bad flash corrupted bios

[kc8adu]

Re: bad flash corrupted bios

i just put a socket on the board and install a chip i wrote with my programmer.
but hotflashing always works.
i had one of those bios saver things but it often left things in a bad state.

[starfury1]

Re: bad flash corrupted bios

OK maybe in reality the resistor is a bad idea with those chips
(which I have no idea on how they work)
but thats basically what I was driving at Harvey (disable when placed on, removed re-enables stuffed bios chip)
that was really just as an example
my days were 27XX eproms not up on the new stuff at all

maybe its doable regardless of MB and thats just "spin doctor talk" on the MB.

I dont know

[starfury1]

Re: bad flash corrupted bios

kikkoman "google is my friend"...thats what got me into trouble with the post
but thanks for the "Correction" to it
...it is as it was thought

Cheers

[starfury1]

Re: bad flash corrupted bios

Thanks I hadn't found that post...and that is really what I was trying to find stuff like that on the tech side of things

Your skills with google are far superior to mine master Yoda.

typing reply's while googling is probably like drink driving its going to get you into trouble!

here's a link to the PDF on that chip W39V040AP

cheers

[starfury1]

Re: bad flash corrupted bios

I still need to work this through from the post link
this is really posted more for info then anything else

few snips from PDF

Reset Operation

The #RESET input pin can be used in some application. When #RESET pin is at high state, the device is
in normal operation mode. When #RESET pin is at low state, it will halt the device and all outputs will be
at high impedance state. As the high state re-asserted to the #RESET pin, the device will return to read
or standby mode, it depends on the control signals

on the replacement socket, /OE (*out enable, active low*) is hardwired
to /reset, thus activating the secondary chip.

So I am assuming THF chip is controlled from /reset line?

on the pcb-side (the empty socket to "Clip" over the pcb-soldered chip)
/OE is wired to an upper NC pin 26 (not connected).
WHY? whats the point.... it must have some purpose to go to the trouble.

All other pins are
wired 1:1* , both chips are in parallel.

Makes sense if you can disable the other chip to a high z stated, as discussed

LPC is designed as a bus
(mind the 3 ID pins).

Guess I'll have to figure out exactly what that means from the PDF, seems there is 2 modes of operation (and ways to program)

I used to think that there might be some resistor or sth. to pull the
pcb chip's */OE-/Init* high (= inactive) while the ECS mainboard
would have like a voltage divider feeding */OE-/Init* to allow this.

seems this is his old theory and I see what he was driving at.

I do not quite get the idea why the pcb-bios does disable it outputs
(if it does)
I think it must but how?

when it is possible that both chips have /OE active (but the
spare chip earlier in sync with /reset). any clue?

whats happening with /OE
why is it taken to an NC pin on the THF socket?

What happens with the /Reset ?



Is there something happing with the combination of pins on the Chip that enable this switching action

Just some thoughts as I go along (and get lost I suppose)

Cheers

[Harvey]

Re: bad flash corrupted bios

on the pcb-side (the empty socket to "Clip" over the pcb-soldered chip)
/OE is wired to an upper NC pin 26 (not connected).
WHY? whats the point.... it must have some purpose to go to the trouble.

Not connected to anything in the chip, but quite likely connected to something on the motherboard.

I'm now even more lost than I was before.

[Paul S]

Re: bad flash corrupted bios

The guide that I posted earlier says to use pin #24 to select which chip to use. This also included cutting the trace, so I don't know if that is a good method.

[starfury1]

Re: bad flash corrupted bios

yeah me too

but quite likely connected to something on the motherboard.

I would have thought so too on the MB

Being in mind I got no idea
Which MODE this thing is in normally while on board

here a bit more blurb

Interface Mode Selection And Description
This device can be operated in two interface modes, one is Programmer interface mode, and the other is
LPC interface mode. The MODE pin of the device provides the control between these two interface
modes. These interface modes need to be configured before power up or return from #RESET. When
MODE pin is set to high position, the device is in the Programmer mode; while the MODE pin is set to
low position, it is in the LPC mode. In Programmer mode, this device just behaves like traditional flash
parts with 8 data lines. But the row and column address inputs are multiplexed. The row address is
mapped to the higher internal address A[18:11]. And the column address is mapped to the lower internal
address A[10:0]. For LPC mode, It complies with the LPC Interface Specification Revision 1.0. Through
the LAD[3:0] and #LFRAM to communicate with the system chipset .

"In Programmer mode, this device just behaves like traditional flash"

Now I am not sure but if you look at table 5 you will see the "Modes" (programmer or LPC)
and * beside function...so do I summarized that depending on mode this will determine if a pin functions actually used?

Read(Write) Mode
In Programmer interface mode, the read(write) operation of the W39V040A is controlled by #OE (#WE).
The #OE (#WE) is held low for the host to obtain(write) data from(to) the outputs(inputs). #OE is the
output control and is used to gate data from the output pins. The data bus is in high impedance state
when #OE is high. As in the LPC interface the "bit 1 of CYCLE TYPE+DIR" determines mode, the read
or write. Refer to the timing waveforms for further details.

Thing is... it must work, question is how, it must be simple

Does the MB design really play a part in how THE THF works?
is the flash firmware on TFH custom to this application?
(does it play a part in disabling the onboard bios)

seems to be raising more questions about it then it answers the pdf
to me anyway

flying somewhat blind here too cause we got no real idea on how the original bios chip is setup on the MB.

cheers

[Paul S]

Re: bad flash corrupted bios

The guide that I posted earlier says to use pin #24 to select which chip to use. This also included cutting the trace, so I don't know if that is a good method.

I should read my own references.

You short out pin #24 to pin #32 to disable it, and connect pin #24 to the board to enable it.

[starfury1]

Re: bad flash corrupted bios

well I dont know of series with these chips and were they are compatible and deviate from each other.
or are just a plain super versions of same

Just on the doc's you posted Paul S from the photos I grabbed the PDF
for the SST MPF 39VF020 Here
(haven't really read it jut looked at the pins)
joining #24 pin /OE to #32 pin VDD would as you say disable and then just use the other /OE to the "modded" PCB is quit easy to understand...and makes sense

point with the THF is it does this without modding the MB and leaving all pins on the MB intact...so they are connected were ever they are even with the THF in place.

So the first question is;
Does the MB actually play a roll in this or "with the right BIOS, is it doable on and MB with that type of BIOS chip".regardless

(he hints at that in the post but basically you do at your own risk)

The method for the Tivo may be doable on the original posters question, MB of this thread , you would need a pre programmed BIOS chip and you would need to perform some micro surgery on MB to see if you could cut the right line.
(and yes a socket would have made this a whole lot easier )

i'll have to have a scratch round the other forum and google some more.


heres the link to all the SST 020 chips pdf's Here
(notice in some of photos its a 010 chip with Tivo

cheers

[kikkoman]

Re: bad flash corrupted bios

here's what I think:

I start to believe that the guy who posted it mixed it up a bit (or i'm stupid and/or can't read)....but connecting the MB BIOS's RST to /OE makes much more sense since RST is driven high all the time in operation and thus disables the first chip.
On this ECS mobo I have here pin 26 actually is connected to GND.

this makes perfect sense. it also means that the THF needs this pin #26 to be grounded although ECS might simply have used the common VSS on pin #16.
if you build your own, I suggest you use that pin instead.

[starfury1]

Re: bad flash corrupted bios

http://www.willem.org/catalog/produc...roducts_id=118

note the "in some cases"

[starfury1]

Re: bad flash corrupted bios

here's what I think:

I start to believe that the guy who posted it mixed it up a bit (or i'm stupid and/or can't read)....but connecting the MB BIOS's RST to /OE makes much more sense since RST is driven high all the time in operation and thus disables the first chip.

that what I was thinking but when I started to work it a bit it didn't make sense so I didn't bother posting on it. (was getting very late)

Rest line when I when to school was active low for a short time then high this would help in disabling if you connect to the right pin

I think it simple how its being done but it eludes me at this point.

I still cant fathom why you would take a line to an unused pin unless you were using it
as a jump point for some other function.
(is that pin 26?)

here a page I found not to do with this but setting up dual bios on a GA MB
(of that site)

I think its from a japan site but not totally sure on that

http://fab51.com/workshop/bios/dual_bios1.html

translation

http://babelfish.yahoo.com/translate...rUrl=Translate

bit hard to make sense of txt wise but think the diagrams should make some sense.

cheers got to run but I'll think on your post later....I too think somehow his description my be wrong

seems you can shut it down by toggling the right pins regardless of some of the other function pins

anyway got to run

[starfury1]

Re: bad flash corrupted bios

OK oh and BTW I missed the data sheet link too Kikkoman ...
thanks for posting it before and DOH! to Me


Remember I still got training wheel on for these chips.

on the replacement socket, /OE (*out enable, active low*) is hardwired
to /reset, thus activating the secondary chip.

So we are now saying its the "Corrupt bios" the lower clip over socket that has the above pins jointed

I agree as far as the function goes the reset would go high taking both pins high and the chip out of action (OP High Z)
(and this below was where I started to say good night Irene)

The "corrupt BIOS" pin (/OE) is soldered to the MB
so now this must connect to some output somewhere (possibly even GND)
what happens when/if it driven low don't we now have a conflict?
(so yeah it may be doable with other MB's but you may have to check that point)

Possible the very action of disabling the chip might force a state that what ever controls this /OE line on the corrupt BIOS is out of action too?

(and why I was wondering if the MB had been designed with this in mind)

this is the bit I have no idea on cause I don't know what the circuit is.

Anyway if the above is the case (and it fits and sound right to me)
it must work lest wise for that MB.

on the pcb-side (the empty socket to "Clip" over the pcb-soldered chip)
/OE is wired to an upper NC pin 26 (not connected).
WHY? whats the point.... it must have some purpose to go to the trouble.

Now that answers "my point" exactly and makes perfect logical sense.

using Pin 26 NC "but ground on the PCB" with the corrupt BIOS
take that to the THF /OE yep I get it.

So apart from My query on the /OE on board BIOS line
this is how it works

(and that plus the pin 26 thing was just adding to the confusion with it)
He seems to have had it right but back to front
(although he mentions something about 2 /OE lines being active??? anyway)

Thanks Kikkoman that sounds right to me

you might be able to tell what drives the /OE line or where its connected on the PCB..

BTW I did find mention of some chips cause flaky behavior, so you might if you are going to attempt something like this (do at your own MB's risk) either use the same type of chip or very close family member technology wise not just a compatible type.
(there was a situation years ago I remember with the Schmitt trigger 7414 74LS14 I think it was, had a different hysteresis or something so some circuits wouldn't behave as they should with the old series)

bottom line if your are going to try this you are going to need a pre programmed (for the MB type) BIOS chip

Yep a socket defiantly would have made life easier!

Cheers

[kikkoman]

Re: bad flash corrupted bios

you might be able to tell what drives the /OE line or where its connected on the PCB..

I've tried that, but it was too tricky in this case. Doable, but lots of work. Most of the EEPROM's lines go to the Super I/O chip via a bunch of resistors, though. I guess /OE does, too.

The "corrupt BIOS" pin (/OE) is soldered to the MB
so now this must connect to some output somewhere (possibly even GND)
what happens when/if it driven low don't we now have a conflict?
(so yeah it may be doable with other MB's but you may have to check that point)

As I said, one might put a 1k resistor between [RST and the 1st /OE] to prevent a short just in case both of these lines are usually driven "hard".
If that is the case, it's simply not gonna work. in this case, you can either cut the /OE trace (or pin) and add a switch OR you have to go the hard way and solder in a socket (but where's the fun with that?)

[starfury1]

Re: bad flash corrupted bios

I've tried that, but it was too tricky in this case. Doable, but lots of work. Most of the EEPROM's lines go to the Super I/O chip via a bunch of resistors, though. I guess /OE does, too.

thought you might have and yeah I remembered one of the posts saying something about it going to a large chip.
(maybe the block digram may show this in the manual, not the /OE line, just talking block layout here)

As I said, one might put a 1k resistor between [RST and the 1st /OE] to prevent a short just in case both of these lines are usually driven "hard".
If that is the case, it's simply not gonna work. in this case, you can either cut the /OE trace (or pin) and add a switch OR you have to go the hard way and solder in a socket (but where's the fun with that?)

I must have missed to bit on the 1K resistor but yeah agree it could be a daunting task
with some MBs
hopefully as you say, lets hope worst case is it don't work but doesn't cause damage

getting something like this to work of cause would be preferred to ripping chips if you not tooled up for it.

but its a risky venture too.

I think any MB thats geared towards OCers and liable to be flashed on a regular basis should really have a socketed chip or a Dual bios setup of some kind

Asus must have fun ...as it seems to be a regular return issue for their MB's
reading posts on other forums

I am now in the Asus club and I did have to flash this to work with G0 stepping CPU.

mostly flashing at Dos type level works ok
but it only takes one dud flash to bring a whole world of hurt


Thanks, Cheers

[gnappoman]

Re: bad flash corrupted bios

so, dear people a simple question 4 you:
I have two identhical MB with plcc soldered bioses, one is good, the other is bad.
Chips are SST49LF080A LPC8 type.
I have also some spare new plcc sockets and one blank SST49LF080A from ebay.
I would like to build my own "Top Hat Flash".
I have no fear of making trouble :-)
Please give your definitive directions on how to do this!
That would be great!
Thanks

[Per Hansson]

Re: bad flash corrupted bios

You need removable BIOS chips
Hotflashing is done by booting up the working system to DOS
Then the BIOS chip is pulled and the corrupted one put in it's place
Next it is flashed...

[gnappoman]

Re: bad flash corrupted bios

mmh ok, but mine is soldered, the entire thread is about how to self-build this

I just need the ultimate advices from the people which did explore this possibility...

[mkumar12345]

Corrupted Bios: Trying to flash the original using SPI Programmer

New member here. Cannot create a new thread yet but thankfully this thread matches what I am trying to do perfectly. I have a tablet thats bricked thanks to me accidentally flashing the wrong Bios.

I found that you can directly flash the Bios onto the chip using a SPI USB Programmer and Clips. So I bought the following two items:
1) SOIC8 SOP8 Flash Chip IC Test Clips Socket Adapter Programmer BIOS/24/25/93, and
2) 24 25 Series EEPROM Flash BIOS USB Programmer CH341A Module W/ Software & Driver

I have the CH341A software and drivers that came with it. The problem I am having is that when I connect the SPI Programmer to the USB port of the computer I am using to flash, the CH341A software status changes to "Connected" when I haven't even attached the clips to the Bios chip yet.

And when I connect the clips to the Bios chip on the tablet the status changes to "Not Connected" and all the buttons get greyed out.

Does anyone have experience with this low level bios flash to debug if I am doing something wrong?