i think i ended up with mismatched xeons!!

[chipmonk010]

wellllll the title says it all i finallly got all my parts together n ive spent the last 6 hours trying to get the damn thing to work right. i got the basic setup
2x xeon lv d1s
pc-dl
i have corsair valueram in right now so i can boot with low ram voltage
5200 gfx
esp powersupply

ok so i can run each cpu by itself in cpu socket 1 no problem at all, but it doesnt want to post with both. i somehow managed to get it to post with both cpus and it passes all test and then said that the cpus did not match!! i tried it with the vcore mode(both cpus modded) and with neither cpu modded alll at stock speed and nothing. what numbers have to match for the cpus to be a match? how can i tell?

please help!

[frostedflakes]

Do they both have the same stepping code? As long as they do, the chips should work fine together.

Did you say you tried them without the U-wire mod and they still wouldn't work?

[Jeff]

Uh-oh... this is going to sound crazy but try either relaxing the screws that hold CPU2's heatsink plate(the bottom metal piece) or try putting nylon washers between it and the motherboard. (I guess you could use a bit of electrical tape in place of the nylon washers in a pinch.)

I just had the EXACT same thing happen on a rig I am putting together for someone and the problem was a shorted trace on the bottom of the PCB near one of the holes for this plate.

I don't know if the newer PC-DLs are different or what but I did not have this problem with the 5 previous PC-DLs I had.

I have two more brand-new PC-DLs here that I'll take a look at tomorrow.

BTW, does your 24pin molex connector have a retaining lip the whole lenght of it? Or does it just have a small one in the middle? The PC-DL I had the problem with was different than the previous boards in that it had a lip running the whole length.

[chipmonk010]

yeah with the u wire or without same deal.....heres my codes

CPU1
1600DP/512/400/1.30v
SL6XK
3318A473-0196

CPU2
1600DP/512/400/1.30v
SL6XK
3318A473-0099

[chipmonk010]

Uh-oh... this is going to sound crazy but try either relaxing the screws that hold CPU2's heatsink plate(the bottom metal piece) or try putting nylon washers between it and the motherboard. (I guess you could use a bit of electrical tape in place of the nylon washers in a pinch.)

I just had the EXACT same thing happen on a rig I am putting together for someone and the problem was a shorted trace on the bottom of the PCB near one of the holes for this plate.

I don't know if the newer PC-DLs are different or what but I did not have this problem with the 5 previous PC-DLs I had.

I have two more brand-new PC-DLs here that I'll take a look at tomorrow.

BTW, does your 24pin molex connector have a retaining lip the whole lenght of it? Or does it just have a small one in the middle? The PC-DL I had the problem with was different than the previous boards in that it had a lip running the whole length.

ahh hah! u may have something there!
yes i have a full thength clip!
lemmie try that n get back to ya
thanks!

[Jeff]

This is the offending hole... (on the backside of the motherboard)


My board did the exact same thing. I was about to RMA it because 5 Xeons ran fine in CPU1 but no combination I tried would run in both. I tried with and without u-wire... with and without Vdimm mods... tried 4 different sticks of RAM... a couple power supplies... a couple video cards...

Then I decided to tear everything down... box it back up... take a break... and start over. When I did this I got lazy with CPU2 and didn't re-attach it's plate. I just tried holding the heatsink on the chip(yeah... I do get that lazy sometimes). All of a sudden no mismatch error and both cpus worked fine. Thinking I magically fixed the damn thing I put everything back together and BAMMM!!!! It didn't work again. Then I got thinking that maybe I tightened the plates down too much and I caused a compression short so I loosened all four screws that held the plate for CPU2 and it worked fine again.

I started to get bummed out because I thought I damaged the board and when I went to tear everything down I took a good look at it and saw those damn traces... after I put nylon washers between the plate for CPU2 and the motherboard everything has worked fine for the past week.

I have a feeling Asus changed something with these boards recently. Like I said, the first five PC-DLs I played with didn't do this.

I hope this babbling helps.

[chipmonk010]

YOU ARE THE MAN!!!!!

it workkkks!!!
haha thats awsome!! i just put those lil red washers on the underside of all eight screw hole n bam first boot no problems at all.

i kept thinking short untill it booted with one cpu, heh, thanks asus.
yeah i was surprized that there was a clip for the power seeing as u said to becareful in ur guide cuz there was no clip..... i guess they made a couple changes.

thanks man, seriously u made my day! i owe u one

[Jeff]

I am very happy for you... but I'm a bit upset at the problem Asus just cause us.

This is going to have to be added to the guide in BOLD and maybe even BOLD AND ENLARGED!!!

[Jeff]

yeah i was surprized that there was a clip for the power seeing as u said to becareful in ur guide cuz there was no clip..... i guess they made a couple changes.

Yeah, I just checked and the two new PC-DLs I just received today also have the extended clip on the 24 pin connector. Now here's a question for you... why not add a similar extended clip on the 8 pin connector? Another one of those "why as why" questions.

Glad your Xeon monster is ALIVE!!!

[chipmonk010]

yeah for 200 bucks we shouldnt have to deal with stupid crap like that, they certainly dropped the ball on this one. thanks again man

[Rodzilla]

I don't get it... What exactly is shorting out?

Jeff... Did you get my PM? is everything coolio?

[frostedflakes]

Good thinking Jeff. Glad you got it up and running chipmonk.

[Jeff]

I don't get it... What exactly is shorting out?

Jeff... Did you get my PM? is everything coolio?

Yep, everything is cool.

The metal plates go on the backside of the motherboard to hold the IWT mounting hardware. The holes all have ground pads much like the holes used for mounting the motherboard to a case. Unfortunetly the hole I circled above doesn't have this ground pad but instead has traces to within a couple hairs of the actual hole. When you tighten everything down you are grounding the plate through the three other holes and then you are shorting to ground the traces around that fourth hole. Pretty stupid if you ask me. Either those traces weren't there on previous board revisions, the metal plates are different than previous revision, or the coating on the motherboard is now thinner and not protecting the traces. At any rate we all need to be careful with that hole I circled above now.

[Jeff]

Here's a backside picture of the hole that I circled above...

See how close those traces are?

[[XC] moddolicous]

couldn't you put like a paper washer type thing around that so you don't short it out?

[frodo]

thats what theyve done to fix it, either that or a thin nylon washer

thanks for pointing that out, as im still very close to buying a dual xeon rig!

what do the boards look with IWTs on, ive never seen a picture with the IWTs on, and how do the fans mount etc etc, just something id like to know

[chipmonk010]

just an update guys, there were no permanent ill effects from this problem im typing this from the rig right now! xp loaded fine and im folding at 3ghz as i type very pleased with it so far. i also like the twinmos mem currently running @ 200mhz 2 2 2 5 @3.2v, the chips did get a little warm tho so i slapped a fan on them, n their fine.

[Jeff]

Try your luck with 3.2GHz. The last two setups I've done have done that 100% stable.

Right now I'm burning in a rig at 2.7GHz with only 1.25v going to the Xeons. 2.75GHz cause a reset a couple hours into it. Still though... if people want a pretty darn quiet Xeon cruncher... 2.7GHz@1.25v lets the fans run at their slowest speed... even with 4 instances of D2OL running.

@frodo, I'll try to get a picture of the IWT mounted for you.

[Rodzilla]

Yep, everything is cool.

The metal plates go on the backside of the motherboard to hold the IWT mounting hardware. The holes all have ground pads much like the holes used for mounting the motherboard to a case. Unfortunetly the hole I circled above doesn't have this ground pad but instead has traces to within a couple hairs of the actual hole. When you tighten everything down you are grounding the plate through the three other holes and then you are shorting to ground the traces around that fourth hole. Pretty stupid if you ask me. Either those traces weren't there on previous board revisions, the metal plates are different than previous revision, or the coating on the motherboard is now thinner and not protecting the traces. At any rate we all need to be careful with that hole I circled above now.

Ouch I can't imagine how this wouldn't damage the board or chips... Glad this hasn't happened yet. So ASUS revised the board and ADDED a problem!!!

BTW Jeff... your pic of the back side of the hole is broken ^^^^

[Jeff]

All my pictures are broken due to Time Warner having thier finger up their butt for the past 24hrs... I have another two fingers they can use too btw.

[perry_78]

This should be made a sticky...

[Rodzilla]

This should be made a sticky...

Jeff put it in the building an Xeon rig guide...