DX48BT2, X38/X48 and Corsair DDR3

[NicholasDX48BT2]

I've just udpated my profile...The problem is I tried removing everything from the motherboard to test the issue - even the graphics card since you can boot without -- same cycling issue. With a new power supply from a different brand, 650W, and only the processor Q9300, a different memory stick tested OK (1333 Mhz 1.5V, and another at 1.7V, and tried down to 800 MHz too), and just the graphics card and CPU fan, the same message appears.

 

***** I've got some news though!! Since the results from your test, Mortimer, and the fact that I did not believe anymore in the Watchdog explanation from the tech at Intel (if I disable it, then on the first boot after power, fans come on but the screen stays blank -- I have to press the button off, then on and it boots normally)... I decided to call Intel again.

 

I got another tech support guy, I explained the situation, and the fact that some other people did the same test and did not find the bug/issue with a POST failure on boot, etc, and was wondering if the tech yesterday really got the error or what's happening here?! He went to a PC with that motherboard, and said that he did the test but got no error?! So I asked him to re-do the test, but really make sure to wait long enough after disconnecting from AC power. He redid the test, and...got the same bug as I did! He said it was the same board as the other tech used yesterday, with 1333 MHz memory, Q9450. He said his board wasn't faulty, they had it for a while, even the release, no problem, and they have the udpated BIOS. So then he said he would do some tests and call me back. 30 minutes later, he called me back, and said that he talked with other tech guys, and he did 10-12 reboots after removing the power for 2 minutes exactly every time. of those tests, he actually got the message only a couple of times, not every time...! And there was a problem somewhere indeed. On my PC though, I did see a normal boot very very occasionaly over the course of my 30+ tests, but I initially thought it was when I did not remove power long enough after the LED light went off...well maybe not then. Maybe indeed it's random like on his PC. He DID confirm he saw the bug I mentioned.

 

He also explained that he thought it might be an issue related to standby power. With the power plugged there is standby power. Unplugged, you remove it, and replug you reapply it, and maybe at that time when you start it initially does't request/get enough power from the power supply, hence POST failure. Also, I noticed how my fans start on high instead of low, controlled by mobo on a normal boot.

 

His explanation could make sense because, the guy on Tom's hardware having issue, along with the review I found on Newegg (if they are really seeing the same issue), the guy at Intel, and myself, ALL have Quad Cores processor new generation 45 nm and 1333 MHz memory. Mortimer has a dual core processor.

 

Anwyay, the guy at Intel said that at this point they needed to investigate further and they will re-contact me in about a week. Well, I may then have found a bug in Intel's board then! Do I get anything for that ;o) At least I hope it's not because of the configuration or faulty mobos.

 

** Another update regarding my PC: I've since then changed my memory, re-tested it for a long long time, with no errors this time. Re-installed Windows, drivers, etc. no problems. Rebooted a few times, tested music, video, Internet, etc. After 2 hours, I decided to re-install Norton (from the Bank Of America trial file..starts with a few words in spanish by the way when extracting the archive ?!)....finishes installation fine this time and reboots (last time it crashed, but I know one memory stick was bad). After that, installation complete, time to register to activate...CRASH. Compete screen freeze. Now I'm suspicious....OK, I've replaced one bad memory stick, but tested the new one for over 8hrs over 12 passes fine. What's going on NOW?! So, I uninstalled. My problems with the PC before happened only when I installed Norton. All freezes where during Norton, and although the 3rd install was OK, display driver not responding etc happened afterwards.

 

So I begin to think that Norton is a piece of...unless I do have another issue with my PC. It starts to be a lot when I, on purpose, chose the best components by leading and reliable companies, with all good reviews...well except for Norton (although it's recommended by PCMag, PC World, etc etc). I know I should have gone with ESET NOD32, Kaspersky etc.

 

So now I'm waiting on an update from Intel, and will test my PC for crashes. If no crashes, then Norton was the issue.

 

SUMMARY (for anyone considering a similar configuration ;o)

- Intel (DX48BT2): leading company, US-based tons of support and documentation and engineers, developing their own processors and chipsets, known for professional solutions and reliability: at least one BUG found, on their best motherboard, with udpated BIOS. (I know I didn't want to be a beta tester and wanted the older D975, but now it's out of stock, so I trusted Intel to deliver a reliable bug-free product, especially when chosing ALL standard components all at default settings). Well, I may also have a defective mobo

- Corsair DDR3 1333MHz XMS3, DHX: one bad stick found. Getting replaced.

- Corsair 620HX: minor comment, but the first power supply fan made a rattling noise; exchanged, the new one is fine.

- Norton: complete PC crash...unless it's my mobo.

 

Wow. Wasn't nearly expecting that much issues for sure. All my previous PCs were trouble free (over 5 PCs assembled before no issues!!)

[RAM GUY]

There are a few threads about this MB and some good information that you may find useful if you have not seen in already.

[NicholasDX48BT2]

Thanks for the link! I did see that post previously and some other links, it does seem that there are a couple of bugs to be resolved...However, in bother his case and mine, there was a defective memory module to be replaced.

 

My new memory modules work fine, no errors. The heat sperader with DHX technology keep them cool, as long as I have two sticks only. With four, it gets really hot, despite all my case fans (if you use fours, then sticks side-by-side on the mobo are close together). Two, they are barely warm.

 

Other the cold boot bug to be resolved by Intel (they have "escalated my case" and are doing testing in addition to the 10+ tests they did for me the other day). Note that the tech guy said: "we'll get back to you within a week regarding this issue that we confirmed and warrants further investigation...hum I meant this "behaviour", not "issue", there's no issue." Ah ah ah!!!!

 

Every thing else worked fine: normal boots are all OK, no issue with detecting the RAM and settings, nothing to change in the BIOS really, installed Vista no issue whatsoever.

 

I also confirmed the issue with Norton. Even on a clean install, crashed my PC. I installed McAfee as a test -- no problems whatsoever, no computer crashes, no issues in the Event Log. I've done a few burn-in tests, no issue, but I have to continue testing a little longer so that it is actually meaningful.

 

So withing a week I'll have answers and hopefully confirm that Intel will have a fix for the cold boot bug or tell me that it's fine, ignore it, and confirm that my PC crashes were due to my defective RAM module and Norton.

[RAM GUY]

OK Great thread and thanks for keeping it up to date.

[forbi]

Hi,

 

I'm also getting the Unsuccessful POST attempt when i first boot up my computer after i shutdown my PC. I off the power which is connected to my computer. But after i went out or sleep , the next morning when i try to switch on my computer. The message of " unsuccessful POST attempt " will appear. I didn't do any overclocking to my computer. Cos im not really that experience to play with it yet :)

 

After reading all the post above, issit because there some power left at the PSU after i shut down my computer and my main power in room. Which cause when i on my power back to have the " unsuccessful POST attempt" and not the ram problem?

 

Sorry if my english was kinda bad.

 

Thanks :)

[NicholasDX48BT2]

Hi,

 

Just to clarify -- did you remove the power from your computer (you unplugged the electricity to your computer, or you changed a switch on the electrical outlet), OR did you choose "shutdown" in Windows?

 

If you choose Shutdown in Windows, you should see a light on the motherboard next to the RAM even with the computer OFF.

 

If you did not remove electrical power, then you may have a problem. if you removed the electrical power, then you have the same problem as me.

 

I've received an email from Intel saying they are still working on the issue. I/they don't know the reason for the bug and if it's due to the specific memory we are using or some other incompatibility/bug elsewhere yet. I'll keep you guys posted when I get an udpate from Intel.

 

** I've found another bug, I posted elsewhere, I was wondering if you (and others) could tell me if you have the same issue in Vista: in the Device Manager, go to System Devices, then Intel ICH9 Family SMBus Controller - 2930, properties---> then in the General Tab, under Device Status, what does it say? And if you go in the next Driver tab, and click on Driver Details, what happens?

Because, if I just click on properties and go to the Driver tab, it shows the correct driver date and the "Intel" name. But if I click on Driver Details it says "no drivers installed", which is a bug I have never seen before. Under Device Status of the first General tab, it says "no drivers installed" but at the same time there is no yellow exclamation point, which also does not make sense. I've reinstalled Vista twice, and the second time with the newest drivers, same issue. I found the issue because Vista at some point gave a popup window in the taskbar saying there were no drivers for SMBus. Everything else and all other Intel drivers work fine...

[forbi]

Hi,

 

Thanks lots for the reply :)

 

Under my Intel® 82801GB SMBus Controller -2930 , This device is working properly. And under drive file details it say Driver files > C:\Windows\system32\drivers\intelsmb.sys

 

I didn't unplug the power cord at the back of my CPU nor from the socket.

I turn off the switch on the socket after i shut down my window. Mine is using a multi plug(i think is the wrong term for it), thoes kind that is in a row with 5 socket.

 

I think it should be some bug with the motherboard i guess?

 

Thanks :)

[NicholasDX48BT2]

Thank you!

 

If you turn off the switch on the "socket of the multi plug" (multiple outlets extension cord or surge protector), then indeed you cut the power to the computer, and you notice the same bug as I saw...

 

You can try to keep the switch on, after shuting down Windows to avoid this issue for now... You can choose "Shutdown" in Windows, but do not switch off power on the multiple outlet extension cord, and it should work... I am waiting on Intel to udpate me with a solution to our problem, I will let you know...!

 

In the device manager, you have "Intel® 82801GB SMBus Controller -2930"? Mine says "Intel® ICH9 Family SMBus Controller -2930", with the old drivers or the new ones, it's the same name...Do you have the DX48BT2? Did you use the drivers that came with the CD in the box or did you use other drivers maybe?

[forbi]

Once again thx for ur reply :)

 

Hmm , i'm kinda noob about computer :)

The only SMBus Controller under my system devices is Intel® 82801 GB SMBus Controller -2390.

I have 2 others which have the word "Family" in it. Which is Intel® ICH9 Family PCI Express Root Port 1 -2940 and Intel® ICH9 Family PCI Express Root Port 5 -2940.

Yup i have a DX48BT2 . I purchased mine from some computer dealer called "Pluto".

[NicholasDX48BT2]

Thanks!

 

I wonder why mine says ICH9 and yours is different...I couldn't find anything on the internet...anybody would have a suggestion?

[RAM GUY]

You might check on the Intel forum and see what they say.

[NicholasDX48BT2]

You might check on the Intel forum and see what they say.

 

I had checked already, it's apparently for ICH7, which is a chipset for older mobos...does not make sense. The joy of drivers...

 

Anyway my memory is working fine so hopefully Intel won't say my other issue (the boot POST issue) is not a compatibility issue with the memory or something. Tried with pretty much everything else removed so...I'll update when Intel gets back to me...

[RAM GUY]

NP Please do let us know what they say.

[eveready1010]

I recently purchased an Intel DX48BT2 and the Corsair TWIN3X2048-1600C7DHX memory, and everything right out of the box has been working fine for me...... The only thing that did not work was a USB-to-Serial adapter. I had tried everything I could except flash the bios to get it to work, so finally I did. The serial port now works (It was a driver/registry edit that fixed it, btw) but now I get the POST error message that has been mentioned in this forum (The system BIOS has detected unsuccessful POST attampt(s)).

 

The message seems to alternate between restarts, one time it shows up, the next time it doesn't, etc. I have not yet changed any of the memory or CPU settings in the BIOS, so this message is strange to say the least.

 

I would be interested in hearing what the solution to this problem is, especially since I have had virtually no problems with this configuration since it was purchased over a month ago till the BIOS was flashed.

 

btw, I tried the power button for 20 seconds with the power cable unplugged, and the very next time it booted, I got the error message.

 

I don't have a whole lot of time to tinker with this, but if I find anything out I will pass the info on. For right now, does anyone have any suggestions on what to try next? This may be a question for Intel, but does anyone know how to go back to the original BIOS ;-)?

[NicholasDX48BT2]

Hi,

 

In my case, the POST error occurs only after I physically remove power: unplug the PSU, or if you have a multiple outlet extension cord with a switch (some of them detect low power components plugged to it and switch off automatically now, I think it's the "Belkin Green" series).

 

Anyway, if I keep the computer plugged and simply use the Vista "shutdown" button, and then start, I do not get the message.

 

There is a bug for sure -- for starters, with the exact same settings, there is no reason the message should be displayed one time and not the other. If you get it even with the power still plugged in, maybe the BIOS is corrupted following the update? You can go back to the original BIOS, I think it's detailed in the manual (BIOS recovery) and the Intel website has previous versions too (I have never tried).

 

Did you change the memory voltage in the BIOS though? I think it detects the settings but not always the voltage correctly.

 

Can I ask you a quick favor? I have two other bugs myself and trying to sort that out: 1) In the device manager, under system devices, ICH9 SMBus controller-2930, click on Properties: what does it say under device status? In the next tab, Driver, do you see the version date? Then, do you get an error message if you click on the Driver Details button?

 

2) I also have a "TCP/IP error" in the event log, saying it has reached the maximum number of concurrent connections allowed. I get that even just after Windows installation, not much software installed (no download clients or anything like that). It happens after startup at some point. I wondered if it shows up in you system at start after installation, at the beginning of the event log?

 

Thanks!

[Andy0x1]

Wow, this is a long thread - I haven't had a chance to catch up on all the reading, but figured I would post my own 2 cents based on the problems I've had.

 

 

I have the same Intel DX48XBT2 board, but DDR3 Corsair rated at 1600Mhz.

 

Computer power cycles over and over on/off/on/off/on/off when you try to power it up. If you let it cycle long enough, sometimes it will eventually boot. (not overclocked)

 

I found that if I pulled the aux PCI-E power connectors from the 9800GTX video card, the computer would power on fine, (but no video obviously)

 

If you connect the power while its running, and hit CTRL+ALT+DEL, the machine will shut off, then start power-cycling again.

 

With a 8600GTS it boots up just fine.

 

Sounds simple, power supply is too weak, or protection is too agressive.

 

Tried a new power supply (600W Thermaltake PurePower) same exact problem.

 

 

Also; a note on the RAM. I previously had Patriot memory that wouldn't run at the clock advertised.

 

Replaced it with Corsiar. The Corsair worked better, but still had issues.

 

I found the solution to be setting the reference bus speed to a higher value when selecting the memory speed.

 

Its absolutely stupid because the resulting clock can be the same, yet it still seems to matter.

 

 

e.g.

bus 266MHz / Reference 266MHz / Clock 1066

v.s.

bus 266MHz / Reference 333MHz / Clock 1333

v.s.

bus 266MHz / Reference 400MHz / Clock 1600

 

You would think all of the above would result in the same result but they don't!!! Intel has some background stuff going on that's not being displayed.

[NicholasDX48BT2]

Thank you for your post! Let me give a quick update based on contacts at Intel and posts elswhere:

 

1. The power cycle on/off was more limited with the new BIOS from May 2008. Yet, it happens the first time you plug the power supply to the AC outlet, with a display from the Watchdog Timer that the POST was unsuccessful and offers to change BIOS options, even though nothing is changed, and then you reboot and it's fine. It's only when you apply power the first time with some configurations (Quad+DDR3 1333 and above?) or if you use a switch on your extension cord/power outlet obviously, even if it boots fine, if you remove AC power it will happen again the first time (not good for the hard drives!!!). Subsequent starups are fine. Change of memory, power supply etc, it's the same issue. Intel has confirmed the bug and has esclated the issue for their engineering department to investigate. I have not received any updates for a while. They did not know why it said the POST was unsuccessful, but they found the SAME bug on their own setup!!!

 

2. The SMBus issue I mentioned earlier is also a bug that I got Intel to finally confirm. They will be looking at it. It happens with the drivers 8.3 and above. I got an error message in Vista saying the drivers were missing, that how I found it, but not every body gets the message, but the result is the same in any case, there is a bug in device manager. IDCC may have additional drivers apparently that could take care of the bug. Other than the bug in the device manager and the error message in Vista that some may see, no other issues found yet with that bug.

 

3. The TCP/IP error I found in my event log is described as Microsoft as a limit to prevent virus spread, since it limits outbound connections to IP addresses that are not valid (25 failed connections per second with Vista Ultimate). So I thought because I'm the only one with the bug (not using peer to peer or torrent), that I have a virus or a bad network card in the motherboard? Noooo. This one is a Microsoft bug. If you select 30 songs or so in Windows explorer and rigt click to add to Windows Media player playlist or simply play (it's like the most basic thing one would do with Vista right??? and I do it after installation to test the sound and channels, before I install iTunes etc)...well apparently Windows connects to multiple addresses and also incorrect addresses, even if you choose all options in configuration for Windows Media Player to NEVER connect to internet for any reason. Well obviously it is still trying to connect, and to bad IP addresses too! I would like to hear Microsoft on that one. Spying? Also I duplicated the issue on other PCs (even at BestBuy!!!). You will see the TCP/IP error, which will limit your network connection and also tell you that this is for Virus protection if you search Microsoft's KB. Good job Microsoft.

 

4. I get some BSODs/"atikmdag display driver has recovered" messages, maybe once a week. It's random and not frequent at all, but usually not long after a startup. There are tons of people with that issue (Google). Other than that, it's very stable. I updated all drivers -- same. The only real issue is when installing Norton from the free 2008 BankofAmerica trial file (I'm saying that because it's the only version I tried), it freezes the computer at random points during or after installation. Then I get tons of BSODs, etc. After yet another reformat, I have no issues with other programs -- McAfee, image word editing tools, music etc, no issues during installatoin, no freeze. So is it really only Norton?? Because I never get a screen freeze otherwise, but I do get some atikmdag display driver has recovered occasionally, just on 2D desktop.

 

RamGuy: I have read on ATI's website that most likely the memory/motherboard can be involved. I will try to underclock my card (HD 3650, not overcloked, default settings) to see if it helps, then try a different memory setting to see if that helps reduce the frequency atikmdag bug. I have 4 GB of TW3X4G1333C9DHX (2x2GB). I set it at 1.7V vs. the default of the DX48BT2 board which is 1.5V (note that I got BSODs with Norton before with TWIN3X2048-1333C9DHX which is rated at 1.5V and I had at 1.5V, but OK one stick was defective and the errors may have been caused by Norton too so it's hard to conclude). 1.7V is what is shown on the sticker on my new memory modules TW3X4G1333C9DHX. I ran memtest for 11 passes overnight with no errors.

 

RamgGuy, What do you suggest as a more stable frequency and timing for that memory to try? I have 1333 MHz and 9-9-9-24 which is what is printed on the modules...what about the bus options mentioned by Andy0x1?

[Andy0x1]

Its absolutely stupid because the resulting clock can be the same, yet it still seems to matter.

 

 

e.g.

bus 266MHz / Reference 266MHz / Clock 1066

v.s.

bus 266MHz / Reference 333MHz / Clock 1333

v.s.

bus 266MHz / Reference 400MHz / Clock 1600

 

You would think all of the above would result in the same result but they don't!!! Intel has some background stuff going on that's not being displayed.

 

 

Just to clarify the above post as I didn't word that very well;

 

This board gives you the option to "reference" the frequency you want against to prevent accidental over-clocking with higher FSB speeds.

 

If you clock your CPU bus at 266 (quad pumped to 1066MHz) and select a memory speed of 1066MHz, you will get a RAM speed of 1066MHz.

 

If you raise your CPU bus speed to 333 (quad pumped to 1333MHz) and select a memory speed of 1066MHz, you will actually get 1333MHz.

 

What is odd, is (at least for me) that would fail to boot.

 

yet if I set the reference speed to 400, and the memory speed to 1600MHz while keeping the same 333 (quad pumped 1333MHz) bus speed, it would result in the same 1333MHz memory speed, but unlike the previous iteration, this one actually worked and allowed the system to boot.

 

----------

 

with a deeper over-clock of 366Mhz (1464Mhz) 3.29GHz (up from 2.4GHz)

 

1464MHz mem speed via 366 / "Default" set 1066 - Fail to boot

1464MHz mem speed via 366 / "266" set 1066 - Fail to boot

1464MHz mem speed via 366 / "333" set 1333 - Fail to boot

1464MHz mem speed via 366 / "400" set 1600 - Works Great!

 

That makes no sense.

 

Other speeds:

1100MHz mem speed via 366 / "Default" set 800 - works fine with corsair but not ********

and etc, on down the pattern.

 

 

Anyone else notice this?

 

latest Intel Bios as of this writing.

[NicholasDX48BT2]

RamGuy,

 

I'm still trying to find less "intensive" settings for my RAM. I have TW3X4G1333C9DHX (2x2GB) at 1.7V and 1333 MHz and 9-9-9-24 which is what is printed on the modules. What do you suggest as a slower speed and more conservative timing? What about 2T/1T??

 

Thanks in advance!

[RAM GUY]

It will depend on the CPU FSB as to what you can set the memory to but they will run at DDR800 or DDR1066 at Cas 7-7-7-20 at DDR1066 most likely with out any problems.

[NicholasDX48BT2]

RamGuy -- thanks a lot. What I have now is:

 

Q9300 @ 2.50 Ghz (not overclocked), Bus speed 1333 Mhz (all default options)

DX48BT2

TW3X4G1333C9DHX (2x2GB) at 1.7V

 

Memory page: custom to be able to change from 1.54V to 1.70V, but that's the only thing I changed:

Reference Frequency = <default>

Memory Frequency = 1333 Mhz

tCL 9

tRCD 9

tRP 9

TRASmin 24

tRFC 60

tRRD 4

tWR 10

tWTR 5

tRTP 5

Voltage 1.70 V

Command Rate <Auto>

 

The main screen shows (display only/not changeable):

System bus speed 1333 MHz

System memory speed 1333 MHz

System memory speed override 1332 MHz

 

What's with the 1332 MHz?!

 

So you would suggest 1066 MHz and 7-7-7-20 but leave ALL other parameters the same, correct? (same bus speed, voltage, command rate, tRFC etc...)

 

Thanks!

[RAM GUY]

If you are not having a problem there is no need to lower the frequency but if you want to see how it performs sure it will not hurt the memory. But we do not test them at those settings so test with http://www.memtest.org to make sure it is stable.

[mlhearn]

This a plea and addition to the current thread.

 

I also purchased TW3X4G1333C9DHX (2x2GB) memory and have not been able to get this to work on the DX48BT2 with Q9450. Performed all the BIOS updates, manually set voltage to 1.7V, etc. System runs cool has plenty of power... When both 2 GB DIMMs are installed in the system, it is highly unstable, couldn't load OS... The system is much more reliable with only one DIMM, and this is how I've loaded OS, applications and BIOS updates, but running applications such as IE resets the system every 10-15 minutes. Either DIMM works in either slot, and both pass all memory tests, but I can't use both at same time.

 

I've read through this thread, the other Corsair threads on DX48BT2 and other web sites, and see that this problem is very common, and I have not found anyone that has been able to get this memory to work. It seems everyone has given up and either bought new motherboards or new memory, though not sure which memory. On this site, some folks tried to get 1600Mhz version of the same Corsair memory to work, but typically ran into issues as well.

 

So, does anyone know of a particular 4GB setup that works? I prefer two 2GB DIMMS so that I can upgrade to more memory with a 64 bit OS in future? I have found a competitor version that has been tested at 800Mhz, but would prefer a faster speed.

 

Any help is much appreciated!

And thanks in advance.

[Dave England798]

Hi all please excuse me butting in i am new to these forums and am experiencing exactly the same problem as the above poster i just cannot get this memory to run at 1333mhz i have tried 9,9,9,24 and upping voltage from 1.54v to 1.7v and the board still reads the memory at 1066mhz my Main BIOS page reads

sys bus speed = 1333mhz

sys mem speed = 1066mhz

sys mem override = 1332mhz

i have updated to the latest BIOS 1782 and still the board states the memory speed is 1066mhz i am not keen to overclock i just want my memory to run at the speed it is intended i unwitttingly assumed it would run at 1333mhz how wrong i was could you please tell me Mr Ramguy what settings if any i need to adjust in order to gain 1333mhz

[RAM GUY]

Dave England798 Please try and set the memory frequency manually and if you are not sure how to set it please contact the MB make as they would be the best place to get help with thier MB and MB BIOS and how it works.

 

 

mlhearn please set the tested settings and test the modules one at a time with memtest.org and let's make sure one is not failing.