I've had to go back to the original 0057; neither 0070 or the newer 0083 worked for me. I reflashewd the BIOS 6 times! Luckily the whole thing works fine.
A shame really as I like the board; it's a lot more stable than some brands I've used.
After my misadventures with 0070, I successfully dropped back to the previous BIOS, which at that time was 0064. Checking back a few weeks later, I found that 0064 was no longer available, and that the only BIOS posted was 0057. I suppose Intel decided 0064 had a serious issue and removed it intentionally; either that or there was a glitch in the space-time continuum, and 0064 only exists in an alternate quantum reality: An alternate universe in which Intel actually fixes its BIOS. Anyway, now I'm running this 0064 mystery BIOS, but the board is running okay, so I figured I should count my blessings and leave things alone. At least until Intel releases an update that seems likely to actually fix the problem of 0070. Hmmm, I wonder what the story behind 0064 is. Maybe it's better that I not know.
I previously had the Intel DX58SO mobo, and the results were similar: Flawless hardware, flaky BIOS updates. For the DX58SO, I eventually found the sweet spot, more or less, and left it there. We seem to have hit a roadblock with this mobo earlier in the development cycle. (And there's nothing particularly exotic about my builds, such that you might expect them to confound the motherboard.)
I stopped at 0070. At first power-up, I could hear the CPU fan hitting 100% -- then a bout of flasher's remorse began to set in. If we can't get the fan right from the get-go, they must have had the music cranked in development lab. I can't get run-of-the-mill G.Skill F3-10666CL9D-8GBRL (this is two sticks, not four) to pass Memtest 4.0a. The program crashes or immediately hangs. Oddly, it does run solid all night with Memtest version 3.5b in 0057.
I don't want anything but reliable. I don't give a crap about BIOS eye candy if I have to trade-in stability. The Intel market is presumably more about getting the job done... stability and reliability.
Has anyone had luck with Crucial? How about Memtest 4.0a or Memtest+ which is older and more like Memtest 3.5b?
I'm using 4 sticks of Crucial Tactical memory CL9 1866MHz 1.5volt for a total of 16GByte absolutely no problems at all.
Like most memory it will only set itself up automatically using the JEDEC spec of 1333MHz which ensures compatibility, but using the Visual BIOS overclocking assistant and moving to 1600 it picks up the embedded XMP settings. For some reason 1600MHz doesn't appear as a selectable XMP profile in the drop down, only 1866 appears, which incidentally worked fine but I don't run at that as prefer keeping to the chip spec. At 1600MHz it is CL8.
Who knows, this particular board doesn't seem to be getting much attention from Intel. Even the release notes didn't detail all the changes on this latest firmware, which is an indication of poor quality control. For example I wanted them to fix hard-drive security so I could take advantage of encryption Intel promote on their SSDs, there was no mention of this fix on the release notes, so if it wasn't for my adventurous nature and an element of 'risk taking', I tried the new BIOS, but I could have as easily skipped this version as it didn't say this particular issue was fixed, and be still be waiting to secure my data. Even then it has bug, if video optimisation is enabled in the BIOS you can't see the log in prompt!
Further, if release notes do not detail every change and the firmware is passed to testing, how do testing know what they should be testing or re-testing?
I also don't understand what issue there was relating to the power LED in the prior version that meant something has been changed with it that has now broken it. Surely when tested someone should have spotted the issue, and this problem has been seen on other boards then fixed with a newer BIOS, so this is a regression of a fixed bug essentially, and we must be some way behind.
The USB issues are hardware related, and Intel has issued an advisory notice that new versions of this board will have a hardware modification to fix the problem, but only when current stock is depleted!
BIOS can't control everything, but it does control a lot. In this situation, they wouldn't pursue a hardware change if it could be addressed with firmware. Have you checked this out? There's more discussion in this thread where I dropped another idea for a workaround: Intel Desktop Board DZ77BH-55K front-panel USB doesn't work properly on both boards purchased
This is the type of problem I would expect on a fast-track low-overhead design and support program. Nobody is perfect. But at least it's tangible, consistent and there are options. Trying to get memory to run if they didn't lay out the board properly to get the timing right -- that's a far more difficult and frustrating situation.
When you see a steady stream of issues like sloppy changes in the BIOS, that suggests it's being driven by a group that doesn't have seasoned coders and/or managers. This worries me. Could be a long road and it's critical to keep the pressure on and share ideas to salvage this board. This is my third manufacturer, and I can't believe how much trash I'm finding out there.
Seeing a good report on Crucial is good news on this end. I have Crucial Ballistix 1333 7-7-7-24 showing up any day now. Decided to stick with the JEDEC spec and hope for greater margins. Got the really fast RAM so I can back it down to CL9 if forced to use that option. If it works, then it should be plenty fast at 7. I'll report the results.
Thanks for reporting your mem config.
thanks for the reply naughtybits, I'm wondering if they can make an extension/adapter that you plug in, but has pin 10 missing, that way it's the same as having it removed. Or has a resister inline with pin 10 on the extension/adapter. Would be cheaper for intel that way, instead of replacing all our defective mobos.
I used one of these adapters and tested pin 10 on the adapter and found it was unconnected to anything. This is used to give me USB3 on the front of the case that internally has a standard USB3 connector. Using this I have had no issues using the header from the board, although I haven't tried any USB 3 device yet, but all USB 2 devices have worked fine. Pin 10 is related to mini/micro-USB connectors, with normal desktop USB3 sockets pin 10 has no use. I think the issue is some USB motherboard connectors are incorrectly terminating or doing something with pin 10 that should only happen with mini/micro USB sockets. A bug in the motherboard hardware means this causes a short, it shouldn't of course, if working correctly at worse the motherboard would just think it is connecting out via a micro USB socket, which has different charging arrangements I think which is indicated using pin 10.
Of course it would be nice to get some confirmation from Intel as to what the issue is exactly, and if it is as simple as ensuring the internal motherboard connector we use doesn't terminate pin 10 anywhere, then Intel should offer to send such a connector to customers free of charge.
Yes, it would be great to get a confirmation from Intel on what the design issue is with pin 10 on the USB 3 header, but I think that would open a can of worms for them. Given that the technical skills to do the following test are basic, we probably have a good chance of finding somebody who has this over-current problem and can do the following:
So far, the evidence suggests that on the systems showing over-current problems we will find a path to ground. If there's a correlation, I'd say that's all the evidence we need to be able to tackle this issue, just short of another person doing a validation.
One detail I forgot to mention...
If the USB 3 header goes to a completely plastic assembly, it will likely get its ground through the USB 3 header. Thus, if the header is unplugged to probe pin 10, you have disconnected its ground path to the motherboard and voided the test.
To describe a more universal test, the meter should test pin 10 against the metal ring or shield surrounding the USB 3 connector on the front panel of your system. Hopefully there is no path and no problem. If there is a path and you have an over-current condition, we can draw some fairly strong conclusions at that point.
When I try to Download BIOS Update [BHZ7710H.86A] 0096 for DZ77BH-55K I get the following message:
"The file that you are trying to download has either been moved, renamed or archived.
Please go to http://downloadcenter.intel.com and choose your product to find your download."
This happens with other versions of Bios for other motherborads too.
How is this possible happening with a company like Intel ?