DFI UT X58-T3EH8 continued - Part 2

Quote:

---

Originally Posted by Sam_oslo

Looking good with 12 RAM stability @4.4GHZ :up:

I have always been using Vcore Droop, and that's what Intel recommends too.

---

Thanks.

my problem is only with CPU and system temperature, they always seem too high for me to continue overclocking the system.
so i think for my daily use 4.3GHz 205x21 is about the limit i'm comfortable with.

BIOS set...
Vcore Droop Control： Enabled,
CPU VID: AUTO,
CPU VID Special Add: 114.40%,
Vdimm: 1.68,
IOH: 1.15,
CPU VTT: 1.32,

BIOS hardware monitor...
Vcore 1.37,
Vdimm 1.66,
IOH 1.13,
CPU VTT 1.36.

Everest software...
Vcore at load 1.344-1.36, no load 0.976~1.108 (CPU multiplier drops)
Vdimm 1.664,
IOH 1.136~1.152
CPU VTT at load 1.36

http://img199.imageshack.us/img199/7...0902210035.png

I need to scrap this C0 and get a D0... its giving me lots of problems and its being temperamental lol

CPU VID: AUTO,
CPU VID Special Add: 114.17%,

load
http://img196.imageshack.us/img196/3...0903031025.png

no load
http://img196.imageshack.us/img196/6...0903031051.png

My chip won't even post at that speed with that voltage. I think I might buy a couple Xeon's and just sell the ones that don't clock good.

Important note for people running 12GB on the DFI - it needs quite a lot of IOH voltage to run.
With 200x20 on the CPU and 8x dram divider (1600Mhz), I get a BSOD and reboot with prime large after a few minutes. Raising the IOHs to 1,25V (both) I am now into 2h already.
Temps are fine, mips blocks on NB and PWM.

Also watch out, some of the later boards have faulty stock PWM cooling. My own personal DFI that I got as soon as it was available actually touches the dual pulse chips, thus reducing the heatload.
The last ones I got DIDN'T touch these chips, causing the board to overheat and crash with OC and stock VRM cooling.

BIOS set...
Vcore Droop Control： Enabled,
CPU VID: AUTO,
CPU VID Special Add: 106.61%,
Vdimm: 1.62,
IOH: 1.10,
CPU VTT: 1.26,

BIOS hardware monitor...
Vcore 1.24,
Vdimm 1.60,
IOH 1.10,
CPU VTT 1.30.

Everest software...
Vcore at load 1.248, no load 0.96~1.088 (CPU multiplier drops)
Vdimm 1.60,
IOH 1.10
CPU VTT at load 1.280, no load 1.264

load
http://img197.imageshack.us/img197/6...0903214316.png

no load
http://img30.imageshack.us/img30/189...0903214712.png

what is the most stable bios right now?

hi there, i have a problem using DFI LP UT X58-T3eH8. Since the first time i've tried to boot, it just shows CF POST Code. i've tried to remove CMOS battery, clear CMOS and other things, but nothing worked. until now, it still can't boot. what should i do? thanks a lot.

What color slots do you have ram in? They should be in Green. If that doesn't work try to just boot with 1 stick in a green slot.

Quote:

---

Originally Posted by jcool

Important note for people running 12GB on the DFI - it needs quite a lot of IOH voltage to run.
With 200x20 on the CPU and 8x dram divider (1600Mhz), I get a BSOD and reboot with prime large after a few minutes. Raising the IOHs to 1,25V (both) I am now into 2h already.
Temps are fine, mips blocks on NB and PWM.

Also watch out, some of the later boards have faulty stock PWM cooling. My own personal DFI that I got as soon as it was available actually touches the dual pulse chips, thus reducing the heatload.
The last ones I got DIDN'T touch these chips, causing the board to overheat and crash with OC and stock VRM cooling.

---

Well, I had to up the NB volts to 1,3 to keep it from BSODing at prime large FFT. Maybe a bad northbridge :shrug:

On my board there is a clearly visible gap between the pwm sink and the puls chips. A thin grey piece of something hangs in the air between them. Does anyone with a correctly mounted cooler have a clear picture of how this should look like ?
My board works ok, but needs a dedicated 70mm fan directed right at the puls chips to allow any kind of oc.

Quote:

---

Originally Posted by Peen

What color slots do you have ram in? They should be in Green. If that doesn't work try to just boot with 1 stick in a green slot.

---

thanks a lot for the reply. i put the ram in the green slots. i've tried with 1, 2 and 3 sticks but nothing worked.

edit :
just now it worked with 1 sticks, but after i exit the bios screen, it shows CF POST Code again.

what should i do? thanks.

Quote:

---

Originally Posted by WeHa

hi there, i have a problem using DFI LP UT X58-T3eH8. Since the first time i've tried to boot, it just shows CF POST Code. i've tried to remove CMOS battery, clear CMOS and other things, but nothing worked. until now, it still can't boot. what should i do? thanks a lot.

---

Taken from eva's guide here http://i4memory.com/f80/initial-note...8-t3eh8-18821/ tip #30

Quote:

---

The manual has neglected to explain the memory configuration combinations for single, dual and triple channel and it's same for all X58 boards.

* For single and dual channel use dimm slots 2+4. Seems you can also use dimm slot 6 for single channel as well.
* For 3 module triple channel, dimm slots 2+4+6
* For 6 module triple channel, populate all dimm slots

---

Also be aware that the X58 and this board specifically seem to be very sensitive to DIMM placement, try numbering each individual stick 1,2,3,etc and place them in different configurations until it works... this should be all it needs. I had the same problem it was very picky with my Corsair 1600C8 stuff.

@sniper, should be no gap there and shouldn't really need much fan to keep it cool. The grey piece is the tim and should be touching both the heatsink and chips. Heres the best pic I could get. kinda hard because it was running and in case but here ya go
http://i29.tinypic.com/ohj5x.jpg


@WeHa, what ram do you have? If this is first time maybe check the bios see what version and date it is. If you can get it to boot and work fine with 1 stick I would update the bios to latest 7/24 beta. Just keep trying sounds like its not DoA so I'm sure it will work.

Thanks Peen. I might try to remount the mb sinks this weekend to see if it helps.

Well my board just died again, went pop same thing as last time. wtf....

Game over, done with this board. Was only at 2.52ghz when it popped AGAIN too

Quote:

---

Originally Posted by Sniper

On my board there is a clearly visible gap between the pwm sink and the puls chips. A thin grey piece of something hangs in the air between them. Does anyone with a correctly mounted cooler have a clear picture of how this should look like ?
My board works ok, but needs a dedicated 70mm fan directed right at the puls chips to allow any kind of oc.

---

Yes, this is exactly what I'm talking about. The temperature is measured from those dual pulse modulator chips. If they are running w/o the heatsink making contact and without a fan blowing onto them, they will heat up more and more and cause the Vcore to break down when they reach approx. 110C, causing the system to BSOD or reboot.

Quote:

---

Originally Posted by jcool

Yes, this is exactly what I'm talking about. The temperature is measured from those dual pulse modulator chips.

---

The CPL-4-50 inductors have no temperature measuring ability built into them.

Really? Funny, because it's definitely those chips that are overheating. The VRMs themselves are cooled well with the stock sink.

Quote:

---

Originally Posted by jcool

Really? Funny, because it's definitely those chips that are overheating. The VRMs themselves are cooled well with the stock sink.

---

The 4-50s have always ran hot. They are designed to dissipate up to 160A per chip. That's what made the early PWM waterblocks that contacted them and the slave chips a waste of money. The only thing that was accomplished was heating up the water.

Dude, I am telling you, they need to be cooled for high loads. I checked contact of the stock sink with the vrms, it was perfect. Still, the vrm overheated and broke down without a fan. With a mips PWM block that cools the two fat chips as well everything is fine.

Quote:

---

Originally Posted by jcool

Dude, I am telling you, they need to be cooled for high loads. I checked contact of the stock sink with the vrms, it was perfect. Still, the vrm overheated and broke down without a fan. With a mips PWM block that cools the two fat chips as well everything is fine.

---

My only intention of the original post was to correct the misinformation that the reported PWM temperature was the inductor temperature. Not the plus of minus of cooling the inductors.

Quote:

---

Originally Posted by jcool

Dude, I am telling you, they need to be cooled for high loads. I checked contact of the stock sink with the vrms, it was perfect. Still, the vrm overheated and broke down without a fan. With a mips PWM block that cools the two fat chips as well everything is fine.

---

jcool praz is right

this is a known fact and was discussed even during the x38/x48's days
that the cpl chip doesnt need to be cooled and theres no internal temp reading from them..

almost everybody who blew their board after replacing the tim on their pwms
was because they had too much contact with the cpl chip.

Hey Praz you seem like you know alot about the board and its components, DFI already RMA'ed my board and repaired it saying there was a faulty component. Well it just died again earlier today making a popping noise. Exact same thing as last time. You have any idea what could die so easily on this board?

http://img3.imageshack.us/img3/2088/20090904212150.png

http://img3.imageshack.us/img3/7012/20090904212232.png

http://img44.imageshack.us/img44/948...0904212258.png


BIOS set...
Vcore Droop Control： Enabled,
CPU VID: AUTO,
CPU VID Special Add: 110.39%,
Vdimm: 1.635,
IOH: 1.11,
CPU VTT: 1.29,

BIOS hardware monitor...
Vcore 1.31,
Vdimm 1.61,
IOH 1.10,
CPU VTT 1.33.

Everest software...
Vcore at load 1.280~1.296, no load 0.96~1.088 (CPU multiplier drops)
Vdimm 1.616,
IOH 1.104
CPU VTT at load 1.328, no load 1.296


http://img177.imageshack.us/img177/7...0904214331.jpg


http://img30.imageshack.us/img30/417/20090904213815.jpg


http://img40.imageshack.us/img40/417...0904222001.png


http://img35.imageshack.us/img35/535...0905003211.jpg


http://img35.imageshack.us/img35/961...0905003152.jpg



========================