AMD Motherboards - [SOLVED] ASRock X370 Taichi cold boot error 54

[SOLVED] ASRock X370 Taichi cold boot error 54

Printed From: ASRock.com
Category: Technical Support
Forum Name: AMD Motherboards
Forum Description: Question about ASRock AMD motherboards
URL: https://forum.asrock.com/forum_posts.asp?TID=6450
Printed Date: 20 Apr 2024 at 10:55am
Software Version: Web Wiz Forums 12.04 - http://www.webwizforums.com

Topic: [SOLVED] ASRock X370 Taichi cold boot error 54

Posted By: Misant
Subject: [SOLVED] ASRock X370 Taichi cold boot error 54
Date Posted: 30 Oct 2017 at 2:50am

Hello,

I'm here to ask some help regarding a weird issue on my new setup:

1) Computer part list :

- Motherboard : ASRock X370 Taichi

- RAM : G.Skill FLare-X F4-3200C14D-16GFX (2 x 8GB)

- CPU : AMD Ryzen 1700X

- CPU Cooler : Cooler Master MasterLiquid 240 (push/pull config)

- PSU : Corsair CX650 - 650W

- VGA : ASUS GeForce GTX 1070 - TURBO

- HDD1 : Samsung 840 Pro SSD 120 GB

- HDD2 : Samsung 850 EVO 250 GB

- Case : Antec P8

- Case Cooling : 9 Fan

Everything is brand new except ssd 840pro and some fan.

RAM is on motherboard QVL

Motherboard is on RAM QVL

2) Issue description:

/!\ Everything in bios is set to default (no overclock or xmp settings for ram)

When I cold boot my computer, it stuck on Dr Debug code 54 and do nothing (no reboot/boot loop, just stuck on code 54).

Sometimes if I manually reboot it pass the code 54 and start normally. Sometime it cold boot normaly (very rare).

In general boot sequence is :

- Cold boot : Stuck on code 54

- Manual reboot : Stuck on code 54

- Manual reboot : Ok it boot and works fine

After that sequence, I can go into bios, windows and no issue. I can hot reboot everytime I want but if completely shutdown my computer it will be stuck on code 54 next time I will power it on.

3) What I've done :

- Clear CMOS (with button, jumper and removed the battery)

==> Nothing changed

- Tested each ram stick in every possible place (alone or with 2 stick)

==> Nothing changed

- Tested multiple bios version (2.00, 2.20, 3.00, 3.20)

==> Bug reproduced in each version but 2.2 seems to boot correctly more often but can be stuck at startup on code 54 like other bios version

- Modified VDDCR_SOC to 1.1, 1.2 and 1.3 and set the LLC to lvl 1

==> Nothing changed exept better boot chance in 2.2 bios version but issue still present.

- To exclude faultly ram stick I did a long HCI memtest (3000 % on 16 process with 800mb tested by each process)

==> 0 errors :

- Intel burn test 20 pass with 8192 ram tested

==> No issue

- Test with custom bios settings (bios 3.20) :

set geardown mode to on/auto
bank group swap to disabled
bank group swap alt to enabled
proc odt to 60 ohms
soc to 1.05 v
soc llc level 2
dram volt to 1.375v
and make sure the am4 boot training is set to auto

==> Nothing changed

- Test with fail_CNT 5/7/9

==> Nothing changed except no F9 error code when I cold boot with XMP settings on (but stuck on code 54)

4) Test request from ASRock Support TEAM :

ASRock Support wrote:

1. Please store your current BIOS settings in a user profile in BIOS.

Load default settings via F9.

Load the XMP profile in BIOS, then set the DRAM frequency to DDR4-2400.

Save via F10 and let the system boot, and do not change other BIOS settings for now.

Memory modules should be in slots A2 and B2.

Please test cold boot with these settings.

==> Done, not working, issue still present

ASRock Support wrote:

If this has no effect then it looks more like a physical hardware problem. It could be related to PSU, motherboard, CPU or a RAM module.

For example a capacitor might not be charging fast enough. Or there could be a micro fracture in one of the components. Then when the system is running, material can expand a little bit and this can be enough to temporarily ?�close??the fracture.

2. Please test each of the RAM modules individually with a cold boot, and make sure that you keep the system powered off at least as long as you normally would.

==> Done, not working, issue still present

ASRock Support wrote:

3. Also test slots A2 and B2 with cold boot individually, one at a time.

==> Done, not working, issue still present

ASRock Support wrote:

4. If possible please borrow another PSU to test with.

==> I had my old PSU from my old 2014 intel build (corsair CX600M). Tested with this PSU but same issue

ASRock Support wrote:

5. If possible please borrow another CPU to test with. Of course I can imagine this step might not be possible.

==> Not possible :(

ASRock Support wrote:

6. Please remove the CPU cooler and the CPU. Place the CPU back into the socket. Close and open the socket lever a few times. Install the CPU cooler. If it is fixed by screws then please make sure that you do not overtighten the screws. Then please test again. This might sound a bit silly. But because (as you probably know) the memory controller is integrated in the CPU, contact issues between CPU and motherboard could also lead to memory problems.

==> Done, not working, issue still present (I did it 3 times. My cooler use manual screw so it's impossible for me to overtight)

5) Conclusion

Now, I don't know what I can do and I need some help from ASRock employes or someone who have better knowledge than me of this kind of issue.

Thanks in advance for your help.

Replies:

Posted By: datonyb
Date Posted: 30 Oct 2017 at 6:29pm

your right
auto should load without issue the weakest cpu off production line and the cheapest ram on amazon

but it dosnt , and its not just asrock its affecting all brands of board to some extent

i found i had cold boot issues that revolved around the ram and the cold boot start up

may i suggest some manual intervention

in bios
set geardown mode to on/auto
bank group swap to disabled
bank group swap alt to enabled
proc odt to 60 ohms
soc to 1.05 v
soc llc level 2
dram volt to 1.375v
and make sure the am4 boot training is set to auto

these above tend to be the main things to help the cold boot issue

-------------
[url=https://valid.x86.fr/jpg250]

[/url]

3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold

Posted By: Misant
Date Posted: 30 Oct 2017 at 7:44pm

Hello,

Thanks for your help.

datonyb wrote:

your right
auto should load without issue the weakest cpu off production line and the cheapest ram on amazon

but it dosnt , and its not just asrock its affecting all brands of board to some extent

Yeah, I know, my point was just to say if I have to modify bios settings to boot, how can I do this if I can even boot (that's why I think I'm lucky to be able to boot sometimes). Maybe I should remove this part from initial post as it can sound agressive (Edit : removed from first post).

datonyb wrote:

i found i had cold boot issues that revolved around the ram and the cold boot start up

may i suggest some manual intervention

in bios
set geardown mode to on/auto
bank group swap to disabled
bank group swap alt to enabled
proc odt to 60 ohms
soc to 1.05 v
soc llc level 2
dram volt to 1.375v
and make sure the am4 boot training is set to auto

these above tend to be the main things to help the cold boot issue

I tried but it's not working, issue still present

Test protocol was :

- Bios 3.20

- Boot into bios (after X cold boot stuck on code 54)

- Changed the settings as requested

- Save

- Boot into windows

- Shutdown

- Wait 5s

- Power on

- Error code 54 and stuck

This protocol was done with

set geardown mode to on

and

set geardown mode to auto

I will update my first post with this informations.

If you have any other settings to try let me know and I will test.

Note : A mail has been sent to ASRock support but actually, no answer.

Posted By: datonyb
Date Posted: 30 Oct 2017 at 8:19pm

" rel="nofollow - it sounds crazy but

have you tried setting the ram to xmp (in case theres a glaring fault with the auto timing settings)

and is the boot training set to 3 or 5 attempts ?

im feeling the problem here is as you seem to feel only the initial cold boot and thus altering the boot attempts will allow things to warm up more and voltages to be more stable when cold booting

-------------
[url=https://valid.x86.fr/jpg250]

[/url]

3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold

Posted By: Misant
Date Posted: 30 Oct 2017 at 8:28pm

datonyb wrote:

" rel="nofollow - it sounds crazy but

have you tried setting the ram to xmp (in case theres a glaring fault with the auto timing settings)

I tried.

It works with hot reboot (didn't test stability in windows) but it doesn't work with cold boot (reboot 3 times with F9 error)

datonyb wrote:

http://forum.asrock.com/RTE_textarea.asp%3cmode=quote&ID=38655&CACHE=128" rel="nofollow -
and is the boot training set to 3 or 5 attempts ?

Where do you change this setting ? is it fail_CNT ?

datonyb wrote:

http://forum.asrock.com/RTE_textarea.asp%3cmode=quote&ID=38655&CACHE=128" rel="nofollow -
im feeling the problem here is as you seem to feel only the initial cold boot and thus altering the boot attempts will allow things to warm up more and voltages to be more stable when cold booting

Yeah it looks like a thermal chainsaw (but it's not always the initial cold boot, sometimes I have to cold boot 10 times to boot. Always no issue with hot reboot)

Posted By: datonyb
Date Posted: 30 Oct 2017 at 8:35pm

yeah should be that fail cnt

i only have the issue with overclocked settings though

if i use stock or xmp its okay i dont even need the am4 boot training

-------------
[url=https://valid.x86.fr/jpg250]

[/url]

3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold

Posted By: Misant
Date Posted: 30 Oct 2017 at 8:37pm

datonyb wrote:

yeah should be that fail cnt

i only have the issue with overclocked settings though

if i use stock or xmp its okay i dont even need the am4 boot training

Ok thanks, I will try this when I'm back from work.

btw, I launched a HCI memtest with xmp settings on to check if it's stable under windows (800MB * 16 process)

Posted By: Misant
Date Posted: 31 Oct 2017 at 3:14am

Misant wrote:

datonyb wrote:

yeah should be that fail cnt

i only have the issue with overclocked settings though

if i use stock or xmp its okay i dont even need the am4 boot training

Ok thanks, I will try this when I'm back from work.

btw, I launched a HCI memtest with xmp settings on to check if it's stable under windows (800MB * 16 process)

So, I tried but issue still present, stuck on code 54 (but now no F9 error when I load XMP settings and cold boot. It not help for the moment but good to know for the futur)

HCI memtest 800% on each process gave no error with XMP settings on.

I had a feedback from ASRock support team. I reported it on my first post with my answer.

Actually, still no solution to that cold boot issue :( but again, thanks for your time and your help.

If you want me to test other settings, tell me :).

Posted By: datonyb
Date Posted: 31 Oct 2017 at 3:19am

you can try to see if this calculator works to help adjust settings

even the ryzen flarex ram may need a tweak

http://www.overclock.net/t/1640919/ryzen-dram-calculator-overclocking-dram" rel="nofollow - http://www.overclock.net/t/1640919/ryzen-dram-calculator-overclocking-dram

-------------
[url=https://valid.x86.fr/jpg250]

[/url]

3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold

Posted By: Misant
Date Posted: 31 Oct 2017 at 5:11am

Again, same issue... I think this will end to RMA (I'm cursed, already RMA CPU, and replaced an Asus prime pro by the Taichi and still have issue :( )

Posted By: datonyb
Date Posted: 31 Oct 2017 at 6:19am

" rel="nofollow - can you not borrow some ram from a friend or family ?

-------------
[url=https://valid.x86.fr/jpg250]

[/url]

3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold

Posted By: Misant
Date Posted: 31 Oct 2017 at 7:37pm

datonyb wrote:

" rel="nofollow - can you not borrow some ram from a friend or family ?

Hi,

Yep I found a random (brand) working DDR4 2133 ram stick from an intel computer and tested with it... same issue, stuck on error code 54. So now we know that ram and PSU can't be the issue. It's CPU or motherboard. CPU was RMA so it should be motherboard or I'm rly rly rly unlucky.

I had an answer from asrock support, I will answer them and update my post later with pics and informations (but as I can see this will end to RMA).

Posted By: datonyb
Date Posted: 01 Nov 2017 at 4:04am

ok thanks for coming back to update us

knowing end results helps us to help everyone

-------------
[url=https://valid.x86.fr/jpg250]

[/url]

3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold

Posted By: ucsmfu
Date Posted: 02 Nov 2017 at 4:03pm

" rel="nofollow - Last few weeks I've been getting a bios code error 30 on my x370 taichi mobo. The computer would randomly restart. Yesterday and today, I've been getting error 00 and my computer would just go black. I'll do a hard reset and my computer works fine again.

Any idea why this is happening ?

Thanks

Posted By: Misant
Date Posted: 03 Nov 2017 at 5:58am

Hello,

So, ASrock support team told me to check if my CPU was rly swapped during RMA process. So answer is yes, first was UA 1713PGT and second is UA 1711SUS

One picture of test with random DDR4 (erk)

And it's confirmed, I have to RMA the board.

Before and after RMA I have to check if problem still present with a test outside the case.

I will do it but actually I'm also checking for segfault issue. I want to use the motherboard RMA downtime to also replace the CPU.

Test is positive :

Next update if I have an answer from AMD.

/!\ Segfault issue and this error 54 issue have absolutly no link between them. Segfault issue can affect every ryzen 7 CPU before week 25 on all motherboard

ucsmfu wrote:

" rel="nofollow - Last few weeks I've been getting a bios code error 30 on my x370 taichi mobo. The computer would randomly restart. Yesterday and today, I've been getting error 00 and my computer would just go black. I'll do a hard reset and my computer works fine again.

Any idea why this is happening ?

Thanks

Could you please open your own thread ? Wink

It will be usefull to give more information (hardware ? bios version ? settings changed in bios ? "...")

In general 00 is overclock failed. Don't know for 30.

Posted By: Misant
Date Posted: 07 Nov 2017 at 6:26pm

Good news, RMA for cpu (segmentation fault issue) was accepted by AMD (rly fast answer).

I'm thinking about waiting for the new cpu, test again and RMA the board if needed (depends of shipping time for new cpu)

Posted By: Misant
Date Posted: 17 Nov 2017 at 8:55pm

Misant wrote:

Good news, RMA for cpu (segmentation fault issue) was accepted by AMD (rly fast answer).

I'm thinking about waiting for the new cpu, test again and RMA the board if needed (depends of shipping time for new cpu)

Hello,

Good news, new CPU received today from AMD. It's week 42.

I will do some tests when I will be back from work and report the results here.

Edit: And... still have the cold boot error 54 :/ so I will RMA the board.

Posted By: Misant
Date Posted: 04 Dec 2017 at 5:30pm

Hello,

New ASrock taichi received from RMA last week.

I did some test and no more issue :)

Everything is now working.

Now I'm near stable (have to finish some stability tests) with 4 ghz 1.392vcore and ram at 3200 with XMP settings (but increased ram voltage to 1.40) and 1.075 SOC.

I did :

- 20 pass intel burn test 8192 mo tested

- 12h prime95

- actually testing ram with HCI memtest (16 x 800mo).

The only issue is the multi bug but the workaroung with offset voltage + pstate is working so it's ok for me (maybe the next bios will fix this).

Edit : I forgot to say, I'm realy happy with the quality of support provided by ASRock. They don't let their customer down when facing an issue (same as AMD, when I did a RMA for the CPU).

Misant