Some Issues with Dual CPU Upgrade 2009 Mac Pro

Discussion in 'Mac Pro' started by KBS756, Jan 3, 2013.

  1. macrumors 6502a

    Joined:
    Jan 27, 2009
    #1
    Installed one CPU and got the Comp to boot up but once the computer logged in the heat sink on CPU A went nuts. (Max RPM Fan). This is without the end Heat sink or cpu installed.

    Then I went ahead and installed the 2nd CPU and I do not get any Bong or boot. Fans don't spin up violently. doesnt boot up.

    Can someone please help me remedy the situation.

    Also I havent modified the fan connectors could that be what is causing the issue? I am using 2 mm worth of washers on all 8 posts

    thanks
     
  2. macrumors 6502a

    Joined:
    Nov 25, 2012
    #2
    First, you got to stop double posting since I just replied your earlier post in another thread. Now that you have added additional info, and I can tell you for sure that it is the fan connector seating problem since the same situation happened to me. All I had to do was using 2 washers instead of 3 on each post, and all problems solved. I did not clip the fan connector either although that would be my next move if I still could not get it to work. Luckily 2 washers on each post (side) and proper tightening were all I needed to make it work. Hope this helps in your case as well.
     
  3. KBS756, Jan 3, 2013
    Last edited: Jan 3, 2013

    thread starter macrumors 6502a

    Joined:
    Jan 27, 2009
    #3
    Sorry about that I somewhat panicked in regards to that.

    Got it to Bong with 2 CPUs in now

    but when I log in only 1 CPU registers as well as only 1 CPU's ram

    iStat Pro Can read CPU Temp as well as the Speed of the Fan over the CPU

    Both Fans seem to register in iStat Pro as runnin at about 1100 RPM (BOOST A and BOOST B)

    But I either get only 1 CPU to boot and register or when I try to tighten or CPU B I dont get the bong
     
  4. xcodeSyn, Jan 3, 2013
    Last edited: Jan 3, 2013

    macrumors 6502a

    Joined:
    Nov 25, 2012
    #4
    Looks like you got uneven tightening on one CPU, therefore the ram not recognized. Loosen the screws and carefully tighten them in one full turn each in diagonal order until they are not too tight not too loose. You know the drill. Another CPU seems still having connector seating problem, it may also indicate that the screws are not tight enough therefore no complete contact with the motherboard connector.

    Edit: You may need to loosen screws and re-tighten them for both CPUs. Remember that uneven tightening causes ram not recognized. CPU not recognized is probably due to improper seating. Both needs fine tuning.
     
  5. thread starter macrumors 6502a

    Joined:
    Jan 27, 2009
    #5
    Can I assume CPU A is fine since it will boot? And the fan on it works

    I have reseated B a couple times now and still dont seem to be having any luck gettin a Boing from when both CPU's are down...

    I see that CPU B's fan does turn though ...
     
  6. macrumors 6502a

    Joined:
    Nov 25, 2012
    #6
    Sorry I had to step away for a while. I think your CPU B may be defective. The only way to confirm is to test it either on another machine or switch it with the other one. You mentioned earlier that both boost A and B fans were running at normal speed, which suggests that both fan connectors are properly connected. This is always a problem when getting new CPUs to work. I have another single-core MP so I could test mine before upgrading. In your case, you may have to switch them to see if both are OK. Good luck.
     
  7. thread starter macrumors 6502a

    Joined:
    Jan 27, 2009
    #7
    K thanks ... think I'm gonna leave it for tonight and bother again tomorrow.

    Would there be any risk running my Pro with 1 CPU posting and the other not as long as the fans in both heatsinks are spinning at a normal speed
     
  8. macrumors 6502a

    Joined:
    Nov 25, 2012
    #8
    As long as the fans are running at normal speed and temperatures are in normal range, I don't see a problem. I'm a bit surprised that you could still post with a defective (unrecognized?) CPU. When you take CPU B off the socket later, make sure to check the CPU socket for any possible bent pins in the socket. Maybe just a few bad connections instead of a bad CPU. Good luck on your testing, and keep us posted.
     
  9. KBS756, Jan 3, 2013
    Last edited: Jan 3, 2013

    thread starter macrumors 6502a

    Joined:
    Jan 27, 2009
    #9
    Reseated the CPU several times didnt see anything obviously bent in terms of pins. Would there be an easy way of checking all the pins. Going to try to make this work one more time tomorrow before I ask the seller to replace the CPU

    And in terms of posting... When I seem to undertighten the CPU so that the Fans on the heatsink work but the CPU isnt recognized I get the boing and only CPU A and it's ram recognized. When I tighten it to a comparable level to CPU A and within that range the comp turns on the fans go to normal start up speed and then just stay there no video grey screen no boing. And if I tighten what I believe is too hard a red light behind the CPU turns on as soon as I turn it on and no boing.
     
  10. macrumors 6502a

    Joined:
    Nov 25, 2012
    #10
    Basically you just take a closer look from different angles about the alignment of those pins in rows and columns. If anything usual pops out, you should find it more easily with a magnifying glass. I did find one particular pin was a little bent in my CPU B socket for some reason, I used a safety pin, with the help of a magnifying glass, to push it back in alignment with other pins. Not sure if it really mattered, but I did manage to make my upgrade work afterwards. I think you already know how much tightening is required for each CPU to work properly, so just make sure both CPUs are in good condition.
     
  11. KBS756, Jan 3, 2013
    Last edited: Jan 4, 2013

    thread starter macrumors 6502a

    Joined:
    Jan 27, 2009
    #11
    K So I got over my laziness and decided to Swap CPU A and CPU B and see what happens. Counted 3.5 turns to get the heat sink down right. Fans still spin on heat sink when I start up now I get the boing, both heatsinks' fans spin up like expected.

    But I get this when booting

    [​IMG]

    and then it restarts the boot up process and I get this screen
    Says something with kernel and goes before I can really read it.

    [​IMG]

    before repeating in an endless loop. Some help trouble shooting would be great.

    Thanks in advance apprciate any help I can get

    Got a ledgible shot of the screen

    [​IMG]

    Btw how do I make these images thumbnails on the forum?
     
  12. KBS756, Jan 4, 2013
    Last edited: Jan 4, 2013

    thread starter macrumors 6502a

    Joined:
    Jan 27, 2009
    #12
    Can anyone please help me out with this ?

    Thx again

    Here's a shot including the text at the top of the screen

    [​IMG]
     
  13. macrumors 68040

    666sheep

    Joined:
    Dec 7, 2009
    Location:
    Poland
    #13
    I assume thet you've done firmware update to 5,1, otherwise your MP shouldn't even POST with hex CPUs.

    Check this out first and try: http://forums.macrumors.com/showpost.php?p=11692308&postcount=264

    Try also diagnostic LEDs on logic board. It might be also worth to try this "not booting" CPU alone in socket A. If it won't change anything, that means this CPU is damaged.

    Use [​IMG]
     
  14. thread starter macrumors 6502a

    Joined:
    Jan 27, 2009
    #14
     
  15. xcodeSyn, Jan 4, 2013
    Last edited: Jan 4, 2013

    macrumors 6502a

    Joined:
    Nov 25, 2012
    #15
    If the weird screenshots happened when you swapped CPU A and B, then they seemed to confirm that you have a defective CPU on hand and it did not work in both CPU sockets. The chance of getting a bad one usually is very low. Install the good one back to CPU A socket and make sure it still works properly, then return the bad one and wait for the replacement. Maybe go try your luck with the lottery while waiting.
     
  16. thread starter macrumors 6502a

    Joined:
    Jan 27, 2009
    #16
    Yea the weird screenshots happened with Problem CPU in Socket A

    Reverted to my older CPU's for now having no issues whatsoever so guess that basically confirms it. Hopefully replacements will arrive quickly and not have any such issues.

    Thanks again for all the help youve been
     
  17. thread starter macrumors 6502a

    Joined:
    Jan 27, 2009
    #17
    Hey everyone if I could get some help on this it would be great.

    Got replacement CPU's yesterday installed them and all works except the 2nd Memory Channel attempting to loosen and tighten not really having a great effect on the issue.

    Could someone give me some pointers on how to get this resolved....

    Also is it dangerous to run the computer while one memory Channel isnt being recognized and the red light is lit on the processor board

    [​IMG]

    Thanks in advance everyone.

    Boots up fine just doesnt recognize this 2nd Channel of memory so it sees 14GB vs 16 GB
     
  18. macrumors 6502a

    Joined:
    Nov 25, 2012
    #18
    This is due to uneven pressure on the processor. So you need to make sure the same pressure of the four screws. It also happened to me in my upgrade. Just loosen the screws and tighten them in diagonal order until the same tightness level is achieved.
     
  19. macrumors 68040

    666sheep

    Joined:
    Dec 7, 2009
    Location:
    Poland
    #19
    Check link in post #13. There's an easy way to fix it.
     
  20. thread starter macrumors 6502a

    Joined:
    Jan 27, 2009
    #20
    I tried doing this several times didnt work for me. But then got the courage to tighten it past where I felt it was safe by 1/8th turns and finally got it recognized. Was a bit afraid of over tightening it though even though I had the 3 washers on each post.

    Can someone tell me what temperatures I should be expecting at idle and under load so I can be confident everythings working correctly. Thanks
     
  21. macrumors 6502a

    Joined:
    Aug 15, 2007
    Location:
    Austin, Texas
    #21
    AHA! I'm having the exact same behavior! So, in the end, CPU B wasn't tight enough to be recognized?
     

Share This Page