somthing like this right? I just need to find the right pin(s).
Printable View
somthing like this right? I just need to find the right pin(s).
I found this thread with info on the Vdimm=Vio mod, but I read here that it won't let you go above 3.6v either :(
can you draw the diagram on this picture plz?
http://img25.exs.cx/img25/4882/DFI32.jpg
Untill we figure out how to turn off the OVP I suppose for testing my lovely BH-5 at 4V and such I can do the same trick as people do with OCZ booster and raise the 3.3v after the board posts.
Holy crap! There is a OVP issue with these boards? I had no idea. This might explain why I have so many cold boot issues, I have the bios set to 3.8V and the 3.3 line at 4V.
It does eventually work after quite a few restarts and removal of ram sticks though. Usually I need to reset the bios through the jumper too. Hmm, this makes me want to go try my benching rig again.
Your board wont post if you set your vdimm to 3.7v or higher in bios if your 3.3V rail is high enough to actually give it 3.7v (i.e. 3.3v = 3.90, vdimm= 3.7). If your 3.3v is only at say 3.6V and you set bios to 3.7v it will post and only be giving you 3.45-3.5
For the Vdimm=vio mod can you check if the point I have marked on the ATX connector is a 3.3v pin, and if one of the legs on the capacitor I have marked reads Vdimm. If what I have marked is correct, connect the 3.3v ATX pin to the pin on the capacitor which reads Vdimm.
edit:forgot to attatch the picture!
http://drunken-student.co.uk/tom/mods/DFI32.jpg
ok well the bottom one reads 2.546 (I assume thats what its gonna read without posting and without a dimm in the socket, and they all read the same)
would the first one work? Or would the second one be a better idea? (i figure more than one line is better either way)
I checked the other cap's near the dimms and there was another one that reads the same...its a 1000uf 6.3V cap instead of 1000uf 10v like the other caps...so maybe each orange wire could goto 2 caps?
Quote:
Originally Posted by Tony420
This seems conflicting to me. Will it boot at 3.7 normally or not?Quote:
Originally Posted by Tony420
not conflicting at all...if you had the board you would know what im talking about. Say you dont have a modded 3.3V line but you install the 4V bios. Nothing will stop you from selecting 3.7V and you will still POST...but your actuall vdimm will only be 3.1-3.2...now if you kept vdimm set to 3.7V and you put 3.3V at 4V you wouldnt be able to post...get it?
also try not quoting segments that make it look like it makes no sense. You obviously didnt read it clearlyQuote:
Originally Posted by isp
if somone says "I like AMD cpu's but I dont like AMD cpu's prices" do you cut and paste what somone says as contradiction with...HEY YOU SAID "I like AMD cpu's " and then you said "I dont like AMD cpu's " gee that sounds conflicting! :slap:
It looks like those are the right capacitors. It would be best to just use one wire to one cap to start with, but if the mod works as you want it to you can add the extra cables later.
ok just to see if it works only use one? and if it does I can add them like the last picture? You think I should bother with the 1000uf 6.3V cap if I do?
Just check it works with one cable, there is much less to go worng. If it works ok, then add more cables. Using 3 cables should be able to provide enough current for the RAM.
ok that sounds good. So pretty much it dont matter where I send the 3 cables then...
It shouldn't matter which connections you use.
Tony420
Calm down buddy, we aren't here to tear eachother down or anything of the sort. I just didn't understand it is all. Thanks for clarifying it for me. I think I know what you're saying now though, and that is; you can't tell the BIOS a certain vdimm voltage and then raise your 3.3 above it and expect to post. Your 3.3 has to be below whatever you choose in the BIOS if you want to post. Is this right?
persivore, couldnt there be a resistor that is part of the vdimm circuit that could limit the voltage to 3.6V? Maybe you can remove this resistor and solder across the 2 points it was on to unlock the full 4V?
-CaT
Keep up the good work guys!
This is interesting.
I've been looking for something on the board which could be providing the OVP so we can work out which resistors are being used to provide the OVP reference voltage, and change their value. I think that it should be being done using a comparator (either a dedicated chip or made from an opamp) comparing the reference voltage to the Vdimm, and triggering OVP accordingly. I've been looking for it around the Dimm slots because that is the most logical place for it to be, but it could be anywhere on the board tbh. The Vdimm looks like its produced using an opamp regulator below the Dimm slots, but I'm not 100% sure. If the Vdimm is being regulated by a dedicated regulator, and not an opamp then it might be easier to mod the OVP, but there doesn't appear to be one on the board.Quote:
Originally Posted by CaTalyst.X
I've had a reply from DFI tech support, but it appears that they don't support the 4v Vdimm BIOS :(
Well, for my benching system the Vdimm isn't what was causing the problems. Even with the 3.3 rail dropped to 3.5V and the Vdimm down at 3.2V it still doesn't like cold booting.
This is rather off topic, but I didn't want to leave my post just out there and have people wonder, "Does his system work or not?"
I'm almost sure that my 3700 and LBUT are trashed, but I like trying random stabs at making it work agian.
Do u think cutting the trace leading to the FB leg of LM358, soldering a 50K resistor from FB to GND could fix the OVP?
-CaT
I don't know tbh. It depends if the LM358 is providing the OVP as well as regulating the voltage.Quote:
Originally Posted by CaTalyst.X
Ill give it a whirl tomorrow :P
-CaT