svidalert explanation

[myth77]

In my history of repairing laptops as a hobby i have found several boards that had either dead pch or ite or.... and after i replaced it i get to the point of having every single voltage except +vcc (core cpu). On cpu PWM CONTROLLER i had +5v, ALL_SYSTEM_PWRGD... every one except svidalert. it would be low in voltage and active. The resistance on vcore was always fine.. but svidalert was on. SVID_CLK AND SVID_DATA are always fine.
According to the datasheets of the controller "SVID ALERT# is pulled low by the controller to alert the CPU of new VR/12/12.5 Status". Can someone explain me what is the meaning of "VR/12/12.5 Status"? What is svid alert checking? is it only cpu or it can be pch? or?
On almost every topic i read regarding subject the conclusion is if you have all voltage but svid alert is low then cpu is dead. But some of them replaced cpu and had exact same problem!?

[loic2907]

Re: svidalert explanation

Same problem here. CPu replaced and problem is the same

[Sephir0th]

Re: svidalert explanation

In most of the cases it is a dead CPU from what I know. However, recently got noticed about a case where it was a burned reisistor around the CPU, so maybe always worth to do a visual inspection before concluding the worst.

[myth77]

Re: svidalert explanation

yes, i have also heard in one place the fault was resistor on pch line.
But, it would be great to know what svidalert is checking...

[piernov]

Re: svidalert explanation

See Figures 9, 10, 12.
Detailed explanation of SVID is in another document that I don't think has leaked.

[ferah831208]

Re: svidalert explanation

See Figures 9, 10, 12.
Detailed explanation of SVID is in another document that I don't think has leaked.

I have same problem with an acer nitro 5, mb la-h501p, original fault was water damage around KBC, replaced it and flashed with vertyanov, founded several traces damaged, and laptop worked for 15 days perfect, then it returned with this specific fault. All voltages except vcore, all pwrok high, I have vcc_sa stable because they used a resistor configuration for NO 0v vboot in this specific rail (ncp81215), so I don't have svid protocol active and have 0v in svid_alert, I suspect also about CPU but with the document u provided I'm a little bit confused. In startup sequence if vboot is programmed to a vid setting other than zero is optional to assert alert# at the end of vboot ramp. But maybe what they mean is it will assert after vboot and de-assert again and wait for svid command, in that case it shouldn't stay asserted and maybe that's why I could assume CPU is not working properly, am I analizying this correctly 🤔🤔?? Thanks in advance

[ferah831208]

In my history of repairing laptops as a hobby i have found several boards that had either dead pch or ite or.... and after i replaced it i get to the point of having every single voltage except +vcc (core cpu). On cpu PWM CONTROLLER i had +5v, ALL_SYSTEM_PWRGD... every one except svidalert. it would be low in voltage and active. The resistance on vcore was always fine.. but svidalert was on. SVID_CLK AND SVID_DATA are always fine.
According to the datasheets of the controller "SVID ALERT# is pulled low by the controller to alert the CPU of new VR/12/12.5 Status". Can someone explain me what is the meaning of "VR/12/12.5 Status"? What is svid alert checking? is it only cpu or it can be pch? or?
On almost every topic i read regarding subject the conclusion is if you have all voltage but svid alert is low then cpu is dead. But some of them replaced cpu and had exact same problem!?

Maybe this can be helpful for somebody, I managed to solve a failure in motherboard la-h501p of an acer nitro 5 laptop, all voltages were ok except vcore, and also svid_altert low, in this case it wasn't CPU or PCH dead, although in many cases it seems to be the case. The point is for some reason in this system u can have all voltages present, all pwrok rails active.indicating correct initialization of system, and PCH is not properly initialized, in my case it was a damaged resistor in a PCH pll rail, but after this i made several test because I have some doubts, u can open 24 MHz crystal and get same result, all voltages and power ok signals without PCH crystal 🤷🏻. The point is u don't have svid protocol active because CPU fails to communicate with vcore chip and signal svidalert is in low level, because of not proper initialization of PCH, I tried several ways to simulate this same result, only when PCH was properly initialized and svid protocol started communicating then svid alter was high level with the pulses to 0v to indicate activity. Hope this can be helpful for anybody