Resolved Mac Pro (Mid-2012) - no video and a very weird mainboard/processor board issue

[tsialex]

By "push pins", I assume you are referring to the pins that hold down the Northbridge heatsink?

Yes, people frequently damage SMD components on the underside of the board near the location of the northbridge.

 

[Valdaquendë]

tsialex - Well, it took awhile but I have had time to thoroughly examine the CPU board. I spent about 30 minutes examining the entire board looking for cratered, missing or burnt-looking components. I could find none. The Northbridge cooler hold-downs look good.

The only things worthy of note (and maybe they aren't noteworthy at all) are the presence of what I take to be flux residue on two areas of the board (both adjacent to the CPU sockets) and an accumulation of dust on that residue. I've seen this type of residue before.

Unless you alert me to something I may be missing, I'll start shopping for a replacement board. I hope to get to flashing the BootROM tonight.

Here are some photos:

 

[tsialex]

This mess is the oily residue from the degraded thermal pads. This board probably spent most of it's life working at full TDP to have so much residue near the CPUs.

This board is too dirty for a visual inspection, I'd clean this with IPA and some solvent (here we have a local brand of paint thinner that is just PERFECT for cleaning that without damaging the solder mask) and reinspect.

 

[Valdaquendë]

I don't have the time to do the cleaning today; Mondays are generally filled with clients' emergency calls from over the weekend and this Monday is no exception. I'm assuming that 91% IPA will be sufficient; is there anything else you recommend as a cleaner? I realize that any use of IPA presents a risk of moisture remaining under SMCs and thought that, once I have the board clean and nominally dry, I would heat my oven to 150˚F, turn it off and place the board (on a silicone mat) in to dry. Sound OK? If I use thinner, of course, I would not dry in the oven.

Also, I find myself wondering: I assume I'm looking for missing, burnt or cratered components. If I DON'T find any, is there any chance of rendering this board operable?

On the other hand, if DO find them, what then? I just got a very nice Hakko soldering station and used it to replace a drone's flight controller board for a client; I'm willing (even eager) to learn but at present my skills are somewhat limited. Once having identified any offending components, is there any practical prospect of saving this board?

I'm just trying to get an idea as to where we are headed, either way, with this aspect of the project. Any words of wisdom are appreciated and carefully considered. I'll post back, probably tomorrow, with an update as to board cleaning. Should I send photos?

Thanks, tsialex!!

 

[tsialex]

I don't have the time to do the cleaning today; Mondays are generally filled with clients' emergency calls from over the weekend and this Monday is no exception. I'm assuming that 91% IPA will be sufficient; is there anything else you recommend as a cleaner?

IPA won't remove all the oily residue, try kerosene/paint thinner on a small area of the PCB that have no solder mask and see what works best.

I tried lot's of chemicals over the years, the paint thinner I have been using is not the best product but it's good enough and it's not a controlled substance like acetone.

I realize that any use of IPA presents a risk of moisture remaining under SMCs and thought that, once I have the board clean and nominally dry, I would heat my oven to 150˚F, turn it off and place the board (on a silicone mat) in to dry. Sound OK? If I use thinner, of course, I would not dry in the oven.

Remove all the oily residue, clean the whole board with a brush and neutral detergent, let it dry for 24h then you put it on the oven.

I do this all the time when refurbishing Mac Pros, since my ultrasonic cleaner is not big enough for the CPU tray/backplane.

Also, I find myself wondering: I assume I'm looking for missing, burnt or cratered components. If I DON'T find any, is there any chance of rendering this board operable?

On the other hand, if DO find them, what then? I just got a very nice Hakko soldering station and used it to replace a drone's flight controller board for a client; I'm willing (even eager) to learn but at present my skills are somewhat limited. Once having identified any offending components, is there any practical prospect of saving this board?

I'm just trying to get an idea as to where we are headed, either way, with this aspect of the project. Any words of wisdom are appreciated and carefully considered. I'll post back, probably tomorrow, with an update as to board cleaning. Should I send photos?

Thanks, tsialex!!

It's a too expensive board to not inspect it fully, look for PCB scratches, see if no tracks are damaged, no SMD components are ripped and etc.

If you find anything, then we can think about what's the best way to repair. If you don't find anything, then it's time to get a new board knowing that you tried your best.

 

[Valdaquendë]

I will try to get this done today.

Would I be correct in thinking that by "neutral detergent" you mean an unscented dishwashing liquid or something of the sort?

I actually have a can of acetone; I got it for a rather delicate cleaning job but I've never used it on a PCB, fearing that it could dissolve the mask or other coatings or damage components that may have plastic in them. What is your view?

Have a great day!

 

[tsialex]

I will try to get this done today.

Would I be correct in thinking that by "neutral detergent" you mean an unscented dishwashing liquid or something of the sort?

You need a PH neutral dishwashing liquid. Some attack metals rather harshly, don't ask why I know this. Usually anything for dishwashing machines or laundry machines is not PH neutral.

Btw, sometimes you really need to use acids, like acetic/citric acid, to remove and neutralize leaked capacitor electrolyte, but you won't need it for the oily residue from the thermal pads and you apply it localized and not over the whole board.

I actually have a can of acetone; I got it for a rather delicate cleaning job but I've never used it on a PCB, fearing that it could dissolve the mask or other coatings or damage components that may have plastic in them. What is your view?

Have a great day!

I've used acetone in the past for cleaning PCBs, but some plastics are damaged by it and the cost is high. If you have enough care and use small amounts on a cotton bud then dry it immediately with a paper towel, for example, you shouldn't have problems.

Test on a small area first.

 

[Valdaquendë]

tsialex - I have now cleaned the board and re-inspected it. I was able to clean it with 91% IPA, a toothbrush, Q-Tips and canned air. I applied the IPA sparingly and found that, while it did not dissolve the residue it DID loosen it from the board, causing it to roll up into rolls or clumps when the brush was applied. Then I used the canned air to gently blow the clumps off and followed up with a dry manila brush.

I was careful to apply the air to the SMCs, as well, to blow any remaining liquid out from under them. The exposed surfaces of the board are now dry but of course I am not taking anything for granted. Unless I hear otherwise from you, I will allow the board to dry for the afternoon and will then place it in an oven that has been heated to 150˚F. and then turned off. I will leave the board there overnight.

Please correct me if I have done anything that would be inadvisable or have overlooked anything. Life is a search for truth; there is no shame in the honest commission (and subsequent recognition) of error ... as long as one doesn't commit it again.

I could find no sign that any components have been compromised; none appeared to me missing, burnt or cratered. Here are some photos:

 

[tsialex]

Did you took photos of the underside?

 

[Valdaquendë]

Indeed. Funny how tight focus can narrow peripheral thinking Here they are:

 

[Valdaquendë]

I can see that I have a bit more cleaning to do around CC739 ...

 

[tsialex]

Nothing obviously damaged, no scratches to give a clue. Did you took photos of the CPU socket pins?

 

[Valdaquendë]

Just got back from a doctor's appt. I'll take and post them now ...





I'll retake CPU-B; it's a bit fuzzy ...

 

[Valdaquendë]

tsialex - It's taken awhile to get back to you on this; other events have intervened, but I am now here to report some excellent news.

On Wednesday morning, I got back to the Mac Pro and inspected the CPU board again. I had noted (above) that there were one or two places that could use further cleanup. I went over the board, again, carefully scrubbing any areas that looked like they might harbor foreign matter. Then I cleaned the board and used canned, compressed air to clear away any debris and dry the surfaces. I paid particular attention to the SMCs, blowing at their bases to dislodge anything that might be harbored in or under the SMC or its pins and blow away any remaining moisture.

I let the board dry out for about 8 hours. Then I heated my oven to 170˚F. (the lowest it would go), turned it off, put the board in the oven (on a silicone mat) and left it there overnight.

Yesterday, I reassembled the board, installed it and, by George, it WORKED.

I was stunned, never having seen this before! Thank you! I have learned an important lesson. I had always assumed that the residue on PC boards was flux or other non-conductive substances used in manufacturing. I never thought that their presence would affect a board's functionality. Clearly I was wrong. In the future, I will clean boards when I run across them in this state.

I ran TechTool Pro and also the extended version of the AHT for this model; no errors were found. I did find that, when running the TechTool Pro 11's Full Suite, the system spontaneously rebooted during the "surface scan" of the SSD but this seems to be a function of the system going to sleep due to "inactivity" while the scan is taking place; when I wake it up, the pinwheel spins and then the system reboots. All RAM tests passed as did all the other hardware tests.

Thanks Again, tsialex!! I'm going to flash the BootROM later this afternoon. If there any other tests, diagnostics or procedures I should perform, please let me know.

 

[tsialex]

tsialex - It's taken awhile to get back to you on this; other events have intervened, but I am now here to report some excellent news.

On Wednesday morning, I got back to the Mac Pro and inspected the CPU board again. I had noted (above) that there were one or two places that could use further cleanup. I went over the board, again, carefully scrubbing any areas that looked like they might harbor foreign matter. Then I cleaned the board and used canned, compressed air to clear away any debris and dry the surfaces. I paid particular attention to the SMCs, blowing at their bases to dislodge anything that might be harbored in or under the SMC or its pins and blow away any remaining moisture.

I let the board dry out for about 8 hours. Then I heated my oven to 170˚F. (the lowest it would go), turned it off, put the board in the oven (on a silicone mat) and left it there overnight.

Yesterday, I reassembled the board, installed it and, by George, it WORKED.

I was stunned, never having seen this before! Thank you! I have learned an important lesson. I had always assumed that the residue on PC boards was flux or other non-conductive substances used in manufacturing. I never thought that their presence would affect a board's functionality. Clearly I was wrong. In the future, I will clean boards when I run across them in this state.

I ran TechTool Pro and also the extended version of the AHT for this model; no errors were found. I did find that, when running the TechTool Pro 11's Full Suite, the system spontaneously rebooted during the "surface scan" of the SSD but I doubt that that is relevant (please correct me, if wrong). All RAM tests passed as did all the other hardware tests.

Thanks Again, tsialex!! I'm going to flash the BootROM later this afternoon. If there any other tests, diagnostics or procedures I should perform, please let me know.

Glad that worked!!!

Cleaning is an old trick of mine and I always start to diagnose boards when completely clean, you probably saw before and after pictures of boards I refurbished on the BootROM thread. Dirty over the PCB is frequently conductive and makes havoc on the circuit.

Btw, flash the never booted BootROM image with the single CPU tray - don't use the dual CPU tray. You probably will need to apply contact cleaner on the socket pins and on each of the PCB connectors.

So, play safe, you don't want to replace the SPI flash memory right now, if something goes wrong when flashing the never booted BootROM image.

 

[Valdaquendë]

Yes; it was an impressive lesson. I will never again ignore residue, etc., on a PCB.

If I understand correctly, I am to apply contact cleaner (I use CAIG Lab's "DeoxIT 5") to:
• the CPU tray socket on the backplane
• the PCI-e slots on the backplane
• the Bluetooth and AirPort sockets on the backplane
• the RAM slots on the CPU tray

THEN I will flash the BootROM with the swap-out (single CPU) tray I got to test this system. Thanks for that advice.

Let me know if I'm missing anything. Thanks a million!

 

[tsialex]

Yes; it was an impressive lesson. I will never again ignore residue, etc., on a PCB.

I will flash the BootROM with the swap-out (single CPU) tray I got to test this system. Thanks for that advice.

If I understand correctly, I am to apply contact cleaner (I use CAIG Lab's "DeoxIT 5") to:
• the CPU tray socket on the backplane
• the PCI-e slots on the backplane
• the Bluetooth and AirPort sockets on the backplane
• the RAM slots on the CPU tray

Let me know if I'm missing anything. Thanks a million!

You are forgetting the connectors, like the heatsink connectors on the CPU tray and the heatsink itself, male/female.

The heatsink connectors are a constant source of SMC problems if the connexion is not perfect.

 

[Valdaquendë]

Right you are. There is also a small connector between the heat sinks that connects a cable that runs to the Northbridge cooler; I'll get that too. Thanks!

 

[tsialex]

Right you are. There is also a small connector between the heat sinks that connects a cable that runs to the Northbridge cooler; I'll get that too. Thanks!

Replace the thermal paste, take utmost care with the push pins/heatsink rivets - people frequently rip SMDs/damage the PCB using pliers.

 

[tsialex]

Btw, maybe you can buy this northbridge heatsink retainer spring kit, or preferably visit your local store and get the same self-locking nuts and screws:

Mac Pro 4,1 and 5,1 Northbridge heatsink retainer spring kit | eBay

This precision spring kit is the definitive answer to repairing your Mac Pro 4,1 and 5,1 Northbridge heat sink broken rivets. Fix your Mac Pro before you have a catastrophic failure like many have already.

www.ebay.com

 

[Valdaquendë]

I read a few articles; it looks like the screws are stainless steel, buttonhead screws, 3mm x 16mm with a nylon lock-nut. The earlier articles suggest using metal and nylon washers on both sides (under board and over cooler spring) but the kit you referenced does not incorporate them. I'm thinking of using thin, nylon washers on the bottom, just to make sure that no improper contact occurs on the bottom of the board. Have you any thoughts on the matter?

 

[tsialex]

I read a few articles; it looks like the screws are stainless steel, buttonhead screws, 3mm x 16mm with a nylon lock-nut. The earlier articles suggest using metal and nylon washers on both sides (under board and over cooler spring) but the kit you referenced does not incorporate them. I'm thinking of using thin, nylon washers on the bottom, just to make sure that no improper contact occurs on the bottom of the board. Have you any thoughts on the matter?

I personally prefer the metal screws and nylon lock-nuts, with the original springs. Easy to find and easy to install/remove when you need to reapply the thermal paste.

I don't think that nylon washers are really needed, it's ground plane there, but for sure you can use it and will protect the PCB from any scratches.

 

[Valdaquendë]

I just called my supplier and they've got all the bits and pieces for just pennies. I'll try to knock this out over the weekend (or early next week at latest) and post with photos.

Thanks again and Have a great weekend!

 

[Valdaquendë]

tsialex - I have a question; it's been nibbling at me since I assembled the materials for the BootROM reconstruction. I noted that one of the necessary factors was the cMP's serial number, barcode, MAC address, etc., as shown on the label. I wasn't clear as to why that was necessary so I looked up a number of your posts that seem to touch on this. If I understand correctly, the serial number, MAC addresses and other information on the label, as well as hardware ID information from the MLB's label are programmed into the BootROM during reconstruction.

So if the backplane has ben replaced, reconstructing the BootROM basically re-synchronizes the serial number and other info so that everything matches. Have I got that roughly right?

Lamentably, I have not yet had a chance to go through the processor board, yet this week; one of my clients (a grass-seed broker) had a full-on network catastrophe, last Sunday and I just got them straightened out by yesterday afternoon. I hope to get to it THIS weekend or early next week.

Thanks and in any case, I wish you an enjoyable weekend!

 

[tsialex]

MAC addresses are not stored inside the BootROM. Each Ethernet controller have it's own SPI flash memory for firmware and configuration (where the MAC address is set). With a MacPro4,1/5,1 with dual Intel Ethernet controllers, there are two SPI flash memories, again, one for each controller.

You can't just replace the backplane and flash the never booted BootROM image or a dump from the defective one, this is cloning and Apple detects at the next security audit, locking iCloud/Messages/FaceTime. When this happens you will have a lot to explain to Apple support to unlock your logins.

It's possible to match a replaced backplane to the case SSN exactly like Apple do when servicing the Mac Pro and a backplane replacement was needed, but this is another reconstruction process to match everything.

Never ever flash the dump from one Mac to another. Apple will detect it and the iCloud owner will have lot's of trouble.