Print

[Noilegnave Siseneg Noen]

yeah, that sounds a little more serious than just having a bad flash on your TSOP.  you get any audio or video at all?

what it sounds like to me is your board is toast.  or it might be shorting on a solder splash or something like that, possible to recover under close inspection.  but that's about the only advice I know to give for that, inspect everything on the board to ensure you didn't accidentally short something out.  anybody else?

[Puskas]

i will go over it, but i dont think ive made any misstakes... ive soldered lots of xboxes, both tsop and chips, hdleds, joypowerbutton etc...

this bahavior have been constant after the bad bios flash, this means the xbox worked before flashing... so my soldering shouldnt be the problem...

thanx for any ideas/advises etc....

[Puskas]

Hey guyz, i've gone over all possible and impossible solder points, all looking anything but bad..
I will try and explain step by step what happened:

this on an old 1.0 xbox....

1. soldered tsop jumpers like ive done many times before
2. copied the memcard linux 007 gamesave hack to the xbox
3. booted up 007
4. loaded the 007 savehack
5. telnet to xbox, login and shit
6. runs raincoat with a working 256Kb x2 bios
7. rebootes
8. shocked that it didnt work
9. still chocked but now almost crying...
10. checked forums and tutorials, found out that one must use 1000kb bios on 1.0 xboxes...
11. solder on my trusting old matrix
12. doesnt work...  the tears are soon coming...
13. figured maybe the matrix for some fucked up reason have lost the bios i flashed 1 year before.
14. reflashes matrix and test it on another xbox, works
15. solder it back on the nonworking xbox, doesnt work...
16. cries
17. find out about switches for selecting different banks, nothing works ...
18...... you tell me what to do and I will..

[Noilegnave Siseneg Noen]

have you tried this?

[mranak]

QUOTE (Ant2483 @ Apr 24 2003, 06:17 AM)

Currently most people are following the tutorials on the web site to create a multibios tsop by grounding points A18 & A19. But using this method you cant use the newest multiversion bios and I was unable to get it to work at all with a V1.1 box. A much better way is to make a switch that switches from A19-->ground & A19-->vcc. This forces the box to only read from half the tsop at a time without comparing different parts of the tsop. The easiest way I have found to connect A19 to vcc is to use this point on the lpc: <picture taken out for space reasons, refer to 1st message> I would suggest flashing the bios for the first time without the switch installed first because once installed (like I said before) the xbox can only read or write to half at a time. Ground will be top half, VCC will be bottom half. But after the switch is fully working you can flash half of the bios at a time with evox. Just boot up the box and have a 512k size bios in c:bios and you can then flash half of your tsop at a time. I have tried this on both a v1.0 and a v1.1 and both have worked fine just make sure you use the correct original bios for each. I also see no reason why this wouldnt work with a switch on both A18 and A19. Then you could switch to 4 different bios banks and could flash to each bank at a time.

So now, if I switch A19 between ground and vcc, then my xbox can only see either the beginning 512K or the ending 512K of that BIOS at a time, respectively.

What did the (old) method of grounding A18 and A19 do then?  If I understand correctly, the xbox boots the last 256K of the BIOS normally, but it can still see the entire 1MB.  Does grounding A18/A19 effectively block the first 768K of BIOS from even being seen?

Now, I also understand that using A18+A19+VCC (is the consensus to use 3.3V now instead of 5V?)+ground is good because you can get 4 256K BIOSes out of the mix.  However, I notice that long since this original post, the XECUTOR2 BIOS is now 512K in size as of build #4978.XX (XECUTOR2 BIOS is 512K).  Therefore, would it be best to split the BIOS into something like the following?  If I got the table wrong then please let me know:

CODE

A18         A19         BIOS ----------------------------------------------------------------------------------------------- Ground   Open       Beginning 512K of BIOS, for the XECUTOR2 which needs 512K. VCC        Ground    3rd 256K of BIOS -> Original BIOS (3944 in my case) Ground   Ground    4th 256K of BIOS -> Some other BIOS (Which one?)

The above table is WRONG.  This one should be right:

CODE

A19   A18      0        1        2        3 N        N        X        X        X        X        1024KB, 0+1+2+3 G        N        X        X        o        o        512 KB, 0+1 G        G        X        o        o        o        256 KB, 0 G        V        X        o        o        o        256 KB, 1 V        G        o        o        X        o        256 KB, 2 V        V        o        o        o        X        256 KB, 3 V        N        o        o        X        X        512 KB, 2+3 N        G        X        o        X        o        512 KB, 0+2 N        V        o        X        o        X        512 KB, 1+3

Of course, I'm assuming that this is worth the trouble for a 512K BIOS.  Is the XECUTOR2 a good, feature packed BIOS?  I realize that just because it's bigger doesn't mean it's better.

Update2: Now according to the the post at http://forums.xbox-s...ST&f=43&t=94172 , the XECUTOR2 is 1MB.  Which is it, 512K or 1MB?  I think it's 512KB.

Update3: I read the nfo files for The Phoenix Boot Loader (v1.3) and I realized that I don't need 007 or any other game for that matter to exploit my xbox and flash the TSOP, nor do I need a memory card or action replay.  The only real 'trick' that I needed to do was to get the xbox to unlock the hard drive and connect the hard drive to my PC to update the xbox's files, thus exploiting the xbox through the dash's font exploit.  It's all here: http://www.maxconsol...=xbox&itemid=87.  So now my xbox has te phoenix boot loader on it and is loading a 1MB verison of the X2 4977 from disk.  When I used EvoX's 'backup' function, I get a 1MB file of the 3944 bios, so EvoX is obviously backing up the TSOP.  So I grounded A19 and ran the 'backup' function again to see if I got a 512KB version of 3944 in my C:backup folder.  Nope: still 1MB.

So my question is, is EvoX simply wrapping around the BIOS and thus copying two copies of 512KB of my TSOP into a 1MB file, or did I screw up modding my xbox?  I have wires going to A18, A19, ground, and 3.3V.  I easily verified that the 3.3V and ground connections are good with my multimeter.

You see, I don't want to flash a 512KB bios to my xbox if the entire 1MB of BIOS is still visible, for obvious reasons (this would cause a FRAG-Flash Red And Green-condition, right?).  So I am trying to verify that A19+ground is working.  I presently have A18 and VCC hanging loose with tape over the wire ends.

I really wish to get an answer to this next question.  I've been waiting for a few months now...
Say my mod did work and only 512KB of my TSOP is visible right now.  What happens if I try and flash a 1MB BIOS file to it (a 1MB BIOS file that is actually either 2 or 4 identical 256KB or 512KB BIOSes)?  If it wraps around the BIOS or if it fails to write the final 512KB from the source flash file, I should still be good, right?  Of course, if my mod failed, then the full 1MB of flash gets written to 1MB of TSOP, no problem.

My final question is why bother flashing the TSOP at all?  The Phoenix Boot Loader (PBL)  is loading another BIOS anyway.  I can load an alternate BIOS from disk with PBL v1.3 by powering on the xbox with the tray open.  Finally, I think that next 'official' release of PBL (v1.4) will have USB support so that I can select from a variety of BIOSes upon boot.  Update: Well, the boot process doesn't always seem to work with PBL.  I'm not talking about the clock problem which causes the xbox to keep rebooting itself.  Instead, either the screen goes black after the initial xbox boot process or I see a sort of green wireframe.  What gives?

Thanks

[mranak]

QUOTE (yzo @ Aug 27 2003, 12:02 AM)

A picture of the mod using a surface mount resistor is now up for those interested. http://nidek.sakura.ro/switch_mod/switch_m...es/Image10.html Enjoy. -------------------- Always remember that you are unique. Just like everybody else.

I don't get it; he doesn't have the write enable points jumpered on the top of the motherboard.  Did he do that mod after this picture was taken?  All of this is useless without jumpering those solder points by R7D3 (MCPX RD# to Flash OE#), right?

btw: thanks for the pictures.  Every little bit helps the rest of us understand things better.

[Reefer]

In reply to XBOX-LINUX post:

"It was suggested to connect A18 and A19 resp. either to GND or to Vcc, to have definitive 1s and 0s, but we cannot recommand this either, because it is quite likely that this will make your MCPX explode!"

I've done this to 50+ boxes and no one has ever complained about their xboxes not working anymore. My own xbox uses this method and it works just fine.. I only use A19 to GND/VCC cause 2*512k is enough for me.

Anyone else have any respons to this? Also at the end of the post it says:
"Because of severe compatibility problems and possible hardware risks, the Xbox Linux project suggests that you do not use this method. If you really need this functionality, please get a modchip with multiple banks."

Did they get bribed by some modmaker?

This really seems funny.. I know the a19 or a18 to ground method is dumb. But the method in this post really works. Never ever has caused any trouble on any box i have used it on.

Give more feedback? Did your mcpx explode?!?.. heh..

[jerryclho]

From the mod of surface mount resister http://nidek.sakura....es/Image10.html.
We should cut a trace and add a 4.7 K resister to re-connect that trace. To me, it is a very risky thing to do, even I have some solding experience from Audio DIY.
I am thinking about a easier way to have this assurance. My idea is to add this resister between the A19 and the switch. Will this work? Any comment is appreciated.

         
Gnd---------------------|-|_
A19-4.7K Resister ---| |_| (Switch)
Vcc----------------------|_|

[xboxlinux]

Can I use the Slayer 2.5 to flash my bios with this multibios setup? Acc>>A19<<GROUND setup?

[mufsta]

quick question what program is everyone using to split the original bios

[xbox_shima]

mufsta i used xbtool and resized it
--------------------------------------------------
Ant2483 I LOVE YOU. i flashed my bios (retail+retail+hacked+hacked) and used the a19>ground a18>ground and only got the hacked bios to work. when i tryed your method works. THANKS ALOT. got live working on hacked xbox (when on retail bios+hdd locked)

[derived]

this is what i know:

1) Change the flash size thing to 0x80000 if u have a 1-meg TSOP, because you've effectively halved its size.
<evox.ini>
ie. Flash = 0xadd5,"Hynix - HY29F080",0x80000
not Flash = 0xadd5,"Hynix - HY29F080",0x100000
It might work without this but I think you should set it with whatever software your using.

2) The reason multiBIOS v1.1 + v1.0 doesn't work without switching for both high and low is cause there are incompatible bootloaders, but it isn't very hard to change it. In fact with info from Paul B.'s guide at xb-linux its possible to change multiver BIOS to V1.0 and vice versa, u can do this instead. (kinda need to swap bootloaders using xb-tool, xbflash, or maybe just versions of xbflash (code it for multiver/w/out 2bl decryption, not as hard as it sounds but most ppl can't do it :-(), maybe look at those keys up the top of 2bl as well? - i dunno, i found it trivial enough when i did it, i might have just been lucky i guess).

3) Thin gauge wire does not matter. It will NOT overheat, melt or cause any damage, unluss your box has a problem. Reason being is heat is proportional to current * resistance, and there's bugger-all of either of those.

4) Use 3.3V if your TSOP can still detect that strength (it can). It doesn't really matter, but its a little safer - and cause it doesn't change states frequently you won't risk corruption, so the standard 5V isn't really necessary.

5) The TSOP is not what risks getting damaged (unluss your really stupid installing). Its whats connecting to the TSOP A19 (i personally believe this won't get damaged this way). When it is, for example, its trying to send high when you've grounded it => short circuit. But this short doesn't last long and ppl have been doing this for so long now without problems, it seems it's ok. Holding high while its supposed to be low shouldn't be any different. (sounds dodgy i know)

6) There is no advantage using a "real" modchip since you no longer risk screwing your xbox when flashing, just so long as theres two working non-official BIOSs :-)

[A@ron]

Well this worked for me. I was scared sh*tless because I flashed using raincoat prior to installing the switch and the machine would only frag. Once the switch was installed it was all good!



A@ron

[shank69xo]

Okay guys, I tried this. Everything went great. I now can boot my v1.1 TSOP with X2 4979 and Original 4817. I also used a switch with a middle "off" position. So I can always boot my modded BIOS then flip switch to the "off" and reflash a whole 1MB BIOS, instead of a 512K. I have teasted this. My only concern is al this talk of resistors. I do not have one in the circuit anywhere. Do I need one? If so what kind, and where do I put it in the circuit? Please let me know ASAP. I do not want to kill my XBOX.

[disaster]

I've read the whole thread and just want to double check for my own sanity a minimal set of things I will need after the write points are soldered/jumpered.

a 512Kb BIOS (probably going with X2)
27 or 30ga wire
SPDT switch
soldering iron
007 AUF exploit (already have)

Should I have anything else?

With all those I solder 3.3v -> A19 <- Gnd to the switch
Boot evox choose flash bios and choose the 512K X2 BIOS and let it reboot.

After that I should have retail+retail+X2+X2? No need to flash the other half at all?

Of course I don't even know what version Xbox I have yet so if it's 1.2 I can only have 1 BIOS anyways.