 2017 at 8:33am
What do You think. Does this MB handle 3200Mhz after OC ?
Back to Top
berner View Drop Down
Newbie
Newbie


Joined: 23 Apr 2017
Status: Offline
Points: 23 		Post Options Post Options   Thanks (0) Thanks(0)   Quote berner Quote  Post ReplyReply Direct Link To This Post Posted: 07 May 2017 at 1:16am

Originally posted by parsec parsec wrote:

Originally posted by berner berner wrote:

it's possible that the temps are real (my ryzen starts throttling at 85° so maybe yours is at the temp limit?) sadly asrock chose a very strange way of fan management.
It looks at the mainboard CPU temp which is something different from the  Tdie.
This temp seems to be a diode on the motherboard, for me it "laggs" behind Tdie and is around 10° C lower.
This leads to the fans not reacting properly.
Maybe you can check at what speeds your fans run. Also you may want to edit the fan curve.(In Bios or f stream)


For the CPU_FAN1 header, ASRock is using the Tdie temperature to control the fan's speed. The mother board temperature is not being used for the CPU_FAN1 header.

The Chassis Fan headers, or some of them depending upon the board and UEFI/BIOS, have the option of selecting the mother board temperature, or CPU temperature to control the fan's speed. The default setting is the mother board temperature, which normally does not change much at all. That is easily changed to monitor the CPU temperature in the HW Monitoring screen.

What do you mean by "mainboard CPU" temperature? Yes, the mother board temperature won't change at all compared to either Tdie or Tctl. If you have your CPU cooler's fan connected to a Chassis fan header, either configure it to monitor the CPU temperature, or connect the fan to the CPU_FAN1 header. Or is your CPU cooler's fan a three pin fan?

There are 2 CPU temps reported for my R7 1700.
One is under the CPU tab in HWinfo and i assume this is the correct one. I call this one Tdie. 
Then there is a temperature reported under the mainboard tab (so i assume it's reported by the mainboard). 
Those 2 Temps are very different from each other, in idle they are similar but under load cpu die is about 10°C higher than the "mainboard cpu temp".

It also takes the "mainboard cpu temp" much longer to react to load difference so i assume it's a temp sensor located on the motherboard (maybe directly under the CPU?)

If you take a look at my fan settings you can see that at 70°C the fan should spin at 80% or 1306 rpm.

It does however spin 1059RPM which is much closer to around the 60°C the "Mainboard CPU temp" reports.

This leads me to the assumption that the mainboard uses this temp to controll the fans. (Otherwise there would be a much bigger problem with fan control)




While i can compensate the offset of about 10°C in my fan settings i can not do the same with the slower reaction time of this temp sensor. 
When i tried 1.4V the CPU reached 85°C (Tdie) and begann throttling before the fans could react. 
After they reached higher RPM the Temp got lower again.


In conclusion i think it makes no sense to use this "Asrock house made" Mainboard temp to control the fans when we have a perfectly good Tdie directly from the CPU.
I think this is a serious flaw of my motherboard (k4 x370) and maybe other asrock Boards. 
If you can, please report this to Asrock since this should be a very easy problem to fix with some lines of code.



Edited by berner - 07 May 2017 at 1:23am
Back to Top
justqw View Drop Down
Newbie
Newbie


Joined: 06 May 2017
Status: Offline
Points: 26 		Post Options Post Options   Thanks (0) Thanks(0)   Quote justqw Quote  Post ReplyReply Direct Link To This Post Posted: 06 May 2017 at 10:19pm
This MB looks very good in specification but these problems with memory compatibility are the only block to click "buy now" for me and I suppose for many people either.
We all wish to run our builds in worst scenario in 2933Mhz
So.. we are just need to wait for stable,well prepared bios - hard to tell but let's be Patience : )
Back to Top
AlbinoRhino View Drop Down
Newbie
Newbie
Avatar

Joined: 28 Apr 2017
Status: Offline
Points: 153 		Post Options Post Options   Thanks (0) Thanks(0)   Quote AlbinoRhino Quote  Post ReplyReply Direct Link To This Post Posted: 06 May 2017 at 7:46pm
I don't understand why they chose a temperature offset for the X series chips either.

I mean, I get that they want your cooling to kick in with enough headroom to be effective before you reach high temps, but all of the 7 series chips are unlocked so why bother just offsetting the X chips and not all of them?
Back to Top
parsec View Drop Down
Moderator Group
Moderator Group
Avatar

Joined: 04 May 2015
Location: USA
Status: Offline
Points: 4996 		Post Options Post Options   Thanks (0) Thanks(0)   Quote parsec Quote  Post ReplyReply Direct Link To This Post Posted: 06 May 2017 at 7:11pm
Originally posted by Rares Rares wrote:

Successsss! 3.9GHz@1.3025v... but the temps are high, HWinfo reports 84 degr. C T(ttcl) in Prime95 (torture test). But I think the temp. is not real... HWinfo has 2 temps: T(tctl) and the "normal" one.


AMD has two temperatures reported from a Ryzen processor, or at least from the Ryzen 7 series.

They are Tdie and Tctl (control), both may be reported in HWiNFO depending upon the Ryzen board, given what I have seen. My X370 Killer SLI/ac shows both.

Tctl has a programmed in offset of 20° C relative to Tdie. So a Tdie of 25° C will be a Tctl of 45° C.

AMD supposedly did this for use with their stock Ryzen CPU coolers, and different Ryzen processors. How that is supposed to work I don't understand.

Originally posted by berner berner wrote:

it's possible that the temps are real (my ryzen starts throttling at 85° so maybe yours is at the temp limit?) sadly asrock chose a very strange way of fan management.
It looks at the mainboard CPU temp which is something different from the  Tdie.
This temp seems to be a diode on the motherboard, for me it "laggs" behind Tdie and is around 10° C lower.
This leads to the fans not reacting properly.
Maybe you can check at what speeds your fans run. Also you may want to edit the fan curve.(In Bios or f stream)


For the CPU_FAN1 header, ASRock is using the Tdie temperature to control the fan's speed. The mother board temperature is not being used for the CPU_FAN1 header.

The Chassis Fan headers, or some of them depending upon the board and UEFI/BIOS, have the option of selecting the mother board temperature, or CPU temperature to control the fan's speed. The default setting is the mother board temperature, which normally does not change much at all. That is easily changed to monitor the CPU temperature in the HW Monitoring screen.

What do you mean by "mainboard CPU" temperature? Yes, the mother board temperature won't change at all compared to either Tdie or Tctl. If you have your CPU cooler's fan connected to a Chassis fan header, either configure it to monitor the CPU temperature, or connect the fan to the CPU_FAN1 header. Or is your CPU cooler's fan a three pin fan?
Back to Top
berner View Drop Down
Newbie
Newbie


Joined: 23 Apr 2017
Status: Offline
Points: 23 		Post Options Post Options   Thanks (0) Thanks(0)   Quote berner Quote  Post ReplyReply Direct Link To This Post Posted: 06 May 2017 at 4:10pm
Originally posted by datonyb datonyb wrote:

no i didnt test with prime
do you think i should try maybe 7 or 8 benches at same time Big smile

No you should not use benches to stress test at all.
Run Prime or occt or even AIDA stresstest (You wanna stress CPU and FPU here, maybe even cache too.)
Back to Top
Pern_au View Drop Down
Newbie
Newbie
Avatar

Joined: 20 Mar 2017
Location: Australia
Status: Offline
Points: 16 		Post Options Post Options   Thanks (0) Thanks(0)   Quote Pern_au Quote  Post ReplyReply Direct Link To This Post Posted: 06 May 2017 at 7:06am
Originally posted by Rares Rares wrote:

Successsss! 3.9GHz@1.3025v... but the temps are high, HWinfo reports 84 degr.


Ryzen master gives accurate temp readings.
He walks among us, but he is not one of us.
Back to Top
datonyb View Drop Down
Senior Member
Senior Member


Joined: 11 Apr 2017
Location: London U.K.
Status: Offline
Points: 3139 		Post Options Post Options   Thanks (0) Thanks(0)   Quote datonyb Quote  Post ReplyReply Direct Link To This Post Posted: 06 May 2017 at 3:01am
no i didnt test with prime

i ran all at the same time
aida stress test
and then started passmark
and then started
superpi
and then started cinebench
and then just to add something alse to the mix repeat clicking to keep it running cpuz bench

so five benches/stress tests at once
i can confirm this did actually quite stress out the system with load LOL
the results were as expected very low scores on the benches Shocked and no naughty lock ups or crashes
after about ten mins (including restarting anything that finished)  i got bored felt sorry for the poor old system and decided id never be pushing that many 0's and 1's thru the cpu and ram ever again

do you think i should try maybe 7 or 8 benches at same time Big smile
[url=https://valid.x86.fr/jpg250][/url]

3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold
Back to Top
Optik View Drop Down
Newbie
Newbie
Avatar

Joined: 23 Apr 2017
Status: Offline
Points: 35 		Post Options Post Options   Thanks (0) Thanks(0)   Quote Optik Quote  Post ReplyReply Direct Link To This Post Posted: 05 May 2017 at 8:59pm
Originally posted by datonyb datonyb wrote:

Originally posted by Pern_au Pern_au wrote:

[URL=][/URL]1.3v seems a little low for 3900?


not really mine has been stable from day one at 1,306 volts and 3.9 on all cores (r7 1700)

milage may well depend on
cpu silicon lottery
power regulation and quaility of the board
and of course the power supply unit

Tested with prime? Try a test with boinc asteroids@home if u don't have a watercooler your system will crash.
And not all ryzen CPUs can run with high clockspeeds
Taichi X370
AMD R7 1700X
Arctic Liquid FreeZer 240
G.Skill Flare X 3200 cl 14@2933
PowerColer RX480
Phanteks EnTHoo Luxe
BeQuiet Dark Power Pro 11 550W
Back to Top
Rares View Drop Down
Newbie
Newbie
Avatar

Joined: 04 May 2017
Location: Brasov - RO
Status: Offline
Points: 47 		Post Options Post Options   Thanks (0) Thanks(0)   Quote Rares Quote  Post ReplyReply Direct Link To This Post Posted: 05 May 2017 at 6:03pm
Well, probably you're right about that, but my guess is that the BIOS is in a early stage and this could lead to errors/wrong hw readings/etc. 

BTW: a new AGESA v1.0.0.6 is already in the house :)


Edited by Rares - 05 May 2017 at 6:04pm
X370 Gaming K4 / Ryzen 1700 @ 3.725 / G.Skill FlareX 14-14-14-14-34 @ 3200 XMP / Sapphire Nitro RX480 / SSD x 3
Back to Top
 Post Reply Post Reply Page  12>
  Share Topic   

Forum Jump Forum Permissions View Drop Down

Forum Software by Web Wiz Forums® version 12.04
Copyright ©2001-2021 Web Wiz Ltd.

This page was generated in 0.172 seconds.