John, It'd help if you create/update your signature so we can see at a glance what hardware you've got.
Printable View
John, It'd help if you create/update your signature so we can see at a glance what hardware you've got.
updated thanks gtj. Hitting the gym to beat out some frustrations at work then will give it another round to see if I can get it stable, have a bit of reading to do on maybe how to get around this stubborn DDR3 timings.
I've loaded up the BIOS toolkit from Intel, and extracted the default settings for the Intel DX48BT2 BIOS 1782. If someone wants to use this as a template for you settings feel free.
John,
I thought you said you have an x38 board, but your sig now says you have an x48 board. Which is it?
If x38, DDR3-1600 (800 Mhz0 is not 'officially' supported, on the x48 boards there are rules on the DDR3-1600, at in dual ranked simms on 2x2GB are not supported. For me, I have set connand rate to 'atuo' and it picks 1T on the ram I have. The G.Skill 2x2GB 7-7-7-18 version, which is what i run at, at 1.90v. add 6-60-6-8-6 I think are the subtimmings.
Now make sure you do not have 'manual' set to reference timmings, even if they are set to the same values as the 'automatic' set. Before 1782, just going to manual, would halt boot and kick off WDT to reboot w/bios defaults.
-R
John,
Honestly your board should run on defaults, minus the changes required to boot it (find disk and boot OS). I always turn on VT/HPET.
I use the outside rows for RAM, ie, the 2nd one and 4th one from CPU. I would not underclock your CPU, just let it run normal, run everything normal. do NOT select the reference memory freq other than 400 if your cpu is 400. I have issues with this when I use other values even when I had 1333 RAM.
BIOS 1814 has been posted. Unlike the 1782, the release notes are more descriptive and contain a number of fixes. Amazingly quick after the previous release.
Listed on Driver downloads and not on the BIOS page yet.
cool. thanks for the heads up.
I flashed my DX48BT2 this morning:
- fixed the power cycling issue that I was struggling with.
- initial XMP settings will at least boot now, I think I'm close to getting 1600mhz FSB w 1:2 divider now.
Will see if I can make some more progress tonight.
Nick,
Yes, I flashed to 1814.
I only got about 30 minutes to play with it but in those 30 minutes my board reacted much more like I would expect. Like you I was getting the random 3 beeps no memory, power off/on cycling, and then just the board powering up with all fans to full and never posting. If i enabled the XMP profile on the board it wouldn't even boot.
After this morning I swapped the DIMMS over to the black slots, and then set the XMP profile to 400/1600 and boom it booted, didn't quite make it to windows but it definitely doesn't sit and spin in the power cycle mode.
Will post more when I get home tonight, but definitely seems positive so far.
I found they ad posted the BIOS update on the RSS feed for this board. Updated to the new FW and re-set my settings and so far all back to normal qx9770 @ 3.6, with one bump on fsb and mch, reset default. RAM is at XMP at 7.7.7.18.60, VT, HPET on and PCIe G set, no test pattern.
-R
Nick, I flashed to bios 1814 and I still have the same problem as you. All my bios settings are on default.
Every time I boot after disconnecting the mains power the following happens:
power on, drives and fans power up, screen stays black; after a few seconds everything powers down (drives getting knackered...); and then it starts up again reaching the "POST failed (press Y for setup screen, etc)" message. After that it boots up normally, and reboots properly provided that mains power is never disconnected.
So Intel still needs to fix this.
That must be a x48 chipset thing because my #2 computer in my sig will do that sometimes when I shut off power but it dont restart it goes to the best last knowning config and starts from the backup bios.
I first saw this issue when I got the board > two months ago and I was concerned because at that point nobody had posted anything regarding this issue on forums. I contacted immediately Intel regarding this issue and spent a lot of time with them doing stupid tests after which they agreed to test on their own computers and were able to observe this SAME issue on their OWN computers to the great surprise of the Intel tech guy on the phone (great design and quality control!!!! they missed that!!!!). Like you, my main concern besides the fact that there is an unsuccessful POST and behavior for no reason while no settings are changed, is the cycling of the hard drives. Everything is on UPS now and if I need to remove power to install stuff etc I unplug the HDD on the first boot after re-applying power.
The Intel "fix" so far according to the Tech guy I talked to last week is:
- Update to the latest BIOS as usual
- Remove the battery for one minute and put the battery flat on a table upside down (this is exactly as it was explained to me)
- Replace the battery, go in the BIOS, and in the memory options, choose Advanced-> User Defined instead of Auto, but without changing any options. Then save.
I haven't tried it yet....does it work for anyone here?
Well I just updated to 1814 with no problem, we see how it goes.
Well updating to 1814 has resolved a lot of the issues for me:
- no more power cycling randomly while trying to post, the off/on that nick_dx48bt2 is suffering from.
- I was able to hit 400/1600 with the same memory, loser timings but it wouldn't even post before.
- board is a lot more predictable.
I tried to see if my reaper memory would boot at 1T with everything set to auto, but no go.:(
With bios 1814 my computer is starting a little faster.
SILVRHAND!!
You tRAS is WAY to low!!! I run 7-7-7-18, at 9-9-9, use 24-28, I have seen 27 commonly used.
but 13, WAY to low! then you can try the other timmings and lower them, make sure th voltage is at raated value of the DIMM before you lower them.
-Ron
gtj
Found your X38/X48 timing calculator and was wondering if this can be used for any X48 board? I had the DX48BT2 but was so frustrated I RMA it and went with another board. Sounds like I should have given the intel board a few more days for the new bios (1814)as I was very tired of the power cycling and the finiky start ups.