Help! Need advice on VDIMM mod for ECS RS480-M (Pictures inside)

[WiCKeD]

Well I decided I would try to do the VDIMM for this board, with the help of the respected modders on the forum. No fan of being helpless and I have been missing out on the warranty voiding fun. ;) The board currently goes up to a painful 2.7v, so it needs some hot, lead lovin'.

Attached pictures. Let me know what parts of the board anyone would need closeups on. You poke the board and I will get some high-res, naked shots of those special places. :cool:

Help me out, please! :D

http://img.photobucket.com/albums/v282/sleepeeg3/ECSOverall_small.jpg
http://img.photobucket.com/albums/v282/sleepeeg3/ECSTopmiddle_small.jpg
http://img.photobucket.com/albums/v282/sleepeeg3/ECSTopright_small.jpg

[persivore]

Can you get a picture of the chips and MOSFETs around the ATX connector area, and a picture of any chips/MOSFETs below the Dimm slots please.

[WiCKeD]

This should make things easier, I hope: http://img24.imageshack.us/gal.php?g=ecsoverall23bk.jpg

Would it help if I said I have a brand new 3700+ Sandy sitting here that is not going to be overclocked until I mod this board? :doh: :help:

[persivore]

Can you tell me what is written on the chips I have circled, and can you also check the voltage at the 2 MOSFET pins I have circled please.
See picture here (http://tom.drunken-student.co.uk/mods/ecsbottominnerright0ct.jpeg)

[WiCKeD]

The longer one reads:
AZ324M-A
429BT

The shorter one reads:
F753835
G501LAB
F152061

The voltage on the LOWER mosfet pin is 4.55v and the UPPPER mosfet pin is 2.62v (VDIMM? That would be the default).

:off: My San Diego is 1.4v stock. Damn. :off:

[persivore]

I think that the longer 14 pin chip is a LM324, with a differnet prefix to confuse people ;)
To confirm that, can you check the voltages at these (http://tom.drunken-student.co.uk/mods/ecsbottominnerright0ct1.jpeg) points.
Can you also check the resistance between these (http://tom.drunken-student.co.uk/mods/ecsbottominnerright0ct2.jpeg) points (resistance between pin 6 and ground (pin 11), and Vdimm read point and ground)

[WiCKeD]

LM324(?) voltages: http://img182.echo.cx/img182/7089/ecsbottominnerrightedit9ce.th.jpg (http://img182.echo.cx/my.php?image=ecsbottominnerrightedit9ce.jpg)

Various resistances: http://img286.echo.cx/img286/8172/ecsbottominnerright2edit4an.th.jpg (http://img286.echo.cx/my.php?image=ecsbottominnerright2edit4an.jpg)

Looks like I am doing something wrong, checking the resistance. This is with the setting @ 200 Ohms. Not getting anything from the other checkpoints at any setting. Do I need to test via the traces underneath? I'll figure out what I am doing wrong and report back.

[persivore]

From those voltage readings it looks like the chip is an LM324, so it is probably being used as the voltage regulator.
You might need to set the DMM to a higher resistance range to get a reading (possibly 2k or 20k).

[WiCKeD]

Actually it turns out it helps if you test resistance with the power off. :) Imagine that.

Here we go:
http://img198.echo.cx/img198/6414/ecsbottominnerright2edit4qj.th.jpg (http://img198.echo.cx/my.php?image=ecsbottominnerright2edit4qj.jpg)

[persivore]

For the Vdimm mod, connect a 50k variable resistor between pin6 and ground. Lower the resistance of the variable resistor for more voltage. It might be ok to use a 20k VR, but I would recomend trying with a 50k to start with.

[Malves]

pin 11 of LM324 (middle pin at the bottom row) can be used for ground, if you wish.

[WiCKeD]

Thank you for the help, persivore! I am off to Ratshack to get the components and will let you know how it goes. :toast:

Just to confirm, can you make sure I am looking at the right components:
http://img131.echo.cx/img131/9493/ecsbottominnerrightvdimm1mn.th.jpg (http://img131.echo.cx/my.php?image=ecsbottominnerrightvdimm1mn.jpg)

Guessing that's right, from what you said Malves. POE.

[persivore]

yea that looks ok :)

[WiCKeD]

Well guys, I don't know how to say this, but... it worked! Persivore you rock! :rocker: Appreciate the tip at the end as well, Malves. Using some liquid tape you recommended - great stuff.

Unfortunately, I found out a major issue with this board: Setting the VCORE above 1.45v does absolutely nothing! The bastards at ECS thought they would be slick and try to deceive average joe overclocker, which is who they seem to think is buying this board. Joe blows.

Not expecting to have to do anything else, but since I am already so far along, I might as well do what I have to. Hopefully ECS will send me a legitimate bios, but I'm not holding my breath.

So... what would it take for a VCORE mod on this board?

[Malves]

You may wanna start here.

[CaTalyst.X]

ouch...have fun with that one guys :D

-CaT

[WiCKeD]

ouch...have fun with that one guys :D

-CaTNot expecting this one to be easy, which sucks... I'll have pics up @ about 6-7 eastern, Malves (hey 1,000 posts).

Fugger must have known I was going to post. ;)

Anyway here we go:
http://img43.echo.cx/img43/457/ecstoprightxtremecloseup28vd.th.jpg (http://img43.echo.cx/my.php?image=ecstoprightxtremecloseup28vd.jpg)
Writing on that chip says:
ISL 6566
CR
N445F3

There are also all of these pics, that I posted earlier:
http://img24.imageshack.us/gal.php?g=ecsoverall23bk.jpg

[Malves]

Just as I thought, it is the ISL6566. Pin 9 to GND via 20k trimmer should work.

But the legs are too small, so you'll have to follow the trace from pin 9. The pic you posted, there's a cap right in front of pin 9, so I can't find it for you. Either post a top view pic or find it yourself.

Please, note: XS owners and members will not be liable for any damage caused to you or your hardware, neither can be liable for any breach of the warranty.

[WiCKeD]

Hehe, no worries, Malves, I am not going to fault anyone and last of all you, man! :p: I've killed far more expensive parts than this board anyway. ;)

Will probably have to dismantle the case to get a pic, which is a major ordeal. Can you see to point out pin 9? Maybe I can figure it out myself.

Edit: Nevermind. Numbers around the corners of the chip. Kind of obvious. Not sure about the traces though, so I'll get a picture up in a sec.

Just so happens I bought a 20k trimmer for the VDIMM, so things are looking up :up:.

[Malves]

I know, bro. That disclaimer is not just for you.;)

[WiCKeD]

Right on. Well here is the best I can do right now:
http://img57.echo.cx/img57/6608/isl65666xz.th.jpg (http://img57.echo.cx/my.php?image=isl65666xz.jpg)
Good pic, probably bad angle. Will be back after I dismantle the case, because I will have to anyway. :shrug:

[Malves]

Here's the point I think goes to pin 9. Persivore may have a way to check it.

[WiCKeD]

Sort of what it looked like to me, but it also looks like pin 9 is going nowhere and both of those traces lead away from 10. Obviously, it's going somewhere, but even up close, I can't tell. That line of paint is not helping... Trying to get a better pic with the camera, because it can pick up a lot higher res than my eye can pick up, but it's not cooperating.

Anyway, can I just use the local screw hole for ground?

Edit: 20 min of high intensity fluorescent and a magnifying glass and I am still not sure:
http://img211.echo.cx/img211/936/isl6566xtremecloseup6nt.th.jpg (http://img211.echo.cx/my.php?image=isl6566xtremecloseup6nt.jpg)

Red is where the traces look like they are headed, purple offshoot looks like a maybe, green X is where I am soldering. Will have to wait until tomorrow though, because I'm wrecked.

[Malves]

Yup.

[persivore]

If you need to find an alternative solder point for the Vcore mod, put one multimeter probe on pin 9 of the Vcore regulator, and probe around the SMD's in the surrounding area. If A point is connected to pin 9, there will be a resistance of <2ohms between the points. Just use that point as an alernative solder point for the mod :)

[WiCKeD]

Just wanted to give you guys an update, since you might be wondering whether I killed my board yet. :) Close.

Basically, it looks like soldering a 20k trimmer to the new green X I marked will let you bump the VCORE a fixed .1v. So you can get the full 1.55v that the board was supposed to provide from the start.
http://img195.echo.cx/img195/3812/isl6566xtremecloseupedit2an.th.jpg (http://img195.echo.cx/my.php?image=isl6566xtremecloseupedit2an.jpg)

Long version:
I measured 1.45v initially, all along the line from pin 9. Tried to get resistance measurements from the start of this mod, but the el cheapo multimeters only occasionally want to work. Grounding them in a molex eventually was the only thing that was working for me when I was testing the VDIMM mod, but eventually, I could not pull any resistances for the VCORE mod. Possible pilot error, but I think it was just these POS tools I am using.

Spent a gawdawful amount of time trying to solder to the end of that resistor(?), because it was a millimeter from the base of that cap. After an extraordinarily long time, I got some solder on the resistor next to it, tried to pick it off of it and ripped the resistor in half! Putting a glob of solder in place of it at least seems to have temporarily fixed it (red spot). Damaging that resistor seems to have dropped my VCORE ~.025v. Beautiful.

At that point, I gave up and tried attaching the trimmer to the new X. At max, it gave me a fixed .1v.

Currently, either the trimmer died, came loose, was set too low, or something else entirely, but I had to remove it, before the system would post. Think the problem was the backplate came loose and was shorting out the board, but right now I am back at square one and will probably try to resolder to that point.

So those were my "let's see what happens" results.

Do yourself a favor and read this:
http://www.extremeoverclocking.com/articles/guides/Soldering_Guide_1.html