cancel
Showing results for 
Search instead for 
Did you mean: 

strange question about consumeII

cdvehicle
Associate II
Posted on April 19, 2005 at 11:08

strange question about consumeII

6 REPLIES 6
cdvehicle
Associate II
Posted on April 18, 2005 at 09:40

could superio's irq infection BIOS read superio's register? if I diconnect irq1...irq12, ofter send post code 0x0c the isa bus's d0..d7 can not toggled.

[ This message was edited by: vehicle007 on 18-04-2005 13:51 ]

cdvehicle
Associate II
Posted on April 18, 2005 at 10:58

I made parallelport float, it can effect superio when BIOS read ''shutdown register''?

marios
Associate II
Posted on April 18, 2005 at 13:20

It should not affect the shutdown register. However, check the datasheet to make sure that you can leave it unconnected. (ot sure about internal weak pullups.)

If you do not see the oscillator running, it may be another proof that the BIOS code is not able to properly turn on and configure the super i/o chip, so you have to solve this issue first.

The reset signal is okay? I'm asking this because I had an issue with the reset signal and another SMSC super I/O chip.

cdvehicle
Associate II
Posted on April 19, 2005 at 00:13

''If you do not see the oscillator running'' ? is it clk32K? It can running normally

Attach is my schematic.

And somebody tell me if I see 0x0d on bus, so the question should be next error code 0x0e, is it right?

[ This message was edited by: vehicle007 on 19-04-2005 06:57 ]

twitwity
Associate II
Posted on April 19, 2005 at 10:50

GeneralBIOS PST CODE table is attatched. code 0D is for accessing CMOS shutdown register

but how can you disconnect IRQ pins while the chips running? if you do that physically while chips are runing it causes clitch to signal pin and the chip may crashes

SIO clk32out and nPOWERON pin should work in normal state and check the CPU. PCI ISA ISA2x and all STPC clk should work normally. after that watch PCI CONTROL signals the PCI control signals should toggle and working. if it is not. at least your board is electrically unstable. factors those can affect all PLL in STPC should be verified. this includes memory and PCI Feedback clock state and any jitter or glitch should be controlled to minimum.

________________

Attachments :

GB-PC.pdf : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006Htiv&d=%2Fa%2F0X0000000aWC%2FlLvq_k9vWrDHzUk3.BPgM6PblcFh57MfsY55RWQQ8J8&asPdf=false
twitwity
Associate II
Posted on April 19, 2005 at 11:08

i got my HW using STPC VEGA at stable state. and VEGA and C2 has same bus topology and i can assume the things inside the chips are same (.. hmm why should they be different?)

1. poweron reset condition should be obeyd. if this is not, the clocks dont show strapped default action. espescially PCI clk runs at a few hundred KHz not in tens of Mhz.

2. ISA activity dies (does not persisits) after a few seconds action. in this case, PCI conrol sognal also dies. those things such as Power place layout, and return path optimization etc. are designed right,

one thing to consider is all the return clocks. i put the clock buffer in PCI and Memory clock path. and put 22ohms of serial termination one at each clk pin. i raised this value upto a few hundreds ohm. this value is totally rule of thum. signal reflection and glitch is somewhat hard to detect. and control, board trace impedance is usually not easy to set up. check visible clks all the time when every things are set right, the board suddenly fonctioning stably.

(check solderint too. resolder, and wash the flux. and watch the signals.. you shouldnt simply ignore board assembly state..)

wish you a good luck.

oh. read the troubleshooting section at the end of the datasheet.. and follow the steps..

[ This message was edited by: Twitwity on 19-04-2005 14:41 ]