No resume from S3 (suspend to ram) - Faulty PSU?

[JimboLatte]

Hi everyone,

as of recent, my PC won't wake from suspend anymore. It still reacts to wake-on-PS/2 from the keyboard, because fans and hdds are spinning up, but the PCI-E video-adapter as well as the USB-Ports won't resume.
I already confirmed that it's not an OS issue. So I'm thinking mobo or psu.
Since the latter (Enermax Liberty 400W) is already seven years old, that seems a good guess. Appart from the sleep issue, the system is fine (albeit being ~5years old):

Asus P5Q Pro w/ Q6600 non-OC
2x2 OCZ Platinum DDR2 @ 800MHz
Gigabyte GTX650 Ti non-OC (recently upgraded)

Because I suspected a faulty +5VSB supply, I took some measurements:
PSU not connected to mobo: 5.15V
PSU connected, system off (S5): ~ 4.8 V
system on (S0): ~ 4.5V
The last value seems a bit low to me. But here comes the weird part. Upon suspend, the system powers down as expected, but so does the standby-line....
system suspend (S3): ~ 3.9 - 4.0 V
Upon wake the voltage ramps to 4.5V again, but as mentioned the screen stays blank etc.

I'm confident the problem is with the low standby voltage, but could someone more experienced comment before I go ahead and buy a new PSU?
Unfortunatly I don't have a spare or one to borrow.

Many thanks and happy new year!

 

[Dr-Kiev]

I would search for bulged capacitors either on motherboard and PSU, for the first.

 

[JimboLatte]

I would search for bulged capacitors either on motherboard and PSU, for the first.

Already opened the PSU and didn't spot anything. But a bulged cap might be easy to miss, so I'm gonna look again. Thanks!

 

[Sunburn74]

Hi everyone,

as of recent, my PC won't wake from suspend anymore. It still reacts to wake-on-PS/2 from the keyboard, because fans and hdds are spinning up, but the PCI-E video-adapter as well as the USB-Ports won't resume.
I already confirmed that it's not an OS issue. So I'm thinking mobo or psu.
Since the latter (Enermax Liberty 400W) is already seven years old, that seems a good guess. Appart from the sleep issue, the system is fine (albeit being ~5years old):

Asus P5Q Pro w/ Q6600 non-OC
2x2 OCZ Platinum DDR2 @ 800MHz
Gigabyte GTX650 Ti non-OC (recently upgraded)

Because I suspected a faulty +5VSB supply, I took some measurements:
PSU not connected to mobo: 5.15V
PSU connected, system off (S5): ~ 4.8 V
system on (S0): ~ 4.5V
The last value seems a bit low to me. But here comes the weird part. Upon suspend, the system powers down as expected, but so does the standby-line....
system suspend (S3): ~ 3.9 - 4.0 V
Upon wake the voltage ramps to 4.5V again, but as mentioned the screen stays blank etc.

I'm confident the problem is with the low standby voltage, but could someone more experienced comment before I go ahead and buy a new PSU?
Unfortunatly I don't have a spare or one to borrow.

Many thanks and happy new year!

I would recommend actually tweaking some voltages on your RAM. I remember having similar issues on a gigabyte board which were responsive to slightly increasing RAM reference voltages. Also, is this predominantly a cool boot/resume issue?

Check the asus forums. I'm sure you will find some info there as well.

 

[JimboLatte]

I would recommend actually tweaking some voltages on your RAM. I remember having similar issues on a gigabyte board which were responsive to slightly increasing RAM reference voltages. Also, is this predominantly a cool boot/resume issue?

Already tried upping the voltage from 1.8 to 1.9V with no effect. Tested both modules separatley. I'm thinking of defective hardware 'cause the problem appeared rather suddenly. Suspend used to work, although I always had issues with premature wake. But at least the hardware was reinitialized properly...

 

[BonzaiDuck]

Already tried upping the voltage from 1.8 to 1.9V with no effect. Tested both modules separatley. I'm thinking of defective hardware 'cause the problem appeared rather suddenly. Suspend used to work, although I always had issues with premature wake. But at least the hardware was reinitialized properly...

I encountered this sort of thing as I attempted to make all five of the desktops of our household use sleep-state to reduce power. If the other machines were over-clocked, it would've complicated my path of discovery, but they weren't. . .

The prevailing wisdom. Some motherboards only allow for S3 sleep versus S0. If they don't "do" S3 regardless -- you can pretty much guarantee it's an aging PSU.

I didn't even bother to test the PSUs in two of our machines. My "sleep-state" obsession occurred around 2011 or 2012; one of the systems had an OCZ PowerStream -- deemed "beefy" and top-end -- which was first purchased in late 2004. The other one used an "ULTRA" model -- cheap but glitzy, and it, too, had been in use for too many years.

We just replaced them both with Seasonic 400 or 450W units, and they sleep and wake very well now. The two desktops are low-end C2D systems with low to midrange graphics cards and either a single SSD or two HDDs.

Didn't even bother with the "PSU tester" or multi-meter.