2009 Mac Pro x5677 Upgrade

[HugeHungarian]

So the title says it pretty well. I'll be upgrading my current cMP (flashed 5,1) from the base e5520 2.26Ghz processors to x5677 3.46Ghz. My incredible wife actually got them for me in a 5 part anniversary gift, and they showed up today. Thanks to my ebay search history, for her research!

So this isn't about me asking how to go about this, as I've done a ton of research on this long before knowing that I was getting new processors. I'll be going down the road of the delidding process using a method such as the one on this thread https://forums.macrumors.com/threads/4-1-cpu-upgrade-pre-delidding-test.1997401/#post-23578030

The point of this new thread is to share my experience, and open another opportunity for others to share theirs, if they're so inclined. I will update with my results upon completion. Thanks to anyone that shares.
[doublepost=1508218683][/doublepost]One thing that I did want to ask that I haven't seen in all my research: Can I use a soldering iron for the delidding? Would I get enough heat? Or would I get enough heat quick enough to avoid baking the processor?

 

[raymanster]

I have the same processors which I de-lidded last year using the mini blow torch method with razors to create tension on the lid. I ended up scrapping a processor as I cut too far with the razor.

If I had to do it again I would try the vice method.

 

[William_si]

I use non delided in a 4,1 chassis with 4,1 dual tray, with common sense and experience with socket B/1366 torgue this is no issue and de-liding seems more risk than worth it.

Just last week i changed again to 2 non delid E5620, first try no boot, minimal adjustment on CPU2, boots fine.

 

[h9826790]

I use non delided in a 4,1 chassis with 4,1 dual tray, with common sense and experience with socket B/1366 torgue this is no issue and de-liding seems more risk than worth it.

Just last week i changed again to 2 non delid E5620, first try no boot, minimal adjustment on CPU2, boots fine.

Delid - more risk on the CPU

Lid - more risk on the tray

IMO, if the CPU is relatively cheaper (or easier to accuire), than delid make sense.

But if the tray itself is cheaper, then washer method make more sense.

In both cases, common sense and carefullness are required to avoid damage.

 

[William_si]

The washer method is safe (as safe as delid in practice), i use none at all.

 

[Squuiid]

If you're going to delid, PLEASE use the vice method. It is so much safer, and easier.
Sep 16, 2017

 

[itdk92]

The washer method is safe (as safe as delid in practice), i use none at all.

For the OP - Remember the extra thermal pads though, or some chips on the board may overheat and get damaged
[doublepost=1508245278][/doublepost]

If you're going to delid, PLEASE use the vice method. It is so much safer, and easier.
Sep 16, 2017

Quote. Definitely safer.

 

[HugeHungarian]

If you're going to delid, PLEASE use the vice method. It is so much safer, and easier.
Sep 16, 2017

Those were my (inexperienced) thoughts exactly, until seeing this: https://forums.macrumors.com/threads/4-1-cpu-upgrade-pre-delidding-test.1997401/#post-23524417

I guess my thought now (still inexperienced) is if I manage to slightly hit a cap, it's possible to salvage; if I rip the whole thing down to the magical rainbow, then I'm hosed for sure. In the end I'm sure it's a numbers game, and the failure rate of vice method is likely lower based on what everyone is saying, just catastrophic when it does fail.

Air plane vs car crash.

 

[Squuiid]

Those were my (inexperienced) thoughts exactly, until seeing this: https://forums.macrumors.com/threads/4-1-cpu-upgrade-pre-delidding-test.1997401/#post-23524417

I guess my thought now (still inexperienced) is if I manage to slightly hit a cap, it's possible to salvage; if I rip the whole thing down to the magical rainbow, then I'm hosed for sure. In the end I'm sure it's a numbers game, and the failure rate of vice method is likely lower based on what everyone is saying, just catastrophic when it does fail.

Air plane vs car crash.

LOL, nicely put. I did 5 CPUS in a vice with no failures. My two cents.

 

[MIKX]

As one who is soon to enter the 4,1 - 5,2 CPU upgrade Twihlight Zone .. and . .after endless Googling this idea came to me in a realistic dream ( no kidding ! ) ..

I purchased a batch of old intel CPUs and using the below method the lids just fell off with so little VICE PRESSURE. Also, the residual solder on top of the cpu ended up uniformly spread out = much easier to remove after the lid is removed.

= no razors.

VICE . . . plus . .. . GENTLE HEATING at the same time to help the vice ?

Surely this will speed up the process.

Thoughts ?

 

[Razzerman]

this idea came to me in a realistic dream ( no kidding ! )

Wow, someone has spent a lot of time researching! Good idea though, softening up the chip prior to the vice.

Another idea, I bought an old, cheap quad core chip with a lid on to practice on if I was going to go the delid route. Went the other way in the end, with washer and all.

Good luck.

 

[HugeHungarian]

As one who is soon to enter the 4,1 - 5,2 CPU upgrade Twihlight Zone .. and . .after endless Googling this idea came to me in a realistic dream ( no kidding ! ) ..

I purchased a batch of old intel CPUs and using the below method the lids just fell off with so little VICE PRESSURE. Also, the residual solder on top of the cpu ended up uniformly spread out = much easier to remove after the lid is removed.

= no razors.

VICE . . . plus . .. . GENTLE HEATING at the same time to help the vice ?

Surely this will speed up the process.

Thoughts ?

I did actually wonder if anyone had tried this; a slight warming to soften the solder, and lessen the chance of internal decapitation. I mean it makes sense. I just might give it a try. I'll be going out to take my son to the dentist soon, and I'll swing through Harbor Freight to see what they have for cheap vices. Maybe some light soldering iron heat while in the vice will make it less violent/risky.

 

[itdk92]

I did actually wonder if anyone had tried this; a slight warming to soften the solder, and lessen the chance of internal decapitation. I mean it makes sense. I just might give it a try. I'll be going out to take my son to the dentist soon, and I'll swing through Harbor Freight to see what they have for cheap vices. Maybe some light soldering iron heat while in the vice will make it less violent/risky.

The solder used in those Xeon family is very very “buttery” and extremely easy.

Never damaged one single CPU with the vice. Heard of people damaging the CPUs by scraping off the remaining soldering though.

 

[HugeHungarian]

The solder used in those Xeon family is very very “buttery” and extremely easy.

Never damaged one single CPU with the vice. Heard of people damaging the CPUs by scraping off the remaining soldering though.

So regarding the vice method, as I've mainly researched the blade and heat method, is there a preferred direction of pressure on the lid? As in, does it matter which side I apply the pressure to so as to avoid damaging caps, or chipping the...chip?

I'm heading to the search bar, as soon as I click 'Post'. Just looking for new info, since this method doesn't seem to be as popular, even if it is better.

 

[ActionableMango]

When I used the vice method I didn't apply any heat. I didn't take much pressure on the vice and the lids popped right off. Worked fine three times. My CPUs simply fell down into a cloth I had laying below it.

The solder is so soft that people have reported you can use a credit card to scrape it off. I used a window scraper.

Just be aware that every single method reported so far has had at least one person end up ruining something, so there is some risk to all of them. I used to think the lidded+washer method was safe, but someone here had problems with that too.

 

[HugeHungarian]

Just be aware that every single method reported so far has had at least one person end up ruining something, so there is some risk to all of them. I used to think the lidded+washer method was safe, but someone here had problems with that too.

True. I guess I have to remember the human factor as well. Not that there's no inherit risk, but some people just aren't as delicate or thorough as others. Not that they're stupid, just maybe a little less careful than others would be. Seems to be the case with many reported failures, in various situations in life.

 

[ActionableMango]

So regarding the vice method, as I've mainly researched the blade and heat method, is there a preferred direction of pressure on the lid? As in, does it matter which side I apply the pressure to so as to avoid damaging caps, or chipping the...chip?

Yes there is a preferred direction. This is the video I followed. He identifies the direction and provides an explanation of why he picked that direction.

Note in his video the lid is on so tight that his arms are shaking, and when the lid finally pops, the CPU goes flying. This was not my experience at all. With the three* processors I delidded, they just sort of fell off. It did not require much pressure at all.

* I did one cheap CPU beforehand for practice...my 2010's original 2.4GHz quad core.

 

[Squuiid]

Note in his video the lid is on so tight that his arms are shaking, and when the lid finally pops, the CPU goes flying. This was not my experience at all..

This was my experience with 4 out of 5. They required quite a lot of pressure to pop.
One did go pretty easily however.
Regardless, none were damaged. Just keep crankin' that vice until they go.

 

[HugeHungarian]

Ok people, so I went into Harbor Freight and found a $16 drill press vice. This was preferable because it is flat on the bottom and easily sits on a desk.

This process went so incredibly easily that I couldn't imagine doing it any differently. I also wonder if some of the extreme tension that some experience is from possibly too much angle in the vice. I tried to keep it as flat as possible in the vice, and the edge that was engaged against the IHS was on the upper face of the IHS and NOT the lower, closest to the chip. I feel like this provides a nice even pressure that doesn't dig into the chip.

You can hardly see the release of tension in the video. When I place my hand over the vice, I have started cranking tension. The most difficult part was keeping the vice flat while cranking, as it was not bolted. Gives the illusion that I had a lot of tension. I did not. Just about as soon as I felt the tension load up, another partial turn and felt everything easily release. As soon as this happened I pulled the IHS off manually, as I didn't want to push it into the caps with the vice. Both lids came of with the same ease.

I will update again when I go to boot. Much thanks to itdk92, Squuiid and the many others that helped me decide on this route, against my original plan!

 

[ActionableMango]

Arrgh, I'm feeling anxiety about all that touching of the CPU contacts.

Looks like the lids popped off nice and easy though.

 

[HugeHungarian]

Arrgh, I'm feeling anxiety about all that touching of the CPU contacts.

Looks like the lids popped off nice and easy though.

Haha, I used to be super worried about it too. Ever since working with this local computer shop that's about as careful with CPUs as he is with pocket change (not that I agree), I've loosened up a bit.

I sure the oils from skin just helps to prolong the life of the metal, haha. I'll make sure it's clean

 

[HugeHungarian]

Well I'm posting this from my cMP with both CPUs installed and showing, but only 12 gb of ram showing. Got the bong quick, and booted even quicker than the old setup. So would you guys try tightening the heat sinks? I have them snug-ish but I don't think quite as tight as the old ones.

UPDATE: Tightened even more, and now 20 of 24 gb of ram showing. Guess I'll try to go tighter...

 

[HugeHungarian]

Ok ended up trying to reset PRAM/NVRAM ( I see them used interchangeably, but I'm sure one is PPC and One is Intel), and SMC. This corrupted something in the EFI boot for Windows/Bootcamp. Blue screen error, and F8 for Startup Options and Enter to reboot. These keys do nothing but flash the screen. I ended up restarting, holding Option and manually booting OS X.

Upon getting in there, I still wasn't reading all RAM, and red light on board. Took out board, swapped CPUs, cleaned all contacts (again) on CPUs and RAM, blew out board with air, reinstalled and all hardware is working. No red light, and all RAM showing. So the transplant is a success. Now just need to fix the boot issue with Windows.

Thanks to everyone! Even with the issues, I am happy with the swap. Much faster machine! Hope this helps someone else looking to do this kind of upgrade.

 

[HugeHungarian]

So this will boot to OS X with zero issues. If I try to boot to Windows 10/Bootcamp using Bootcamp app, I get this:

Must hold power button from here.

If I hold Option key during boot and manually select the first EFI partition, which is on the separate Windows physical drive, it boots Windows with zero issue. See boot options here:

If I choose the second EFI partition, which seems to be located on the Mac hdd, as it disappears when I remove that drive; then I get this screen, and must shut down to get out:

I assume this is the partition I need for Windows to work correctly. Also, before all of this happened, I only ever saw one EFI partition on the boot select screen; now there's two as described.

Anyone have any ideas, or a direction on the interwebs to point me?

 

[thornslack]

Interesting. Be following this for your win 10 boot solution. It is similar to a problem I am having