you can be like me n collect em lol
even thought i just got the best ddr2 board ever created "ut p35"
i will be buying the dfi p45 100%
Printable View
im starting a nice collection myself. maximus formula x38, dfi lt x48, ip35 pro p35... havent installed the dfi yet so im still on maximus converted to rampage. so far the asus is far from enjoyable... not enough control over the board. this is why i went dfi. and im not doing sli as of yet so no nvidia chipsets in my collection... except for an ancient nforce2 mobo (sempron equiped)
now all i need are a few cpu's a couple of gigs of ram (probably going to get another set of what im currently using as i was 100% satisfied w/ these)
my ghetto solution to the cooling the mosfet
a 40mm fan, a long screw and a nut. I just used the existing screw hole on the back sink
works very well...
http://img246.imageshack.us/img246/826/img1344cz5.jpg
im making a back plate for pwm (northbridge already done... did a little write up about it here)
i gotta get the hack saw for the back plate but i got a nice candidate... nice block of copper i sourced from the stock heatsink of 1 of my 2900 pros. just heat it up on a gas stove and let the solder heat up... lifts right off. cut off *still have to do this :(* 2 corners diagonal from each other and the 2 that are left are a perfect fit for the 2 push pins :up: just get some posts w/ springs (got some of those already from extra parts) and you have a nice back plate.
i still have to figure out the fan solution. i have 2 fans that are a perfect fit... just need to figure out the best way to mount them. im thinking using the existing screw holes as you did but making a small bracket from the spare copper fins so i can have it exhaust out the back
great job man, you should really consider watercooling the NB, mines is 35 loaded MAX.
I think I can picture the bracket you have in mind, if your going to exhaust the heat then the fan wouldn't be facing down towards the exposed mosfet area, just the heatsink. The heatsink cools fine but the exposed mosfet area is what really gets hot so you might want to reconsider that..
what are you getting now as your PWMIC temp?
i havent even plugged it in yet! :rofl:
hey, i could easily make a heatpipe cooler with a vice, a torch and my snips w/ the extra parts from the 2900pro hsf assembly! i could have the fans cool it all w/ them sandwiched inbetween the 2...
man im getting ideas! i knew all this junk i have could be useful!
*edit*
just tacked it together... need to do a little trimming but its gonna work. already got the mounting problem solved... if i trim the back corner about 2 mm to make a right triangle, there is clearance for a copper plate that could be attached... now all i need is the plate (looks around room)
i will have pics up in a min.
*edit*
pic of idea
http://i48.photobucket.com/albums/f2...cture002-2.jpg
just have to solder it back together, add a few more fins to make it wider, trim up a few pieces, cover over the solder joints w/ a sticker, polish it and it will be nice! i can still mount the fan on this and still cool both components :up: i think this will be saturday's project. like i said, i havent even fired up this board, but i know i can make it scream.
btw that copper heat sink is recessed into the black pwm heatsink so there is some overlap, but w/ 2 fans mounted over it i dont think it will be a problem.
also, im watercooling the cpu (may consider northbridge once i see temps... but as for now, i have confidence in the cooling solution provided + my backplate mod... that fan on the nb in the pic is a AFB0612HHC 25cfm delta... more spare parts laying around...) so clearance is not an issue... though it is still quite roomy
Greetings,
Having just recieved my UT-X48-T2R and got it at least booting up and vista installed I am finding the system rather sluggish, perhaps I was expecting more from the new processor or as I suspect ive not got it configured correctly.
Installed is the Q9450 and gskill ram 8500 1066mhz ddr2 memory 2x2gb.
I dont initially want to OC it as I need to get some work done and just need it stable but running at a decent pace.
The fsb is set to 333mhz and the memory in the bios shows 333:1066. Using the memset utility it shows fsb/dram ratio 5:8 but shows 533mhz which I presume is the non ddr figure, not sure if this is correct for my intended configuration.
Graphics card is a ati 3870 with 512mb ddr4 ram.
The bios is as shipped as I dont see a newer one.
It doesnt seem to be overly stable and nothing like the performance I was expecting. I would really appreciate any comments on what in the bios could be causing this.
If you need any further information please ask, I will try to reply quickly.
Many thanks
BS
bcsbcs: not setting your ram settings manually will cause instability. It doesn't sound like you've done much apart from install the system. With my g.skill sticks on my x48 and x38 I had to set tRFC to 52-56 manually to even think about being stable. Check to make sure they are detected at one of those settings. Setting your dram voltage is also very important too. Might need a hit to northbridge volts since they're 2x2 sticks with a quad core. I don't think I've had a dfi board + ram combination yet that didn't require me at least entering my ram settings for stability.
Thanks for your input Dryden,
I just got a copy of Everest 4.5 and it reports the following:
Memory Timings:
CAS Latency (CL) 5T
RAS To CAS Delay (tRCD) 5T
RAS Precharge (tRP) 5T
RAS Active Time (tRAS) 15T
Row Refresh Cycle Time (tRFC) 52T
Command Rate (CR) 2T
RAS To RAS Delay (tRRD) 3T
Write Recovery Time (tWR) 14T
Read To Read Delay (tRTR) Same Rank: 4T, Different Rank: 6T
Read To Write Delay (tRTW) 9T
Write To Read Delay (tWTR) Same Rank: 11T, Different Rank: 6T
Write To Write Delay (tWTW) Same Rank: 4T, Different Rank: 7T
Read To Precharge Delay (tRTP) 5T
Write To Precharge Delay (tWTP) 14T
Precharge To Precharge Delay (tPTP) 1T
Refresh Period (tREF) 4171T
DRAM Read ODT 3T
DRAM Write ODT 7T
MCH Read ODT 8T
Performance Level 8
Read Delay Phase Adjust Neutral
DIMM1 Clock Fine Delay 8T
DIMM2 Clock Fine Delay 6T
DIMM3 Clock Fine Delay 8T
DIMM4 Clock Fine Delay 6T
I have litterely just got this and havnt even gone through it myself but as the last PC i built was 3years ago I will have to spend some time reading up what I can.
If anything looks obviously bad then please let me know.
Thanks agian.
BS
bcsbcs
May want to set the memory voltage to 2.0V. Other then that nothing should need to be changed for stability using stock speeds.
I'm not sure if I'm going for the T2R or T3R, but which memory would you recommend for each -- easy to OC, but reasonably fast results?
Thanks Praz,
With them settings it didnt seem too bad and it managed a complete run through of 3dmark 2006 so thats a good base point to continue fiddling :)
I have an almost identical system to build also but with 4*2gb gskill 8500/1066 in the same board and I'll probably have fun trying to get that to work too.
Regards
BS
4x2GB will most likely require increasing the NB voltage even at stock speeds. Possibly a little secondary memory timing manipulation also.
I had quite a bit of trouble with 4x2 gb on the pc2 8000 sticks with the x48 board. I think I settled around 1.35 nb real with 400 fsb on an e8400. In order to go past that it required some significant tweaking for my particular hardware combination. I ended up mailing G.skill and they said that 4x2gb of the pc2 8000 or 8500 sticks wasn't recommended with the x48 chipset at DDR2 900 or higher. I am guessing you'd need at least 1.35 for stock settings with the quad and 4 sticks -- Praz could probably tell you better as I haven't tried a 45 nm quad core yet with this board.
"Also what's the deal with the random C1 hangs with this board? On every cold boot I get a C1 hang which lasts a few seconds then the computer reboots itself and everything is fine. Then on random reboots I get a C1 hang which clears after I reset the computer."
thats dfi for you period.i have a dead 680i lt(back from repair and dead again) and a rd600 dead with c1 errors on each board.even before they just totally stopped working they would do the same thing"c1" crap.im done with dfi i think for a long time,im so sick of there c1 errors even with modest clocks.i used to think dfi was the best baords on the planet,now i know there the best bios,thats it...there dfi ultra d series was the last of the bullet proof boards.
Yeah, I've tried just about everything that I can think of--increasing NB all the way up to 1.45V and increasing Vdimm up to 2.1V--and I still can't get away from the "C1" error on cold boot and the occasional "C1" error on reboots. My settings are now 24 hours stable using both small FFTs & blend. I have also exceeded 20 passes using memtest 2.01 w/o any errors. I posted the same question over on Clunk's site any he couldn't offer any advice either.
I have to admit though that this isn't just a DFI issue because I have had this same problem using an Asus Maximus Formula. The problem didn't arise with that board though until I started using my QX9650, there was no "C1" issue using my x6800. So I am wondering if it isn't the QC that's leading me down this road of frustration....
So I got occt ram test to run 10 hours without crashes, but I still have one core thats slow. In P95, worker thread 3 is very slow, as in 6 passes on the other vs 1, usually causing a bsod shortly. which GTL is assigned to each thread, and do I raise the one, or lower it? Also, I purchased a voltmeter (probably not the name) and I'm wondering if it's worth the risk to just measure from the board and find my values. My goal is atleast 3.95 at 1.55v. (Q6600)
I had trouble with only one set of those same sticks. the 2gbx2 apparently are very finicky to the memory options and NB gtl. Atleast for me.
My NB needs a Minimum of 1.476 bios setting to run stable at the speed in my sig. Seems high, but lower it 1 notch and I get popped back to the Desktop in games.
Larry