K7S5A, batterie Problem

[Steve]

Quotes from here:

In an ATX board that is in soft-off state (which is what happens when it powers down on its own, or the power button is pressed), the power supply's Standby rail provides the backup voltage to the RTC and CMOS RAM, as well as the chipset south bridge so you can push the power button again to make the south bridge wake the PSU and the rest of the system.

The "lost CMOS" issue comes in when during this power supply transition from Standby-Only to Full-On, the Standby rail takes a nosedive below the threshold where the RTC chip sets its "low voltage" flag. BIOS in turn uses that flag to then assume the CMOS RAM contents cannot be trusted, and there's your CMOS error message.

The battery on the board is only used when the board is in Physical Off state.

And:

Standby power going away (to physical Off state) isn't the problem - that's covered by the battery taking over.

The problem is the transition from Standby to On, where on many cheap power supplies one can observe the Standby rail dropping below the limit tolerated by the RTC while firing up the main rails. Now while the board is in On state, the Standby rail is supposed to feed the RTC, doesn't, and causes the RTC to complain.

In simple words, the Standby voltage may drop or go away altogether in Standby state (transitioning to Off), but in On state, it must be stable.

 

[Rottie]

If you need some help there is a K7S5A forum they might have some answer for your problem.

 

[LTC8K6]

Link is not working, sm8000.

It sounds like a test is in order on my back up computer.

 

[Steve]

Link fixed