I know, that title looks odd but I'll explain.
Yep, it's on the PS3 again too ![]()
First off, the PS3 has two main chips, a Cell BE and an RSX chip (dunno if there are more techy descriptions but those'll do for me
)
Now, I thought I had found the problem in that the Cell BE chip on my dodgy PS3 was rapidly heating when applying freezer spray, good I thought, I have the problem sussed.
Had a wee peek aboot online and found someone had lifted their RSX chip off the board, or so they thought.
In actual fact both are big square heatsinks that are then greased to attach to the outer casing and fan (stupid setup if you ask me, no wonder they overheat if there's two heatsinks involved!)
So, the RSX chip wasn't seeming to heat up at all, now I know it's the *heat sink* not getting warm.
My logic is now making me think the Cell BE that I thought was overheating is actually doing exactly what it should and sending it's heat straight to the heat sink whereas the RSX isn't.
In that case, I'd expect the RSX chip would overheat and cause the system shutdown..
Popping this heat sink to get in and regrease or check this is gonna be difficult so I would just like those in the know to either confirm my logic or totally wipe it out.
As an added thought and kinda confiming my line of thought is that the RSX is the Reality synthesiser so I guess thats graphics related, the PS3 lasts slightly longer booting up to nothing or playing a movie than it does if I whack in a game, so I'm thinking the graphics chip is working heavier and overheating faster.
All sensible options and opinions gratefully received ![]()
Jimbo : oÞ
"Making Teenagers depressed is like shooting fish inna barrel"
This item was edited on Tuesday, 9th March 2010, 13:23
Don't own a PS3, didn't read the other thread much. But anyway...
A picture would help! But what you are describing (from someone who hasn't seen it) as a heat sink is the casing that the chip is built into, which really shouldn't come off at all.
Do you have a hairdryer with a cold setting, or a fan heater that operates with it off? They can be useful for blowing on chips to see if it is a heat problem or not.
Editor
DVD REVIEWER
MYREVIEWER.COM
"The Green Belt policy is a Labour Policy, and we intend to Build on it"
John Prescott
Quote:
A picture would help! But what you are describing (from someone who hasn't seen it) as a heat sink is the casing that the chip is built into, which really shouldn't come off at all.Found some earlier for a lad who had pulled his off, s538.photobucket.com/albums/ff344/saltru18/
Possible you are right, it is in fact a casing.
Quote:
Do you have a hairdryer with a cold setting, or a fan heater that operates with it off? They can be useful for blowing on chips to see if it is a heat problem or not.Using freezer spray Mr S, and finding that when powered up and no fan etc connected, the Cell BE chip (the one which hasn't come apart on that lad's pics and seems to be the main CPU etc) is heating up at a helluva rate, fast enough for me to watch the freezer spray melt itself back off as I expected it to.
However, the RSX chip (the one missing the outer in those pics) doesn't appear to be heating up quickly at all.
Jimbo : oÞ
"Making Teenagers depressed is like shooting fish inna barrel"
This item was edited on Tuesday, 9th March 2010, 18:53
So,, providing you haven't removed the casing of the chip, that should be getting heat transferred. In which case the problem definitely is the transference of heat from the top of the chip case to whatever contraption the PS3 uses (some kind of heat pipe/sink combo I'm guessing).
Now one thing I know from bitter experience is the part of the casing that gets hot is the middle, way more than any other part of the casing. And unless that part is flush against a heatsink it won't transfer much heat at all. Paste won't fill in this gap, all paste can possibly do is fill in the microscopic imperfections on each surface to improve efficiency, it won't bridge any gap.
Things that can cause heatsinks to not sit flush off the top of my head... screwing down one side too tightly, not screwing down a side tightly enough. Bad weight distribution (but the PS3 probably doesn't have a huge giant tower fan on it so won't be this).
All of that is probably no help to you though, sorry. ![]()
Editor
DVD REVIEWER
MYREVIEWER.COM
"The Green Belt policy is a Labour Policy, and we intend to Build on it"
John Prescott
OK haven't removed the casing as of yet.
The RSX he casing anyway)
The Cell BE chip is heating up very quickly.
If I cool it with freezer spray it still kicks in as an overheat, which is why I think it isn't the Cell BE but either the RSX or something underneath it's casing.
I've tried two old Skt478 coolers clamped against each chip (granted only clamped with tie wraps so not mega forced) which still overheated, tried spraying in cooling spray, same result.
When I'm spraying it with the freezer spray the fan is removed and the chips (or casings) exposed so I'm definitely hitting the right places.
It just seems odd that one chip is boiling hot, the other isn't and it still overheats without the fan assembly getting warm enough.
The fan, btw, is about 7" wide!
Jimbo : oÞ
"Making Teenagers depressed is like shooting fish inna barrel"
I don't think freezing spray is much use really, unless you are basically holding down the spray from the second you turn it on until it shuts down. Some chips are capable of generating more than enough heat to shut themselves down in under a second without a heatsink on them. Old AMD chips used to fry themselves in less than this time.
You could have a faulty temp sensor in (or under) one of the chips, though I'd be surprised if that was the problem.
Have you got an extremely aggressive fan you could use? Or a blower (like a fan heater that works with the elements off) of some sort?
Editor
DVD REVIEWER
MYREVIEWER.COM
"The Green Belt policy is a Labour Policy, and we intend to Build on it"
John Prescott
Quote:
Or a blower (like a fan heater that works with the elements off) of some sort?I have 2.
I shall try them in the next half hour or so.
Jimbo : oÞ
"Making Teenagers depressed is like shooting fish inna barrel"
Right, tested with fan heater blowing cold, and the chips hit by freeze spray beforehand.
Same results.
Touched both chips (or casings) and the RSX is still very cool with the Cell BE only being slightly warm.
Definitely not what I'd consider a shutdown temperature so I'm back to thinking a different component is either overheating or a faulty thermal sensor somewhere. Problem is I have no idea where the said sensor is ![]()
<edit> Also when in playing with it this time I decided to freeze the chip casings right down, turn it on and try pulsing spray on them, again, no joy, almost instant shutdown and the chip casings were cold.
Not even slightly warm, they were cold.
So, either a diff component overheating or one of these buggers isn't transferring out heat which I'd guess is the RSX as the Cell BE is warm to touch in most cases. <end_edit>
Jimbo : oÞ
"Making Teenagers depressed is like shooting fish inna barrel"
This item was edited on Thursday, 11th March 2010, 16:09
Could it be the PSU? Hard to test without having a spare, unfortunately, though the symptoms might be similar. No domed capacitors anywhere on the power board at all?
Editor
DVD REVIEWER
MYREVIEWER.COM
"The Green Belt policy is a Labour Policy, and we intend to Build on it"
John Prescott
Similar to a PC, completely seperate PSU.
The PSU results in a different error from what I can read online and shuts down without the "system hot" message.
It's getting to be a vendetta now....
Jimbo : oÞ
"Making Teenagers depressed is like shooting fish inna barrel"