You really ought to buy a digicam that's not from Kodak or HP sometime... :p:Quote:
Originally Posted by Tony420
But in all seriousness, a digicam with a decent macro mode would do the trick
Printable View
You really ought to buy a digicam that's not from Kodak or HP sometime... :p:Quote:
Originally Posted by Tony420
But in all seriousness, a digicam with a decent macro mode would do the trick
Ohk...you guys might as well forget the vdimm = vio mod because its not gonna fix the OVP issue..
-CaT
Ok after buying A booster modding it, then spending bloody hours working out whats up with my board, the penny dropped OVP :stick:
So is there anychance of a fix, my OVP kicks in @ 3.55 3.6
I'm not sure about this, but I'm thinking that the OVP might be done using the winbond hardware monitor, and the OVP level is set in software. What BIOS' are people running on the board?
Im on the 1/28 4 volt but I have tryed just about every bios including non 4 volt
My board OVP seems to kick in @ different voltage sometimes 3.5 using booster 3.6 using bios. driving me :mad:
Same results here...
I've tried most of the bios's with the booster, varying drivers just in case, nothing got me past 3.6vdimm, and even upping once booted in gave me no results, so I'm assuming ovp still blocks anything over 3.6 from the booster...
Gray
Quote:
Originally Posted by persivore
I think you maybe onto something, after reading some pdf on winbound chip it is indeed capable of OVP. This could explain having to remove and swop Ram to get it to boot again. All we need now is a Bios Guru that can maybe unlock it :banana: :banana:
If the chip does control OVP could you not just cut trace to the correct ping out?
If I find pin out for chip could a bright spark help out?
I don't know if cutting a trace would work, and I'm not even sure if the hardware sensor is controling the OVP. Cutting the trace might also cause a problem if the pin is left to float, becasue it could chose to float high, which would constantly trigger OVP.Quote:
Originally Posted by jiff
Personally I dont think that there is a specific OVP circuit on the UT board.
My own system becomes unwilling to post with the 3.3v line too high regardless of the vdimm setting in the bios.
I can get it to go higher if I boot the system at 3.3=3.3 then increase the 3.3v line by tweaking the PSU up.
But this is messy and not very safe IMO.
There may be some other component using the 3.3v line that causes posting problems when you have it too high.
I think the best way to get high Vdimm on this board may be to cut the 3.3v line input into the vdimm mosfets and feed in 5v. (as on the 939 board)
This may need you to retrack vref and might be awquard to acheive, but I think it might work.
Unfortunately I dont have any money to buy a new board, should I kill mine modding. So I wont be trying this out.
I think that there is still the problem with OVP when using a DDR booster :(. Do you think the problem could be caused by the mem controller on the CPU not liking the high voltage, or can other s754 boards run 3.6v+ through the RAM?Quote:
Originally Posted by Holst
I dont think its down to the CPU.
You might damage the CPU running high Vdimm (although instance of this is very unusual with the newer A64 steppings) but I think the CPU can take 4v+ as it does on other boards (and with the UT if you mess around enough)
I cant test using DDR booster as my fortron PSU wont work with it.. I dont understand how my PSU works enough to get it working. I think the PSU detects a grounding/short fault and wont power up.
Maybe that's hit the nail on the head, and that's the reason...
If you volmod your psu beyond 3.6v on the 3.3rail it can get unstable and not boot...
Perhaps it's not really ovp as such, but anything over 3.6v makes the system unstable due to voltage leaking into the 3.3rail...
Perhaps it's affecting other components as it's, from the look of it, connected pretty much directly to the 3.3rail, as raising the 3.3 also raises the vdimm...
A custom cap/diode to block >3.3v from leaking out of the vdimm circuit?
Beyond me...
Gray
Its not the 3.3 rail the OVP kicks in with booster as well @ the same voltage as the 3.3rail mod.Quote:
Originally Posted by Gray Mole
I may buy a Neo, I can only get 3.5volts and underload the droop is awful
Have you guys tried with the NF3 to boot at 3.3Vdimm - let's say - into Windows and afterwords play with the Vdimm?.......If it's a "post" OVP, then the mobo must work OK........but if it's a "main" OVP then providing more Vdimm within Windows, mobo will freeze or even shut down..........Which of these two happens?......
Quote:
Originally Posted by hipro5
ok I put booster back in :eek: booster booted @ 3.4 volts turned up to 3.8 in windows no problem apart extreme droop :rolleyes: Turned the booster down to 3.5 rebooted :nono: boot had to wait an hour with cmos battery out to get it to reboot. (without booster)
My OVP floats between 3.5 3.6
So it sences OVP only on boot, could this be the Winbound chip?
Could you try to lower Vdimm inside Windows VERY-VERY slowly from the DDR Booster to see if it'll reboot or not......?
Booted with booster set to 3.3 lowered VERY slowly down to 2.8 no problem raised slowly back up no problem.
Tryed it with Bios Vdimm set to 3.3 and 2.8
Hi Jiff,
I got your PM. Do you guys know what is the vdimm regulator for this board? I read Catalyst.X saying something about LM358, is that the one? BTW, how do you expect the regulator to work if you cut the trace that feeds FB and solder a trimmer from FB to ground?
Anyways, if it is LM358 indeed, then you could check the volt reading on pin 2 and 6 of it, so Persivore can tell us which one is for vdimm.
If it is pin 6 then solder a 50k pot between pin 5 and pin 8, or if it is pin 2, then solder a 50k pot between pin 3 and pin 8. Lower resistance for more vdimm.
I believe that should take care of limited vdimm.
As always, do at your own risk.:)
Cheers M8Quote:
Originally Posted by Malves
I think the problem is noone knows where or how the OVP is controlled.
Be nice if you guys could put ya brains together Im sure you would get mod work out of it. (sure as hell got me)
looking on here there is ALOT of DFI NF3 about. Coupled with the Maximizer I can see the demand rising.
My OVP has gone mad (if its the ovp)
Now i have to remove battery also for almost 2 hours to get up and running again.
Before i could just lower volt and reboot.
Any news on this ? im willing to try anything and if it kills the board i dont care cuz its going to trashcan anyway if not solved.
I got 50k pot and solder iron ready if you show what of the 2 chips to solder on.
There is one above dimms and there is on below dimms.
I too would be very grateful if someone with knowledge of such stuff could help figure out how to remove the OVP on this board. I was playing with mine a little yesterday, OCZ 520W and 2x256 HyperX BH-5..... With PSU 3.3v @ 3.67-3.69v and vdimm at 3.63-3.64v under load (measured from dimm slot, mobo is wire-modded) my max benchstable is ~270mhz. I feel certain that with ~3.7-3.8v I could do higher, like high 270mhz+ range, maybe 280mhz if I got real lucky. But as is, that is pretty much the max V's I can run. When I adjusted the Powerstream higher, to ~3.73v on 3.3v rail (which would put vdimm at ~3.68v), my cathodes and fans came on but rig wouldn't boot up =/
Btw side-note, I wish the Powerstream pots weren't so fookin sensitive, makes it hard to precisely tune the voltages....
im totally there...
my OVP kicks in w/ my 3.3 @ 3.600 on a DMM... if i have it set to 3.59 it boots. in windows i can crank it up... the ovp is on POST only.. not constant. after post you can up it above 3.6... but it sucks because its a pain when im trying to bench and restarting = turning pots.. taking readings... re-posting below 3.6v and then upping the voltage again.... and so on.. its no good.
i do not know enough to find this mod but if anyone needs any readings or tests i will sodder wires and get numbers or anything anyone wants to get this done... also i have some vdroop issues.. with my vdimm @ 3.4 idle on a DMM and have my cpu above 2.8ghz or more.. on load my vdimm drops to 3.2-3.1 and my PSU stays solid as a rock.
Insane, that's cool to hear that you can get around it by increasing volts after boot. I thought about that but didn't know if it was gonna instantly fry something or whatnot...
As for your vdroop issue, if you solder 2 thick wires on backside of mobo from PSU 3.3 rail to dimm slots, your vdimm will always be very close to your PSU 3.3 rail and will only drop about as much as the PSU underload. My vdimm is always ~.05v less than actual PSU rail, and it drops .01-.02v under load :)
yes i see that mod and i think i will have to do that this weekend. im going to move my setup around and face the OCZ Powerstream pots towards me. and go to town.. i wish someone could really figure out this OVP thing though...
do you loose some stability or some cleanliness of the power if you bypass and feed the ram directly form the PSU and not through the mofsets and caps and such?