GA-X58A-UD7 won't post, error code 2B (no system speaker, so can't hear beeps)

[jamespetts]

I am building myself a new system based on the i7 950 and the Gigabyte GA-X58A-UD7 motherboard, with a triple set of 2Gb Corsair MMS memory (6Gb total) and a PowerColor Radeon HD 5870 (which has the custom quiet cooler). I am also using an Antec P183 case and an Antec CP850 PSU. I am using a Noctua NH-D14 cooler.

Having assembled the system carefully, and attaching only a monitor, keyboard and mouse (with nothing in the expansion slots save the graphics card), I tried booting. Unhappily, no signal was ever received by the monitor.

The Antec P183 case does not include a system speaker, so I cannot determine whether there are any POST beeps. However, there is a little LED display on the motherboard which displays progress while booting. What I find is that it reaches 2B (defined in the manual as "invoke video BIOS"), hangs there for some time, then resets to C1 (the initial state) before going all the way through to 2B again and repeating the cycle endlessly.

I have tried re-seating the graphics card, and checking that the two PCI-E power connectors were secure (they were a little loose initially, but they are secure now), but I get the same behaviour. The keyboard lights (numlock, etc.) do not illuminate, but the LED in the USB mouse does at the later point in the cycle. Unplugging the mouse does not alter the behaviour.

I should be very grateful for any assistance.

[Dark Mantis]

Hi and welcome to the Gigabyte Forum.

Have you made sure that the EATX 12v  8 pin power plug is securely inserted near the CPU?

Check the connections on the 24 pin motherboard connector.

If you have a multimeter check the voltages that the PSU is supplying.

[jamespetts]

Thank you for your reply. All of the power cables seem to be seated correctly.  I then started up again and used a multimeter to test some of the lines: the red and black on one of the MOLEX plugs was 8.97-8.98v and the yellow and black was 1.0v exactly.

However, while I was testing, I noticed that the screen came into life and was actually displaying the logo! It had started up, however it was not working reliably. After a short while of running normally into the BIOS, the screen would either go blank or become covered with artefacts and distorted, and the computer would stop responding until hard reset. On one occasion on saving a setting and exiting in the BIOS, the computer turned itself off and back on again in the space of a few seconds. The failures do not always happen in the BIOS screen: some happen at the logo screen stage. The time interval before the failures occur seems to be unpredictable, but in all cases very short: anything from about five seconds to about two minutes (if that).

I did manage to get as far as looking at the "PC health" section of the BIOS screen, noting that the CPU temperature was well within a sensible limit (34C), and the North Bridge was not much more at 44C. Any suggestions in resolving this oddness would be much appreciated!

[Lsdmeasap]

Sounds like a faulty video card to me!

Here is the error code descriptions in regards to your issues.

2b
When the system powers on and the Debug LED shows the usual POST code sequence but hangs on 2b or you get a beep code as soon as the sequence reaches 2b, then the Mainboard cant detect your Videocard properly. Note that the POST code sequence might not halt at 2b but continue, so the final POST Code it displays could be FF, 00 or 7F which means the Maiboard booted, entered BIOS, or OC recovery kicked in and is waiting for Input to enter BIOS. Since the Videocard couldnt be detected properly you will only see a black Monitor and either 2b, FF, 00 or 7F. Check if there is something wrong with the Videocard, the PCIE slot or BUS, the Monitor or monitor cable. Turn the Monitor on and off, switch the Monitor cable from one D-SUB or DVI plug to the other in case your Videocard has two. If this doesnt help then reboot and load the optimized defaults. If that doesnt help either try another Videocard if possible, or try the Videocard in another system to make sure its working properly. Try another monitor or try the Monitor on another System.

7F
When the system detects that a critical error happened during the last boot attempt or several boot attempts failed, so OC recovery kicked in. There should be a message displayed on the Display informing you that OC revocery kicked in and that its waiting for your input to enter BIOS. If you cant see this message cause your Display stays black, please check the description for POST Code 2b.

00
Your in the BIOS right now, if you cant see the BIOS cause your Display stays black, please check the description for POST Code 2b. If this is part of a loop please check the point "E8 -> 01 -> b2 -> 00 looping" above.

[jamespetts]

Lsdmeasp,

I don't know whether you read my last post, but the system does sometimes get as far as displaying an image on the screen, and I can interact with the BIOS; it then fails a short time afterwards as described below. This is usually after several cycles of 1C - 2B. When it gets into the BIOS, it displays a code 75. I don't have any spare PCI-E graphics cards to test, however, I'm afraid; but do the latest symptoms suggest a graphics card problem in any case?

Edit: On trying booting to Memtest, it reaches the application and starts working, but, again, after a short while, the screen goes blank and the keyboard lights (Numlock, etc.) stop responding to keypresses.The LED displays "FF" on this occasion.

[Lsdmeasap]

Ya I read it, but the artifacting also still made me think it may be a graphical issue, because of those comments and the 2b code you mentioned previously.

75 is normal for when you are in the BIOS.

Do you have an older PC around the house or maybe at a friends, any PCIE or PCI card would work fine for testing, or if you can take your card to a friends or work and see if it acts funky there too or not.

FF means normal operation (Fully Functional), but since memtest freezes it would saty on FF anyway.   Maybe it's a memory issue, are you using the latest version of Memtest86+?   Here it is if you are not sure
http://www.memtest.org/#downiso

Do you have any, or can you borrow any, other DDR3 for testing in case it's that?

Are you setting your own Ram voltage, and QPI/Vtt voltage?   Or leaving everything on auto?   You may need to set it all up manually, auto doesn't always work for very long.

I can help with full settings if you need me to, if so please provide me a link to your memory and let me know what overclock on the CPU you want to run, if any. 

[jamespetts]

Ahh, thank you for your reply. I might have an old PCI graphics card around somewhere (but it may take many hours to find, if it can be found at all), but I don't have a spare PCI-E card, nor do I know anyone who could lend me one. All my other PCs and spare graphics cards are AGP. It was a rather old version of Memtest: I needed to write a new Ubuntu CD in any event, so I shall do that in the interim.

As to the memory, it's this Corsair memory. I have not got as far as overclocking or changing the voltage levels or anything of the sort - I can't get into the BIOS for long enough to allow me to do that! I assume that all the settings are still on automatic. I did press the clear CMOS button on one occasion when I was having difficulties earlier. The two things that I have managed to get as far as changing in the BIOS was to make the PCI-E card the first graphics card tried (as opposed to PCI) and telling it that I have no FDD. On each occasion, when I pressed enter after saving the BIOS settings, the computer did not simply restart, but shut down (fans powered down, etc.) and then powered on again a few seconds later. Is this normal?

Edit: One thing that possibly militates against it being a graphics card problem is that, after the last freeze described above in MemTest, I turned off the computer using the motherboard's power switch. A few minutes later, I noticed that the keyboard Numlock light was still illuminated, which only went out when I turned off the rocker switch on the PSU.

Edit 2: I tried removing one of the DIMMs to see whether that would help (I'd have preferred to have removed two, leaving only one, but the CPU fan was in the way of the final two). Starting it again, it reached the BIOS screen then became corrupted on the logo screen, as it often does. The code displayed after a time was then 7F. Does this help, diagnostically?

[Lsdmeasap]

Hopefully you can find a PCI card that will work, any should be fine.  Then we should be able to quickly rule that out as part of the issue or not.

As for your memtest issue, if the new one does the same then we obviously wont be able to adjust anything until we figure out the cause first.

So... For now lets do this quick memory test.   Test each module, one by one, in any white slot and see if the board will boot and run, and stay running longer in the BIOS than before.   So I'd say just test each one, one at a time in any white slot, and enter the BIOS and sit there for 3-5 minutes and see if it stays running.   If so, move on to testing the next stick, this will rule out any one stick as being possibly faulty and causing these issues when you use the whole set.

Yes, it's normal to shut down after making BIOS changes, sometimes it will fully power down, other times only partially depending on what you are changing.    Nest time you can get in there please try to load optimized defaults as quick as you can and then hit F10 to save/apply/reboot.

Some USB keyboards and or mice remain powered up while the board is shut off, so that is normal, PS/2 as well sometimes depending on the BIOS settings.


Memory must be in the white slots only for 3 modules or less.    So please number each module with a sticky note, or some tape and paper, and see if you can do the above test or not.

7F error is noted above, mainly it means that settings failed and fail safe defaults are loaded, which is why I mentioned for you to try to load optimized defaults as soon as you can.   If you do end up being able to, once you reboot go back into the BIOS and set 1.66V ram voltage, this setting is in the MIT section, then into sub-section 4 at the bottom of that section.

I'm still on the fence as possible graphic issue, but it could be memory related as well (Settings or faulty dimms)

Do you have a case speaker connected?   If so, do you hear any beeps with you see 2b on the debug LED?

[jamespetts]

Thank you very much for your reply - that is most helpful. I eventually managed to track down a PCI graphics card. With that in place instead of the PCI-E graphics card, the system works fine. The problem is definitely related to the graphics card, therefore. The remaining question is this: does that mean that the card is certainly dead, or is it still possible that the issue could be caused by a setup or configuration issue?

[Dark Mantis]

Just to cover the obvious, you have got both power connections attached to the GPU?

[jamespetts]

Just to cover the obvious, you have got both power connections attached to the GPU?

Yes, I'm afraid that I checked that several times.

[Lsdmeasap]

Thank you very much for your reply - that is most helpful. I eventually managed to track down a PCI graphics card. With that in place instead of the PCI-E graphics card, the system works fine. The problem is definitely related to the graphics card, therefore. The remaining question is this: does that mean that the card is certainly dead, or is it still possible that the issue could be caused by a setup or configuration issue?

Awesome to see you were able to narrow it down, sorry to hear it is the more expensive part though!    Best to setup a RMA on the card unless you can test it somewhere else to be sure, or test another PCIE card in your board to be sure, if you were thinking it's the slot instead of the card or something like that.  

And if you want to test the slot, if that's what you meant I mean, you can test the card in any of the other PCIE slots and it should work normally if it's not faulty.  It's not a setup or driver issue, driver issue wouldn't appear until windows so you know for sure it's not that, and you really can't mess up the setup of a graphics card as long as the power is connected properly it should work.

[jamespetts]

I just tried putting the graphics card into another slot: exactly the same result. Will have to return the card, I think :-(