Mac Pro 5,1 AHT reports error

[rolltide4life]

My Mac Pro 5,1 intermittently shuts down for no reason at random times. I decided to give AHT (Apple Hardware test) a try. I had to download and install it on my usb via GitHub. Anyways, here is the screen shot - https://imgur.com/a/vZaI3

Does that mean the logic board is failing? What exactly is it caused to fail?

Thanks.

 

[adam9c1]

Is this a dual CPU model?

Download TGPro and look at Northbridge temperatures.

Perhaps you have a problem with the Northbridge heat sink (common on dual socket systems).

 

[h9826790]

What's your spec, especially if you installed any 3rd party GPU . PCIe card. AHT didn't decide to handle those hardware.

If possible, you should run ASD but not AHT.

 

[rolltide4life]

thanks for the replies. Here are my current specs for Mac Pro 5,1:
dual 3.46 ghz system (x560)
96 GB of Ram
nvidia 1080 ti (macvidcards edition)
1GB Apple PCIe SSD with sintech PCIE Adapter


Do you know where I could download and run ASD?

Thanks again for all of your help.

 

[Squuiid]

nvidia 1080 ti (macvidcards edition)

How are you powering that?
Very likely the reason for your shutdowns. For me the 1080 is the sweet spot for a cMP. The Ti draws too much power for comfort.
http://www.macvidcards.com/blog/the-pesky-power-issue-with-pascal-1080ti-and-titan

 

[rolltide4life]

How are you powering that?
Very likely the reason for your shutdowns. For me the 1080 is the sweet spot for a cMP. The Ti draws too much power for comfort.
http://www.macvidcards.com/blog/the-pesky-power-issue-with-pascal-1080ti-and-titan

Thanks for the reply. I have two 6 pins to one 8 pin adapter and sata to 6 pin adapter to power 1080 ti card. I have tried using external 500e power supply to power 1080 ti card only. I’m Still getting random shutdowns.
I have replaced power supply for my Mac Pro 5,1. with no luck. Thats why I provided the screenshot of AHT error in this thread to see if anyone can decipher what the error message means.

Someone suggested that I should try ASD, but i dont have it. Does anyone know where i can download and run ASD?

Thanks again for the replies.

 

[h9826790]

thanks for the replies. Here are my current specs for Mac Pro 5,1:
dual 3.46 ghz system (x560)
96 GB of Ram
nvidia 1080 ti (macvidcards edition)
1GB Apple PCIe SSD with sintech PCIE Adapter


Do you know where I could download and run ASD?

Thanks again for all of your help.

AFAIK

4SNS mean System sensor problem.

VeAS means V (Voltage), e (PCIe slot), but I don't know the details about AS and those numbers, most likely are the details of the error value (e.g. which slot).

Since it's a sensor. So, not really the logic board failing. If logic board fail, the code should be 4IRP, or 4MLB, etc. But not 4SNS.

However, a sensor failure definitely can report an excessive value which eventually trigger the shutdown protection.

Since not sure which slot has problem and you only need to use two slots for GPU and SSD. I suggest you move the GPU to slot 2 and SSD to slot 4. And see if this can avoid the trouble.

If not, and you have another GPU. I suggest you replace the 1080Ti by a known good GPU. Just in case the flashed / modded 1080Ti's power drawing patten trigger the sensor error.

Anyway, there was no 1080Ti or PCIe SSD back in the AHT written. It's quite normal that AHT consider those unknown value as error (e.g. VeAS may actually means the Voltage sensor on the graphic card, but not really the slot. And the SMC cannot understand 1080Ti's sensor's reading, therefore, consider it as an error).

For Mac Pro 5,1. You need ASD version 3S149. It may be a bit hard to find, but Google should able to help.
[doublepost=1506018480][/doublepost]

Is this a dual CPU model?

Download TGPro and look at Northbridge temperatures.

Perhaps you have a problem with the Northbridge heat sink (common on dual socket systems).

Cannot be Northbridge. 4SNS is sensor's error, even if that's the Northbridge's temperature sensor. The code should be TNxx but not Vexx.
[doublepost=1506018584][/doublepost]

Thanks for the reply. I have two 6 pins to one 8 pin adapter and sata to 6 pin adapter to power 1080 ti card. I have tried using external 500e power supply to power 1080 ti card only. I’m Still getting random shutdowns.
I have replaced power supply for my Mac Pro 5,1. with no luck. Thats why I provided the screenshot of AHT error in this thread to see if anyone can decipher what the error message means.

Someone suggested that I should try ASD, but i dont have it. Does anyone know where i can download and run ASD?

Thanks again for the replies.

So, even though the GPU is at idle, the Mac Pro may still shutdown randomly?

 

[rolltide4life]

h9826790, I was able to find 3S149 and ran the test. Here is the results - https://imgur.com/a/tb2Gn

That tell us a little more what VeAS is - PCIe Boost A, 12V test failed. Just want to come here long enough to post this and see what y'all think of this error message. I haven't done much other than wanted to run the test using ASD really quick to give you guys the results. I will post more tomorrow.

I do appreciate all of you to take the time to reply to this thread. Thanks again.

 

[FireArse]

This is an interesting one.

You’ve already had great feedback. From memory, you can ask ASD to continue past errors. It would be interesting to see the next 5-6 tests after the failed test.

That’ll help correlate whether the sensor applies to all your PCI slots, or just one of them.

I’d have a supported GPU handy while you consider your options. Moving the card to another slot offers the quickest, safest resolution..

F

 

[rolltide4life]

This is an interesting one.

You’ve already had great feedback. From memory, you can ask ASD to continue past errors. It would be interesting to see the next 5-6 tests after the failed test.

That’ll help correlate whether the sensor applies to all your PCI slots, or just one of them.

I’d have a supported GPU handy while you consider your options. Moving the card to another slot offers the quickest, safest resolution..

F

[doublepost=1506052830][/doublepost]I went back and did some more testing on this Mac Pro 5,1 tonight. I have moved GPU to slot 2 and SSD to slot 4 as per h98 suggested and that did not alleviate the issue - I still got VeAS is - PCIe Boost A, 12V test failed via ASD and I have continued the test past one error and there are no other more errors. The only error I have is PCIe Boost A, 12V test failed.
I have linked two screenshots below:
https://imgur.com/a/Ccdl9 and https://imgur.com/a/ZRqub

what you see is the screenshot of:
PCIe Boost A, 12V failed with reading of 0 V
PCIe Boost B, 12V Passed with reading of 12,160 V

Can anyone tell me where is PCIE Boost A and B located in Mac Pro? Is it on Logic board? Do you think they may be
the two mini 6 pin connectors that feed the power to GPU? Perhaps its fried that it can't handle the power load for 1080ti ? just thinking out loud here.

I appreciate any more feedbacks and suggestions.

Thanks again.

 

[h9826790]

h9826790, I was able to find 3S149 and ran the test. Here is the results - https://imgur.com/a/tb2Gn

That tell us a little more what VeAS is - PCIe Boost A, 12V test failed. Just want to come here long enough to post this and see what y'all think of this error message. I haven't done much other than wanted to run the test using ASD really quick to give you guys the results. I will post more tomorrow.

I do appreciate all of you to take the time to reply to this thread. Thanks again.

That’s exactly why I recommend ASD, so much clearer. No need to decode.

Anyway, if that’s the voltage below lower limit. It should not shutdown the Mac. And it should not affect anything when you power the GPU by external PSU.

Correct, the mini 6pins are the PCIe Boost A and B.
[doublepost=1506066782][/doublepost]Sensors are just sensors, not necessary means the system not working. e.g. a sensors can shows a fan has zero RPM when it’s actually running at full speed.

The sensor cannot read the voltage doesn’t mean that the power line doesn’t work, just you can’t monitor it. And as I said, under volt should not shut down the Mac. But of course, if it’s really zero volt (sensor report correct but abnormal reading), then it can’t be used to power any GPU anymore.

 

[rodrigore]

Hi, sorry to resurrect this thread, but Im having the same report of apple hardware test and I also have random shutdown.

Did you fix the problem? or maybe someone can answer me if is possible to fix the PCIe Boost A or I need to change the logic board?

Thanks

 

[h9826790]

Hi, sorry to resurrect this thread, but Im having the same report of apple hardware test and I also have random shutdown.

Did you fix the problem? or maybe someone can answer me if is possible to fix the PCIe Boost A or I need to change the logic board?

Thanks

Which GPU you are using now? Is that connected to the mini 6pins?

 

[rodrigore]

Im using an NVIDIA 1060 3 gb superclock which have an 6 pin, so I need only one 6 from the mac pro. I use this card in my other mac pro, because I know that I dont have any issue.

Now I run the ASD and I get the exact same message that the author of this thread: " Error - Sensor is reading below the low limit".

If I need only one mini 6 pin from the mac, can I use the PCIe Boost B only?

 

[h9826790]

Im using an NVIDIA 1060 3 gb superclock which have an 6 pin, so I need only one 6 from the mac pro. I use this card in my other mac pro, because I know that I dont have any issue.

Now I run the ASD and I get the exact same message that the author of this thread: " Error - Sensor is reading below the low limit".

If I need only one mini 6 pin from the mac, can I use the PCIe Boost B only?

Yes, you can only use the Boost B.

If you Mac is working flawlessly, then I won't worry too much about Boost A considered INOP.

If you need more power for GPU later, Pixlas mod can give you enough power to drive even a Titan Xp.

 

[rodrigore]

Thanks for answer about the Boost B.

Im not sure if choosing the PCIe Boost B fix the problem, because the shutdown are entirely random. Maybe one week with no shutdown, or maybe a day. From yesterday im using the Boost B but maybe tomorrow will be suffer a shutdown. In that case I will report about this problem.

In case that I want to fix the PCIe Boost A, the only solution is to change the logic board?

 

[h9826790]

In case that I want to fix the PCIe Boost A, the only solution is to change the logic board?

That should be the easiest way to do it. Otherwise, you will need to know precisely which exact electronic device causing the issue, has the spare parts, and know how to replace it (with proper tool).

Which also means, buy a whole used cMP, and then sell the current one (or the remaining parts) should be an even cheaper as well as easier way to leave this situation.