I think I need assistance from more knowledgeable people. I just bought and built a new rig.
New parts:
Case - Raidmax Atomic (copycat of the Raijintek Metis)
MoBo - ASRock FM2A88X-ITX+ (bios v2.40, released 15/8/2014)
CPU - AMD A10-7800 APU Boxed
RAM - G.Skill Ares DDR3-2133 2x4GB (CL9-11-10-28) (jan2015 production)
SSD - Samsung 850 Evo 250GB
PSU - Fractal Design Newton R3 600W
Old parts:
"Monitor" - HDTV (via DVI-HDMI)
CPU cooler - Noctua NH-U12F
Case fan - Noctua NF-S12-1200
So, first I built everything inside the case, assuming everything will be fine. Left the side panel open. And actually, with the first boot, everything was fine, except for the CPU and case fans running at full RPMs. Checked the UEFI bios for temps, and they were perfectly fine, around 37-43 celsius. Set the fan speed control to "Silent Mode". Restarted, but the fan speeds stayed the same (maxed).
Checked the RAM, and it was running at 1600MHz, instead of 2133Mhz. The bios recognized the XMP timings that were exactly what the RAM was specced for: 2133-9-11-10-28. The motherboard is also specced to natively support the 2133MHz (even 2400MHz is supported via OC). So chose the XMP 2133 profile for use. Restarted, got back into bios, RAM was now running at 2133MHz, temps staying the same. Tried to further experiment with different fan profiles, chose custom, set the temps/RPM like this: 35C/20% -- 45C/50% -- 55C/80% -- 65C/100% (or something like that). Restarted, but the RPMs stayed the same. Tried different lowest values: 35C/0%, 45C/10%. Restarted, no change, fans still running full blast, temps still the same (around 40C). Checked the internet (from another computer), and it seems AsRock's fan speed control is notoriously very poor, not working like it's supposed to. Manually checked the heatsink to make sure that the CPU is really running cool, not just the bios saying it is. And indeed, it was fairly cold to the touch, like a metal would feel like. Decided to leave it for now.
Decided to boot into a Live USB (Linux, USB2). Everything worked fine (though fans still at 100%), system booted up, WiFi was working, I partitioned the SSD, all going great. Shut down the system, tried another Live USB (USB3). Everything worked fine. Turned off the system, switched to the old computer to make a new Live USB (to replace the one in the USB3 stick). Made it OK, turned the old computer off. Still once tried the USB2 Live USB with the new rig, everything still in order. Decided to close the case, so shut down the computer, aligned the wires to fit inside the case, screwed the side panel shut.
This is where the problems started:
Plugged in the USB2 for final test (before I would install the Live USB (USB3) to the SSD). So, switched the computer on. Fans still at full blast, like before. But: BLANK SCREEN. No bios screen, no Live USB screen. Nothing. Something must have come off when aligning the wires or something, right? So opened the case and started to check through the components one by one. But everything seemed to be in order. This required a more thorough inspection. Now started hunting for the culprit.
First tried different video connections (DVI, HDMI, D-Sub) - nope
Unplugged mouse and keyboard - nope
Unplugged the SSD - nope
Unplugged front panel audio - nope
Unplugged front panel USB2&USB3 - nope
At this point, took everything outside the case, at the same time plugged in the PC speaker for error beeps. Visually checked all the connections, but there was nothing out of the ordinary.
Reseated the ATX12V - nope
Reseated the Power LED, HDD LED, Power Switch cables - nope
Reseated the MoBo 24pin power cable - nope
Removed the WiFi adapter - nope
Tried all the different RAM combinations (Ram1 in Slot1, Ram1 in Slot2, Ram2 in Slot1, Ram2 in Slot2) - nope
Reseated the CPU - nope
Cleared the CMOS with jumper - nope
Removed the CMOS battery for 30 seconds - nope
(also checked the battery charge: 3.04V -- anything above 2.5V should be good to go)
Removed everything except the 24pin and ATX12V power cables, CPU+cooler and PC speaker - nope
Removed the CPU+cooler - nope
At this point, there isn't much left. The 24pin cable is mandatory, I can't remove that. And PC speaker is necessary if I actually want to know what's wrong with the system. Even tried without ATX12V, but that thing is also mandatory, if I want to get the CPU to operate.
Tried the PSU on the old Computer - works just fine, so at least the PSU shouldn't be at fault
Built the core system again: power cables, MoBo, CPU, 1x RAM (+PC speaker) - nope
Tried discrete PCIE graphics card from the old rig - nope
In the end, I think I've now tried every logical combination that I can think of (probably have forgotten to mention some of them, too). No beeps whatsoever. I did test the PC speaker, it is working.
So, from what I've understood, that pretty much leaves MoBo and CPU as the last remaining potential fault points. Considering that the CPU/APU has the integrated graphics unit, which therefore handles the image, there's a chance it has a fault in there. But, I haven't heard a single beep from the PC speaker, which would lead to MoBo fault, instead. Apparently it could also mean a faulty PSU, but like I said, it's working OK on another rig.
There's no electrical burned smell, and no bulging capacitors or bent pins that I could see. Visually checking nothing seems to be wrong. The only potential anomaly would be three capacitors running hot; two higher caps near the USB3 header, and one in front of the SATA connectors. All the other caps stay cool, but these three are hot to the touch after a while (after about 30-60 secs of running), unless separately cooled. By hot to the touch I mean I can touch them for 2-3 seconds, then instinctively have to pull away my hand. They are not bulging, and they haven't leaked or anything. Apart from the heating, they seem perfectly fine. There's also a NB/SB (?) chip with a passive heatsink, which stays cooler than those three capacitors.
Those specific capacitors are apparently voltage regulators for the CPU, as they stay cool if I remove the CPU. The NB/SB chip still warms up even if no CPU is connected. Or would this rather hint of a CPU fault? My instincts would say it's 60% chance that it's a broken MoBo and 30% chance that it's the CPU. 10% for everything else.
Also tried sleeping over the night, then rebooting again, when the system is cold. Still blank screen.
Anyone have any ideas to try? Or opinions on what could be the problem? I don't have any other FM2 MoBo, CPU or DDR3 memory to try out, though.
New parts:
Case - Raidmax Atomic (copycat of the Raijintek Metis)
MoBo - ASRock FM2A88X-ITX+ (bios v2.40, released 15/8/2014)
CPU - AMD A10-7800 APU Boxed
RAM - G.Skill Ares DDR3-2133 2x4GB (CL9-11-10-28) (jan2015 production)
SSD - Samsung 850 Evo 250GB
PSU - Fractal Design Newton R3 600W
Old parts:
"Monitor" - HDTV (via DVI-HDMI)
CPU cooler - Noctua NH-U12F
Case fan - Noctua NF-S12-1200
So, first I built everything inside the case, assuming everything will be fine. Left the side panel open. And actually, with the first boot, everything was fine, except for the CPU and case fans running at full RPMs. Checked the UEFI bios for temps, and they were perfectly fine, around 37-43 celsius. Set the fan speed control to "Silent Mode". Restarted, but the fan speeds stayed the same (maxed).
Checked the RAM, and it was running at 1600MHz, instead of 2133Mhz. The bios recognized the XMP timings that were exactly what the RAM was specced for: 2133-9-11-10-28. The motherboard is also specced to natively support the 2133MHz (even 2400MHz is supported via OC). So chose the XMP 2133 profile for use. Restarted, got back into bios, RAM was now running at 2133MHz, temps staying the same. Tried to further experiment with different fan profiles, chose custom, set the temps/RPM like this: 35C/20% -- 45C/50% -- 55C/80% -- 65C/100% (or something like that). Restarted, but the RPMs stayed the same. Tried different lowest values: 35C/0%, 45C/10%. Restarted, no change, fans still running full blast, temps still the same (around 40C). Checked the internet (from another computer), and it seems AsRock's fan speed control is notoriously very poor, not working like it's supposed to. Manually checked the heatsink to make sure that the CPU is really running cool, not just the bios saying it is. And indeed, it was fairly cold to the touch, like a metal would feel like. Decided to leave it for now.
Decided to boot into a Live USB (Linux, USB2). Everything worked fine (though fans still at 100%), system booted up, WiFi was working, I partitioned the SSD, all going great. Shut down the system, tried another Live USB (USB3). Everything worked fine. Turned off the system, switched to the old computer to make a new Live USB (to replace the one in the USB3 stick). Made it OK, turned the old computer off. Still once tried the USB2 Live USB with the new rig, everything still in order. Decided to close the case, so shut down the computer, aligned the wires to fit inside the case, screwed the side panel shut.
This is where the problems started:
Plugged in the USB2 for final test (before I would install the Live USB (USB3) to the SSD). So, switched the computer on. Fans still at full blast, like before. But: BLANK SCREEN. No bios screen, no Live USB screen. Nothing. Something must have come off when aligning the wires or something, right? So opened the case and started to check through the components one by one. But everything seemed to be in order. This required a more thorough inspection. Now started hunting for the culprit.
First tried different video connections (DVI, HDMI, D-Sub) - nope
Unplugged mouse and keyboard - nope
Unplugged the SSD - nope
Unplugged front panel audio - nope
Unplugged front panel USB2&USB3 - nope
At this point, took everything outside the case, at the same time plugged in the PC speaker for error beeps. Visually checked all the connections, but there was nothing out of the ordinary.
Reseated the ATX12V - nope
Reseated the Power LED, HDD LED, Power Switch cables - nope
Reseated the MoBo 24pin power cable - nope
Removed the WiFi adapter - nope
Tried all the different RAM combinations (Ram1 in Slot1, Ram1 in Slot2, Ram2 in Slot1, Ram2 in Slot2) - nope
Reseated the CPU - nope
Cleared the CMOS with jumper - nope
Removed the CMOS battery for 30 seconds - nope
(also checked the battery charge: 3.04V -- anything above 2.5V should be good to go)
Removed everything except the 24pin and ATX12V power cables, CPU+cooler and PC speaker - nope
Removed the CPU+cooler - nope
At this point, there isn't much left. The 24pin cable is mandatory, I can't remove that. And PC speaker is necessary if I actually want to know what's wrong with the system. Even tried without ATX12V, but that thing is also mandatory, if I want to get the CPU to operate.
Tried the PSU on the old Computer - works just fine, so at least the PSU shouldn't be at fault
Built the core system again: power cables, MoBo, CPU, 1x RAM (+PC speaker) - nope
Tried discrete PCIE graphics card from the old rig - nope
In the end, I think I've now tried every logical combination that I can think of (probably have forgotten to mention some of them, too). No beeps whatsoever. I did test the PC speaker, it is working.
So, from what I've understood, that pretty much leaves MoBo and CPU as the last remaining potential fault points. Considering that the CPU/APU has the integrated graphics unit, which therefore handles the image, there's a chance it has a fault in there. But, I haven't heard a single beep from the PC speaker, which would lead to MoBo fault, instead. Apparently it could also mean a faulty PSU, but like I said, it's working OK on another rig.
There's no electrical burned smell, and no bulging capacitors or bent pins that I could see. Visually checking nothing seems to be wrong. The only potential anomaly would be three capacitors running hot; two higher caps near the USB3 header, and one in front of the SATA connectors. All the other caps stay cool, but these three are hot to the touch after a while (after about 30-60 secs of running), unless separately cooled. By hot to the touch I mean I can touch them for 2-3 seconds, then instinctively have to pull away my hand. They are not bulging, and they haven't leaked or anything. Apart from the heating, they seem perfectly fine. There's also a NB/SB (?) chip with a passive heatsink, which stays cooler than those three capacitors.
Those specific capacitors are apparently voltage regulators for the CPU, as they stay cool if I remove the CPU. The NB/SB chip still warms up even if no CPU is connected. Or would this rather hint of a CPU fault? My instincts would say it's 60% chance that it's a broken MoBo and 30% chance that it's the CPU. 10% for everything else.
Also tried sleeping over the night, then rebooting again, when the system is cold. Still blank screen.
Anyone have any ideas to try? Or opinions on what could be the problem? I don't have any other FM2 MoBo, CPU or DDR3 memory to try out, though.
Facebook
Twitter
Instagram