XO1.5 DCONIRQ ECO

From OLPC
Revision as of 03:23, 15 October 2009 by Wad (talk | contribs) (New page: =Scope= This describes a change to early prototypes of the XO-1.5 motherboard. It fixes an error in GPIO assignments made on the B2 (and earlier) prototypes. Boards witho...)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Scope

This describes a change to early prototypes of the XO-1.5 motherboard. It fixes an error in GPIO assignments made on the B2 (and earlier) prototypes. Boards without this ECO will not reliably detect a DCON interrupt.

This only applies to XO-1.5 B2 phase motherboards.

Problem Description

The problem is that SMBALRT# is synchronized by the RTC clock (32 KHz) before being used as an interrupt. The DCON interrupt pin is asserted for one line period, or roughly 15 uS, and is thus usually missed.

GPIO12 provides a general purpose input capable of generating an SCI/SMI, which isn't synchronized to the RTC clock, nor is it "owned" by ACPI.

One concern is that GPIO12 is used by the processor when PCIRST# is asserted to determine the FSB clocking rate. This requires a pullup to +3.3V. If for some reason the DCON were to assert DCONIRQ during that time,

Procedure

  1. Run a wire between R123 (the pad that isn't connected to +3.3VSUS --- this part is located on the "top" side of the motherboard) and the pad of R264 that is connected to R263 (these are located on the bottom side of the motherboard).

Alternatively:

  1. Run a wire between U17 (the DCON), pin 6 and the pad of R264 that is connected to R263 (these are located on the bottom side of the motherboard).