** OFFICIAL ** Intel DX38BT & DX48BT2 BoneTrail Thread

[s0lid]

BTW is here still way to disaple C1E from bios in newest bios? I haven't yet tested that 1478 bios
From everest it's possible...

[gtj]

Hi gtj

Yes, I can't boot at all. it is a coincidence we both having these problems with E8000, but I can't even say which BIOS version I have since I still couldn't boot.
I'm gonna try the suggestion of populating other slots other than the first, maybe it'll work, although intel states that this slot should always be populated.
I don't have a E6000 processor to test, I thought of it, maybe lower FSB would set lower mem freq and it would boot but unfortunatelly I can't find one to test.
Regarding DMM I do have a couple of Flukes laying around somewhere. Do you measure against a ground spot somewhere or between the other 2 VR pins?
Thanks for your help.

Definitely try each stick by itself in all 4 slots to see if you can at least get to the BIOS.

The measurement is between the tab in the picture and any good ground.

I did´nt see that earlier, changelog of bios 1478:

Be careful with 1478. Others are reporting issues with sleep and hibernate.

BTW is here still way to disaple C1E from bios in newest bios? I haven't yet tested that 1478 bios
From everest it's possible...

Nope.

[Forsaken1]

Measured ram volts.Bios set at 1.9v,meter is reading 2.0v.

Thanks for location to measure gtj.

[Manel]

I'm getting pretty pi#$ed off now, tried all slots with either stick and it's still a no go. No normal boot, no diagnostic, nothing...
One thing happened a couple of times though - several times, in slots other than the first, the board wouldn't beep at all but wouldn't boot as well, power was on, board was powered, fans would rotate but no display as well. But only sometimes this happened, next boot would go 3 beeps again.
I really don't know what else to do.
I guess next step is find that Fluke.
Any thoughts?!

[gtj]

I'm getting pretty pi#$ed off now, tried all slots with either stick and it's still a no go. No normal boot, no diagnostic, nothing...
One thing happened a couple of times though - several times, in slots other than the first, the board wouldn't beep at all but wouldn't boot as well, power was on, board was powered, fans would rotate but no display as well. But only sometimes this happened, next boot would go 3 beeps again.
I really don't know what else to do.
I guess next step is find that Fluke.
Any thoughts?!

I'm wondering if you don't just have a bad board.

[Randy]

I'm getting pretty pi#$ed off now, tried all slots with either stick and it's still a no go. No normal boot, no diagnostic, nothing...
One thing happened a couple of times though - several times, in slots other than the first, the board wouldn't beep at all but wouldn't boot as well, power was on, board was powered, fans would rotate but no display as well. But only sometimes this happened, next boot would go 3 beeps again.
I really don't know what else to do.
I guess next step is find that Fluke.
Any thoughts?!

Have you swapped out video cards and PSU's ? Just a thought.

[xMrBunglex]

i'm starting to wonder if Intel had some quality control issues with their first release. i sent my RMA out a little while ago and am expecting my replacement to arrive soon. i have high hopes for the new one.

[vipertwo]

Hmmm. I think I posted an interim version by mistake. You shouldn't have needed any patches. Sorry.

No problem, it was close enough for me to get it working -- perhaps you could update the code at http://www.peakin.com/x38/asc76xx.zip for other people.

Will you soon be submitting your asc76xx kernel module for inclusion in the Linux kernel? Thanks again.

[GJW]

Here is where I am on my lack of ability to change core voltage.

I tried the last three bios versions that support a QX9650, 1358, 1397 and 1478. This is actually all the bios's that support a QX9650. Based on Everest no matter what you change the CPU voltage to, there is no effect on CPU voltage. However if I use the 300mV offset (I know this is bad but max CPU temp was 145F during 2 minutes of 100% load so I don't think I'm in the danger zone) Everest shows the CPU voltage going to 1.48 volts (it seems to read about .05 volts low based on what I have observed). So if it's running stock voltage of 1.238, add .300 volts you get 1.538, allow for Everest error, and it's accurate. While I can certainly live with this voltage based on the core temp, I'd really like more ability to change the voltage other than +.300 volts.

I have turned it on and off, cleared CMOS, tried just about every combination I can think of to see if it works and every time the changes I made to CPU voltage don't happen, based on Everest. At this point I really don't know what else to do. I have not tried putting the QX6850 back in and trying the various bios's and what not, but I'm not optimistic. There is one older version I can run with a QX6850 for testing purposes to see if that has any impact, 1308.

If I knew for sure this was a board problem, I'd buy another board in a heartbeat, but I just don't know for sure....I'm open for anymore thoughts.

jerry

[clark_carty]

Here is where I am on my lack of ability to change core voltage.

I tried the last three bios versions that support a QX9650, 1358, 1397 and 1478. This is actually all the bios's that support a QX9650. Based on Everest no matter what you change the CPU voltage to, there is no effect on CPU voltage. However if I use the 300mV offset (I know this is bad but max CPU temp was 145F during 2 minutes of 100% load so I don't think I'm in the danger zone) Everest shows the CPU voltage going to 1.48 volts (it seems to read about .05 volts low based on what I have observed). So if it's running stock voltage of 1.238, add .300 volts you get 1.538, allow for Everest error, and it's accurate. While I can certainly live with this voltage based on the core temp, I'd really like more ability to change the voltage other than +.300 volts.

I have turned it on and off, cleared CMOS, tried just about every combination I can think of to see if it works and every time the changes I made to CPU voltage don't happen, based on Everest. At this point I really don't know what else to do. I have not tried putting the QX6850 back in and trying the various bios's and what not, but I'm not optimistic. There is one older version I can run with a QX6850 for testing purposes to see if that has any impact, 1308.

If I knew for sure this was a board problem, I'd buy another board in a heartbeat, but I just don't know for sure....I'm open for anymore thoughts.

jerry

Seeing similar results with a QX6850. I can set the voltage to 1.4V for example in the BIOS, and it shows the change. However, Everest only shows the default 1.35V. Severely putting a damper on my overclocking efforts. BIOS 1307. I've got the FSB at 375 with my memory now running at 1500MHz stable, however, I can't get the multiplier past 8 (3GHz).

[GJW]

Seeing similar results with a QX6850. I can set the voltage to 1.4V for example in the BIOS, and it shows the change. However, Everest only shows the default 1.35V. Severely putting a damper on my overclocking efforts. BIOS 1307. I've got the FSB at 375 with my memory now running at 1500MHz stable, however, I can't get the multiplier past 8 (3GHz).

Don't take this the wrong way, but I'm glad you are having the problem. That means I'm not going crazy and validates my issues with at least one other person. I also had a QX6850 in this board with the same results. I really feel there is a bios problem with this board with respect to these CPU's. I did try to pass this information to Intel but the didn't want to hear anything about it.

j

[xMrBunglex]

so nice to know Intel is having trouble writing a BIOS for... Intel chips. very reassuring.

[GJW]

Seeing similar results with a QX6850. I can set the voltage to 1.4V for example in the BIOS, and it shows the change. However, Everest only shows the default 1.35V. Severely putting a damper on my overclocking efforts. BIOS 1307. I've got the FSB at 375 with my memory now running at 1500MHz stable, however, I can't get the multiplier past 8 (3GHz).

A few more things. What is the rest of your hardware? I am really hoping that Intel just hasn't messed up that much and that there may be something we have in common in our setups that is interacting and causing this. What brand of memory, video card, etc. If we have something in common that we may think may be interacting and causing lack of core voltage control, I'll get some other hardware and try it and let you know.

Second, I have found that much over about 360 on the bus speed and it gets unstable. I used 360 and 10 on the QX6850 and the QX9650, with stock core voltage, with pretty decent sucess and stability. If I went much over about 365 on the bus speed it would be unstable. I also found a mention of this on a pretty detailed review of the board from several months back. You were probably shooting for 1500 MHz FSB. Try 360 and 10 and see what happens. Not that it's a huge improvement, but it's something to try.

jerry

[Manel]

Hi again

Been away for a couple of days.
Bad board is an option, but I wanted to be sure. Haven't swaped any hw yet since I have no other spares to.
I'm about to buy a stick Kingston Value RAM which was tested with the board by Kingston (funny how Intel directed me to their mem support page for mem sticks compatibility but there are is no mem tested with this board) and it's rated at 1.5V, if this does not boot then it will be the board.
Another option is to take it all to the shop and have them test it with other sticks which I believe I'll try first.
By the way, the Fluke measured 1.539V in the spot you mentioned Jerry, thanks. I believe we can rule out supply to mem problem, it's even a bit higher than the 1.5V it should be.

[Manel]

Hi again

Been away for a couple of days.
Bad board is an option, but I wanted to be sure. Haven't swaped any hw yet since I have no other spares to.
I'm about to buy a stick Kingston Value RAM which was tested with the board by Kingston (funny how Intel directed me to their mem support page for mem sticks compatibility but there are is no mem tested with this board) and it's rated at 1.5V, if this does not boot then it will be the board.
Another option is to take it all to the shop and have them test it with other sticks which I believe I'll try first.
By the way, the Fluke measured 1.539V in the spot you mentioned Jerry, thanks. I believe we can rule out supply to mem problem, it's even a bit higher than the 1.5V it should be.

[Manel]

New BIOS 1500 out. Check the downloads page.
Left mobo to test, did anyone test this 1500 yet? S3 problem solved?
I asked Intel regarding the S3 issue, they said they didn't know about it, but would check.

[clark_carty]

New BIOS 1500 out. Check the downloads page.
Left mobo to test, did anyone test this 1500 yet? S3 problem solved?
I asked Intel regarding the S3 issue, they said they didn't know about it, but would check.

I see they read their bug reports! I filed a bug on it over a week ago!
I'll try it tonight when I get home.

[s0lid]

Still no word about disabling C1E :F

[rchelicopter]

Does anyone know where the Motherboard temperature sensor is? I'm measuring about 45c with everest and was wondering if that's to high. All other temps are very low as it is a watercoooled system, also I have the fan installed on the NB.

For some reason I always thought the Motherboard temp in everest Was the NB, is this true..

[gtj]

Does anyone know where the Motherboard temperature sensor is? I'm measuring about 45c with everest and was wondering if that's to high. All other temps are very low as it is a watercoooled system, also I have the fan installed on the NB.

For some reason I always thought the Motherboard temp in everest Was the NB, is this true..

It's between the cpu socket and the leftmost dimm socket. The 3rd temp reading in everest is between the dimm sockets. What does it read compared to the motherboard?

[xMrBunglex]

don't ever let anyone tell you Intel has a bad return policy. they received my bad motherboard two days ago and shipped a replacement 24 hours later. nice quick turnaround.

[Manel]

I think Jerry is correct regarding the sensors positions, however, both MCH and ICH are supposed to have thermal sensors too I believe.
I think it may be a limitation of the monitoring chip aSC7621, it measures 2 remote diodes, it's own internal diode as well as a PECI, which I'm not sure if it's used for the processor digital thermometer or if the processor uses one of the remote sensor inputs. If it doesn't use the PECI, which I think it doesn't, this is not in favour of Intel since the digital is more accurate, although more difficult to handle I think. This is a rather old and limited chip.
My mobo is being tested at the moment, they'll let me know (soon I hope) if it'll be RMAed or if it hasno problem at all, and it's somewhere else.
A couple of days ago it stopped the 3 beeps (mem related), it would just power the fans but still no boot, without the sticks it still beeps. I'm not sure of anything anymore. I'm wondering if my vid card is ok? Can anyone confirm if this mobo during normal boot beeps once, like most do, when everything is ok? Is there any sound during normal boot?
Also, does anyone know what revisions of the board there are, and what differences between them? Mine is 503 but I believe I saw somewhere there is a 505, am I wrong?

[clark_carty]

BIOS 1500:

S3 Standby Mode Still Busted with Vista 32bit - Upon powering up, the machine hangs, then needs to be power cycled. After power cycling the machine resumes from Hibernate mode, so it never entered S3 mode to begin with.

Still no core voltage override sticking on the QX6850.

USB Keyboard at Startup - Issued fixed.

[Manel]

Hey Clark

Intel is really at it, they just released BIOS 1507, it seems to fix the S3 Vista issue, if it was related to RAID that is. Did you have RAID enabled in your system when S3 fails?
Let us know if it's fixed.
wow, these guys left me a good impression now, they seem to be working hard to fix all issues.
Soon enough this will be a hell of a board, can't wait to get mine back.

[s0lid]

Found this from 1507 bios change log:
Updated X48 MRC timing table values.

So they have "accidentaly" proofed the X48 Bonetrail

But here is the whole list:

- Added option for memory command rate.
- Updated X48 MRC timing table values.
- Updated BIOS Setup to detect the supported XMP profile speed.
- Fixed long boot time due to no media in SATA optical drive when
set to AHCI/RAID mode.
- Fixed issue where Aggressive Link Power Management could not be
enabled on the ICH SATA AHCI controller.
- Fixed RAID Fail to Resume from Standby (S3).
- Fixed Windows* Vista 0xC4 BSOD when running Driver Verifier on
TPM.SYS.
- Fixed time-out failures with Extended INT 13 Seek Function (INT
13 Function 47).
- Fixed issue where certain SAS HBA card was not booting.
- Fixed issue where certain USB hard drive did not work in USB
legacy support.
- Fixed IDCC2 on handling automatic mode for PCIe values.
- Fixed IDCC2 to reset the last known good values upon WDT reset
from OS.
- Fixed the title of the BIOS setup page get overlap by certain
condition.