Print

[scaremy]

hi new here

i haven't gotten the 3RLOD but i cant play for 5 min without the screen freezing and looking all weird
so should i give this a shot?

also why not also put a small block "shim" under the GPU as was mentioned before as standard practice? surely that would keep it from bowing in either direction and make it  nice and flat?

  one last thing im confused on

witch is more important?
the GPU side or the CPU side or both? can i get away with just one for now?
also is the GPU hs the short silver one?
and if so why is there 2 processors on the copper hs side? a big one and a little one?






im so sorry at my n00bieness LOL

[compuguy01]

[quote name='scaremy' date='May 1 2007, 01:22 AM' post='3976822'
witch is more important?
the GPU side or the CPU side or both? can i get away with just one for now?
also is the GPU hs the short silver one?
[/quote]
 

Both have to be perfectly flat, Ive been having great success following RBJ's tut not perfect yet but im well on my way its all in the adjustments.   The GPU is under your DVD Drive "the short one" (IMG:style_emoticons/default/smile.gif) .

This post has been edited by compuguy01: May 1 2007, 01:58 AM

[Boddah]

Just finished putting mine back together tonight...seems to be working perfectly after a couple of hours of gaming. I'm not sure of the error code, though. (no error LEDs)

I'll post in the success thread after a few more days.

PS, thanks for the tut RBJtech!

[SnufftheCrimeDog]

Well this seems to be successful fix. I know, my xbox has been working for about a month now no problems. My friends xbox who i had a bear of a time with is also working great now. On the other hand the first xbox 360 we fixed, using Lawdawgs method failed a few weeks ago. Of course we got it back up and running with RBJTechs. So kudos to you RBJTech. Not to totally bash Lawdawgs method in theory there the same fix, but I think that because of the lack of the lock-nut the screws actually loosen up overtime and cause graphical freezing, because of it. I would really be interested in starting a poll, I know there is already one for successful fixes, but a poll to see how long people have been running with this fix, and too see if there still running.

[booker]

QUOTE(SnufftheCrimeDog @ May 2 2007, 01:58 AM)

Well this seems to be successful fix. I know, my xbox has been working for about a month now no problems. My friends xbox who i had a bear of a time with is also working great now. On the other hand the first xbox 360 we fixed, using Lawdawgs method failed a few weeks ago. Of course we got it back up and running with RBJTechs. So kudos to you RBJTech. Not to totally bash Lawdawgs method in theory there the same fix, but I think that because of the lack of the lock-nut the screws actually loosen up overtime and cause graphical freezing, because of it. I would really be interested in starting a poll, I know there is already one for successful fixes, but a poll to see how long people have been running with this fix, and too see if there still running.

Excelent idea...

[RBJTech]

QUOTE(SnufftheCrimeDog @ May 2 2007, 01:58 AM)

Well this seems to be successful fix. I know, my xbox has been working for about a month now no problems. My friends xbox who i had a bear of a time with is also working great now. On the other hand the first xbox 360 we fixed, using Lawdawgs method failed a few weeks ago. Of course we got it back up and running with RBJTechs. So kudos to you RBJTech. Not to totally bash Lawdawgs method in theory there the same fix, but I think that because of the lack of the lock-nut the screws actually loosen up overtime and cause graphical freezing, because of it. I would really be interested in starting a poll, I know there is already one for successful fixes, but a poll to see how long people have been running with this fix, and too see if there still running.

Have to agree.  A LOT of people seem to think you need to do the overheat first - you do not.  If you do any form of overheat, then you are likely fixing the problem with a warped board and it will fail again in time.

You should not get the 3ROL after doing the X-Clamp fix - if you do, then you have not fixed the problem - simple as that.

If you do then the options are : 1) Less washers 2) More torque 3) Less torque or 4) Shim

PS - The X-Clamp+Shim mod I posted a while back is going strong, no issues so far but I'll give it a bit longer b4 I add to the poll ..

[RBJTech]

I've added this diagram to the tutorial which shows you what difference the number of washers makes to the overall result :-



The diagram is obviously over emphasised, as the board flexing no where near that amount graphically ( ) but you get the idea.

You can clearly see WHY it works - pressure is put on the outer BGU connections.  The less washers you have, the more pressure you put on it.  Remember the X clamp did the opposite, it put pressure on the inner pins leaving the outer pins to come away from the flexing board (which was not held firm).

I'm just trying to figure out why the 'shim' method worked (3 washers+shim) when the 2 washer method didn't - looking at it now, putting a 1mm thick shim on the die, according to the above diagrams would just make the board flex into a 2 washer design.. Hmm

[brywalker]

I really hate to disagree with you, but I have to.

The amount of flex that the board gets with less washers is insignificant. The pressure is now equal on all four corners, so you are acutally not bending the board - moreso just pulling it foward. If there was any flex on the board it would be outside the BGU package. The heatsink flexes far before the board, actually. I don't have pics - but I put a GPU heatsink on a throw-away board just to see how far I could torque down before I cracked the die.

I was able to screw it down all the way. The corners of the heatsink touched the board - and the entire bowing was at the heatsink! I don't recommend doing this - but it shows that the aluminum will bend far before the board bends out of spec. The CPU heatsink probably won't flex as much, you would crack the die before you got that far down because it is thicker - even though copper is super soft.

This post has been edited by brywalker: May 2 2007, 02:30 PM

[Bravoexo]

Hey, I didn't realize I was following RBJtech's method instead of lawdawg...

(IMG:http://img256.imageshack.us/img256/3351/xbox1uz5.jpg)

I ofcourse decided to move away a bit to the designed method and think about using that washer instead.  In this configuration.

(IMG:http://img186.imageshack.us/img186/3135/mg6703jm3.jpg)

This way, I didn't have to drill through the HS's or use locking nuts.  Just tightened it in a crosswise manner (since were' talking screws here, tightening one post first all way would tilt the HS to that corner disapparingly)

What do you guys think?  The nylon washer seems to be the right height of the original portion of the oem HS leg that was underneath the mobo (held by the x-clamps)

[RBJTech]

QUOTE(brywalker @ May 2 2007, 02:28 PM)

I really hate to disagree with you, but I have to.

The amount of flex that the board gets with less washers is insignificant. The pressure is now equal on all four corners, so you are acutally not bending the board - moreso just pulling it foward. If there was any flex on the board it would be outside the BGU package. The heatsink flexes far before the board, actually. I don't have pics - but I put a GPU heatsink on a throw-away board just to see how far I could torque down before I cracked the die.

I was able to screw it down all the way. The corners of the heatsink touched the board - and the entire bowing was at the heatsink! I don't recommend doing this - but it shows that the aluminum will bend far before the board bends out of spec. The CPU heatsink probably won't flex as much, you would crack the die before you got that far down because it is thicker - even though copper is super soft.

 

I'm not with you on the heatsink bending I'm afraid ..  Are you saying a 14mm thick vaned aluminium bends more than a 2mm thick piece of fibreglass ?   The heatsink will obviously be 'weaker' going against the vanes, but that is still a 2-3mm thick piece of ali, that will take some force to bend over the short distance (60mm) we are playing with ... ?

Yes the bending is outside the BGU package, (that's the whole idea) - to press the outer pads back onto the BGU package.  If we had enough pressure to bend the package, it would snap (it's ceramic)..

We are probably applying negative (pulling) pressure on the centre pins now by doing this, which is why if possible, I recommend 3 washers to keep everything level as possible.










 

QUOTE(Bravoexo @ May 2 2007, 02:51 PM)

Hey, I didn't realize I was following RBJtech's method instead of lawdawg...

I ofcourse decided to move away a bit to the designed method and think about using that washer instead.  In this configuration.

This way, I didn't have to drill through the HS's or use locking nuts.  Just tightened it in a crosswise manner (since were' talking screws here, tightening one post first all way would tilt the HS to that corner disapparingly)

What do you guys think?  The nylon washer seems to be the right height of the original portion of the oem HS leg that was underneath the mobo (held by the x-clamps)

It's a hybrid of the two

Yes using M5's means you don't have to muck about with the heatsinks but it also means that you apply all the force by using the thread on the bolt, not uniform and flexible pressure from the spring washer - but it works ..

The nylon washers look fine, I assume you are adding extra washers onto the bolts when the board is in prior to attaching the heatsinks ?  (ps - I bet that's fun and games as the bolts arn't secured to the case !)

[Bravoexo]

Yeah, I think it is a hybrid of both methods, he he...

At first I added the same washer on top of the mobo... but no go, no thermal grease actually touched both HS... he he, instant kill switch (+ 2 L overheat signal) from my box. (I actually see why now that I look at your pic...)

Removed any washers and it's just a direct bolt on.  Screwed in the HS in a crosswise manner... and everything is working fine for about 3 days now.  Did even a marathon of 25 games in one sitting (playing the intro, a few minutes in, or downloading updates) and worked a-ok.

I think my washers actually would work with the original HS legs...too.  But can't try it anymore coz I already drilled through my metal case.  Oh well.

This post has been edited by Bravoexo: May 2 2007, 03:30 PM

[ptr.exe]

I've done this method on two 360s now.

The first one I used two washers, as three was too high. This left a small amount of room, I tightened the bolts as far as possible by hand (normal ones, as I had no lock nuts at the time) and then did a half turn using pliers.

The second one I ran out of washers and so couldnt put any. I was careful how much I tightened the lock nuts, so that the GPU heatsink ran parallel to the board, so no flexing occured. And there was very little pressure on the CPU heatsink, letting the spring washers do the work - again so there would be no flexing of the board.

The first time, it fixed the 360s red lights, which I know were caused by a bent board.

The second time it did not fix the red lights (intermitent 0020) as I knew this was because of the board being warped near the AV area, I did the mod just as a precaution against future flexing. However once I put it all back together the 3 ROL never happened again...

I can only guess that the heatsink bolts are now holding the board straight in line with the stanard 360 case bolts that screw into the AV connector. Quite random that this should help such a problem, but thanks all the same  

[booker]

QUOTE(RBJTech @ May 2 2007, 01:01 PM)

I've added this diagram to the tutorial which shows you what difference the number of washers makes to the overall result :-



The diagram is obviously over emphasised, as the board flexing no where near that amount graphically ( ) but you get the idea.

You can clearly see WHY it works - pressure is put on the outer BGU connections.  The less washers you have, the more pressure you put on it.  Remember the X clamp did the opposite, it put pressure on the inner pins leaving the outer pins to come away from the flexing board (which was not held firm).

I'm just trying to figure out why the 'shim' method worked (3 washers+shim) when the 2 washer method didn't - looking at it now, putting a 1mm thick shim on the die, according to the above diagrams would just make the board flex into a 2 washer design.. Hmm

What is the more efficient method?

more flexing, less flexing, perfectly flat?..

thanks

[RBJTech]

QUOTE(booker @ May 2 2007, 04:20 PM)

What is the more efficient method?

more flexing, less flexing, perfectly flat?..

thanks

Ideally as flat as the 'original' - but this won't always fix the problem - so start with 3 and work down to 1...

[booker]

QUOTE(RBJTech @ May 2 2007, 07:57 PM)

Ideally as flat as the 'original' - but this won't always fix the problem - so start with 3 and work down to 1...

I´m not using this method, but stil the concept is the same, so i´ll ask you what you think about it.
Right now i have 2washers, each washer have 1mm each.
So the space between the Mobo and the HS is 2mm (obviously).

The question is... i had the system working flawless for 1 week, but then it started freezing after 2 or 4 min into the game (dashwork was fine though)

You think that maybe 2mm is too much and i should try 1 washer only?
I believe this looks a bit agressive, since it could damage de DIE.. i believe the space between the Mobo and the HS original is around 2mm.. but you are the expert here.
What do you think?

Using 1 washer, you don´t have a chance of breaking the cpu?... it looks a very thin space there.