Asus CUR-DLS

[JEWilson]

This may be a bad cap issue. Unsure to date.

First, this mobo would not post.
Inserted post diag card and status displays 0C 06.
I understand 0C in Asus Award BIOSes to be decompressing BIOS.
But BIOS is not being loaded.

Surmised it may be a faulty BIOS chip.
Removed device and obtained a BIOS image with a Willem programmer.

Tried erasing the BIOS chip and reloading the original BIOS image above.
This appeared to be complete ok.

Reinserted BIOS chip into mobo and powered on.
Post diag card now reads 00 00 - completely dead mobo.

Used b14 and scanned the obtained BIOS image.
This reports errors on checksum for modules.
However, identified from BIOS string - this a v1009 BIOS (latest).

Visited Asus website and DL's the v1009 BIOS.
Decompressed with b14 and unpacked ok - no errors reported.

Note, the DL'd version reported all the correct modules whereas,
the installed original version did not.

Erased the device again and attempted to install the DL'd v1009
BIOS. Flash fails repeatedly at 67%

Diagnosis - possible the BIOS device is faulty.
Possible, I am doing something incorrectly with the Willem.
Any suggestions or insight?

Thx

[Scenic]

Re: Asus CUR-DLS

sound like the bios chip is toast. happens sometimes..
had the same problem with a foxconn mainboard recently..
flashed the bios file on another bios chip and bingo! the board works fine again

[zandrax]

Re: Asus CUR-DLS

I agree with Scenic: get a new bios chip, the original one may be damaged.

Zandrax

[kc8adu]

Re: Asus CUR-DLS

to test the chip erase it and do a blank check.
or fill buffer with 00 and write it .
then fill with ff and do the same.
if chip bad it wont pass verify.

[JEWilson]

Re: Asus CUR-DLS

Y

Erased, verified and did the blank check.
Passed ok but would not write the DL'd v1009 BIOS.
Odd...

The original bios image can be rewritten back to the device
with no errors but as that image came originally from the chip,
it could be suspect as discussed.

I will need to get an equivalent for the SyncMOS F29CS1004T or
another.

BTW, pulled the mobo out of the case.
Found two Fujitsu RE FPCAP starting to dome at the VRM.
These are 680uF 6.3V on the O/P side of the VRM. ESR is ok
The remainder of the VRM comprises 6.3V 1500uf Nichicon HD.
ESR is ok.

Will replace the two bad FPCAP with 560uF 4V Sanyo OSCON SEPC as
I have these in stock. Voltage should be ok as the CPUs dial
up 1.75V only.

I intend to replace the Nichicon HD with Pana FM 1500uF 6.3V
as the mobo is dated from a preventative maintenance PoV.
Will replace 1 of the 1500uF with 2200uF to take account of
the drop in capacitance re - SEPC in the VRM.

Will let you know of progress

Thx

[kc8adu]

Re: Asus CUR-DLS

i doubt the drop from 680 to 560 will hurt anything.
the sepc likely have lower esr to balance.
you mean that you cant write the .bin to the chip with your programmer?

[JEWilson]

Re: Asus CUR-DLS

Y - the bin file, the BIOS image

Let me explain

First, removed the BIOS chip from the mobo
Got an image straight away with the programmer and saved

Next, erased the BIOS chip above ok with the programmer
Read the chip ok
Verified the chip ok
Tested chip empty ok all with the programmer

Took the saved image and wrote this to the device i.e. programmed
with the programmer.
Read the chip ok
Verified the chip ok

Installed the BIOS chip back into the CUR-DLS, got nothing
at boot (as prior post) i.e. 00 00

Checked the attributes of the BIOS image saved originally
with the same verision DL'd from Asus - used B14 utility.
Both are the same version i.e. v1009

Now the original saved image reported CRC checksum errors
and B14 only listed 1-3 of the bios modules.
With the Asus sourced bios image, all modules reported
including scsi option rom

So...
Took the BIOS chip from the CUR-DLS and programmed it
with the V1009 bin file from Asus but programmer fails
as discussed above at 67%

Conclusion - the BIOS chip is faulty

I have attached both of the reports from b14 util if you want
to see what I mean

Also, the Fujits and the SEPC have exactly the same spec
i.e ESR 0.007, ripple 6100mA

Thx

[Scenic]

Re: Asus CUR-DLS

I will need to get an equivalent for the SyncMOS F29CS1004T or
another.

as i couldnt find anything for that thing on g(.)(.)gle i searched for the manual of that board... 4MBit BIOS Chip on a S370 platform...?! holy $/-/17!

[Wizard]

Re: Asus CUR-DLS

divide by 8 to get bytes.

which is 512KB bios. Not big.

Recently, I had to update very, very recent Samsung TVs that has firmware that weighs in at massively 20MB.

Cheers, Wizard

[pfrcom]

Re: Asus CUR-DLS

I had no problems programming DLS_ASU.009 into an SST 39SF040 chip, using a Willem Enhanced with software 0.97ja

What make & model is the chip that came out of the board ?

Maybe previous owner substituted an LPC type chip like SST 49LF040, instead of necessary JEDEC type like SST 39SF040

[pfrcom]

Re: Asus CUR-DLS

Just noticed you already said the existing chip was a SyncMOS F29CS1004T

Should that be F29C51004T ?

Anyway, SST 39SF040 would be a good replacement - ASUS download site says support for that chip went into BIOS version 1003

[Scenic]

Re: Asus CUR-DLS

divide by 8 to get bytes.

which is 512KB bios. Not big.

Recently, I had to update very, very recent Samsung TVs that has firmware that weighs in at massively 20MB.

Cheers, Wizard

well.. most socket370 based boards (including some server boards) i had here had a 1MBit or (sometimes) a 2MBit chip... but 4MBit..?
i've only seen 4MBit BIOS Chips on more recent boards until now (Socket939, LGA775, ...)

edit:

Just noticed you already said the existing chip was a SyncMOS F29CS1004T

Should that be F29C51004T ?

i think so..
no result for F29CS1004T.

but with F29C51004T:
http://www.datasheetarchive.com/F29C...datasheet.html

[JEWilson]

Re: Asus CUR-DLS

Y

My mistake. The device is a SyncMOS F29C51004T

Interestingly, the manual states the board supports a
4Mbit FWH. However, the device above is 5V only.

FYI - Intel sourced boards have supported 4Mbit FWH since
i810 days.

As the device had a label on it which denoted RDLS9 1005,
I suspect this indicates a v1005 BIOS.
Even more odd, is the fact, I do not believe a v1005 BIOS
exsits as I cannot find it on the Asus website.
In any event, it was updated to v1009.

It may be an OEM BIOS but from the reports attached,
the BIOS string makes it clear, it is an Asus mobo.

I have a number of 4Mbit FWH devices, it may prove
worthwhile to try a FWH. Then again, the device may
get fried!

[pfrcom]

Re: Asus CUR-DLS

Only reference to FWH in downloaded CURDLS-100.PDF refers to item 17 of the component locations

I reckon that's a typo

BIOS history shows support being added for SST 39SF040 and MXIC MX29F040, neither of which are FWH chips

If you really want BIOS 1005, try ftp://ftp.asus.com/pub/ASUS/mb/sock3...works/cur-dls/

[JEWilson]

Re: Asus CUR-DLS

Agreed - typo

Thanks for the link - appreciated

[pfrcom]

Re: Asus CUR-DLS

On my Willem, for programming a 4mbit chip like an SST39SF040, it's necessary to change an additional jumper - it's just to the left of the DIP switches on mine

If I don't change that jumper, it only gets half way through programming

Could that be your issue ?

Compare the diagram shown by the Willem software when you specify device SST39SF040 versus SST39SF020 (a 2mbit chip)

[JEWilson]

Re: Asus CUR-DLS

Ordered a AM29F040B from RS Components

Flashed with the v1011 BIOS
The mobo POSTs!

Will get round to the recap of the VRM now

Thx

[kc8adu]

Re: Asus CUR-DLS

i figured it would post with a valid bios.
i have seen a few boards with damaged flash eeproms act like that.
just had a given up for dead vp6 land on my bench with a eeprom which would not program and pass verify.would display garbage on post screen.
dead giveaway was the checksum.
a fresh chip,topcats modded bios,and a recapping yielded a working vp6.

[JEWilson]

Re: Asus CUR-DLS

I've recapped the CPU VRM.
I understand this to be a +5V VRM and there is one for each CPU
but these are run from separate VRM controllers.

Q - the remaider of the bulk capacitance on the board is
16 Sanyo 1000uF 6.3V series CA and CG. There is also
a number of 16V 100uF GEC caps. The latter I'm not
bothered about.
However, I am concerned these Sanyo caps are under
typical voltage spec. for a mobo i.e. with the exception
of the voltage block for memory, I would have expected
to see a number of 10V caps.

I suspect at production, these 6.3V caps have been
inserted possibly (or typically) as a result of there being
no 10V parts to hand.

Any suggestions or insight?

[PCBONEZ]

Re: Asus CUR-DLS

On motherboards both 10v and 6.3v caps are used on 5v and 3.3v power.
Interchanging 6.3v and 10v is no problem what so ever.
In fact, though they are commonly used, 10v caps aren't needed at all on motherboards.
.
They (10v) get used when 6.3v of the desired specs or size aren't available.