Desktop boots without display or turns off occasionally.

moltenskull

Journeyman
My desktop's configuration (built around 2012-2013):
1) CPU - AMD FX 8350 + Corsair H80i GT
2) GPU - Sapphire R9 280X Tri-X OC
(Received upon RMAing R9 280X Dual X)
3) RAM - G. Skill RipjawsX 16GB.
4) Motherboard - MSI 990FXA-GD65 V2.
5) PSU - Corsair TX 850M
(bought in August 2018)
6) Peripherals
(mentioning because they have some relation with the problem)
- Logitech G29
- Logitech G402
- Corsair K55
- Thrustmaster TH8A
- Logitech Extreme 3d Pro
7) Windows 10 Home Single Language v1809
(no updates available)
8) 1TB WD Blue + 2TB Seagate HDD
(No problems in Crystal Disk Info and HWiNFO)
No component is overclocked.
Problem I am facing:
On booting up, the fans spin, HDD light glows, USB devices (namely, the G29, the G402 and the K55) calibrate but both displays remain in power off mode with a "No signal" message. Restarting with the restart button or shutting down with the power off button does not work. Only force restart works. There are no BIOS beep codes (checked multiple times) and this also happens occasionally after the system is running normally , irrespective of whether it's idling or under load. All temperatures in HWinfo remain well within limits and the only error message is - " Radeon Wattman Settings were reset to default. " (GPU temps are always within 72-78- No AC room).
Replacing the power supply in August last year was the most recent change.Was done @ Cost2Cost,Nehru Place by their staff.
Running Kaspersky Total Security (fully updated).
Please advise as to what could be the problem. Thanks.
 

whitestar_999

Super Moderator
Staff member
Confirm whether it is display issue or something else by logging remotely into your pc from another system(or even android mobile with MS remote desktop app) in same local network to see if you can login when display is not working.If yes then issue is with graphics card else issue is probably with mobo.
 
OP
M

moltenskull

Journeyman
Happened again today on cold boot. Tried logging in with Android version of Microsoft Remote Desktop with IP address on WiFi. Connection failed and could not login. What do you suggest next? @whitestar_999
 

cute.bandar

Cyborg Agent
Restarting with the restart button or shutting down with the power off button does not work. Only force restart works.
restart button means reset button ? Try holding down the reset button

Are the GPu temps 78 @ load or idle?
Try running without ram to see if it generates beep codes.
Try booting with only essential components - PSU, mobo, CPU, RAM

Google suggests a problem with amd wattman software/driver Default Radeon WattMan settings have been resto... | Community "Default Radeon WattMan settings has been resto... | Community
 

whitestar_999

Super Moderator
Staff member
This means issue is probably with mobo(most likely),ram(least likely) or psu(moderately likely). If there is some local pc hardware shop nearby then you can get your mobo+cpu checked with their psu & gfx card or get your psu+gfx card checked with one of their systems.
 

patkim

Cyborg Agent
One more simple test you can do it swapping RAM to another free DIMM slots. If you have 1 x 16GB then put it in another slot. If 2 X 8 then put them in other slots. I guess your mobo has 4 DIMM slots and in case of 4 x 4 then reduce it to just 2 x 4 and retry.
Is the HDD LED glowing continuously when this issue occurs?
 
OP
M

moltenskull

Journeyman
@cute.bandar Using a CM HAF 912,which has a reset/restart button and a power down button. By force restart, I mean I have to hold the power down button for 5 seconds before the system shuts down and I can power it up again. The restart button does not work then. GPU temps at 78 while under load,idle is 37 Deg. C. Will try other suggestions . I checked the thread that you posted ,in it finally the OP had a problem with thermal paste application but I think that should not be the problem in my case. CPU temps are fine in both AMD overdrive and HWinfo.

@whitestar_999 I guess I will take it to Nehru Place and get it checked. Local PC shops in my area would refuse,they are just interested in selling. :p Btw, the motherboard AUX temp (never got a definite answer as to what that means) in HWinfo is quite high at idle - 57 Deg C. Could you shed some light on this?

@patkim I could try that , I have 4 DIMM slots and a 2x8GB configuration. Now that you mention it, the last time this happened, the HDD light was off but all other USB devices/fans/lights were working.
 

topgear

Super Moderator
Staff member
Keep only the Keyboard and mouse connected while booting up. Disconnect every other USB controllers ie joystick, gamepad, racing wheel. Connect those only when you need. See if it improves the situation.

If not then use different sata data cable and if possible different sata power cable from PSu too.
 

whitestar_999

Super Moderator
Staff member
Different mobos have different temp sensors so can't say for sure but anything below 60C for any mobo temp sensor is good enough(for reference when overclocking/heavy load of ryzen 6 cores(& these are more power efficient than FX series) any cheap mobo temp sensors easily hit 90C+).
 
OP
M

moltenskull

Journeyman
Hi all, thanks for all the suggestions and inputs. Turns out the problem (for now) was that SATA cable for one of the HDDs was loose at the motherboard port (probably happened while cleaning). Pushed it back in and the system has been performing fine since then. :)

Thanks a lot,
moltenskull.
 

whitestar_999

Super Moderator
Staff member
Hi all, thanks for all the suggestions and inputs. Turns out the problem (for now) was that SATA cable for one of the HDDs was loose at the motherboard port (probably happened while cleaning). Pushed it back in and the system has been performing fine since then. :)

Thanks a lot,
moltenskull.
Install crystaldiskinfo free & check ultraDMA CRC error count value(raw value).Any sata cable issue will result in this having non-zero raw value.
 

topgear

Super Moderator
Staff member
Hi all, thanks for all the suggestions and inputs. Turns out the problem (for now) was that SATA cable for one of the HDDs was loose at the motherboard port (probably happened while cleaning). Pushed it back in and the system has been performing fine since then. :)

Thanks a lot,
moltenskull.

great and glad to know.
 
OP
M

moltenskull

Journeyman
Install crystaldiskinfo free & check ultraDMA CRC error count value(raw value).Any sata cable issue will result in this having non-zero raw value.
UltraDMA CRC error count is having a non zero value on one of the HDDs. I Googled a bit and after that , downloaded WD's Diagnostic Tool for HDDs. Extended generic test came up fine and on one post, I read that as long as the value remains constant, it's okay? Is it so?

Also,another fault I found in the BIOS was that CPU voltage was too low and set to a manual value. I forgot to change it back while trying to undervolt the CPU a few months back. Since the then CMOS battery was so old, it was refusing to accept any changes and when I had the CMOS battery also changed at Cost 2 Cost, I forgot to set it back to Auto. Changed it back and now seemingly, all is okay.
 
OP
M

moltenskull

Journeyman
Turns out the final problem with the system was the GPU,which after some days after posting the above reply,gave a red screen of death*. Replaced with a new GPU and having no further problems . Thanks for the help.
Edit:*Was stuck at a red screen of death
 
Top Bottom