Printable View
here are some pics
http://img12.photobucket.com/albums/...k/DSCN1550.jpg
http://img12.photobucket.com/albums/...k/DSCN1549.jpg
http://img12.photobucket.com/albums/...k/DSCN1548.jpg
http://img12.photobucket.com/albums/...k/DSCN1547.jpg
http://img12.photobucket.com/albums/...k/DSCN1546.jpg
Composite:
http://img12.photobucket.com/albums/...composite2.jpg
it shoudl be similar to the 4pda2+, the voltage measuring points are similar
up up and away
Gonna start trying with this. Can you check the volt reading on points 1,2,3,4?
in a moment, lemme go grab my multimeter
1- 5.53
2- 3.13
3- 3.13
4- 3.31
heres another image
http://img12.photobucket.com/albums/...gereadings.jpg
hmmm...let me try again. :)
alrighty sounds good, thanks malves
Ok bro
Gimme the readings for these pins. Also, try to find the vdimm FET. Check the voltage on those FET's, and see which one matches the BIOS.
vdimm fet:
http://img12.photobucket.com/albums/...k/DSCN1569.jpg
voltage points you asked for:
http://img12.photobucket.com/albums/...pca3vdimm2.jpg
Ok, bro
Think I got it. Solder pin 6 to pin 11 (GND) via 200k trimmer. The 200k is just to be on the safe side. You can find the correct trimmer value. Just remove (unsolder) the trimmer and check the ohms on it, when the vdimm starts going up.
Please, do it at your own risk.
which pin of the trimmer is ground
middle or legs?
im pretty sure its the legs, but just wanna be absolutely sure
trimmer middle => pin 6
timmer front => pin 11
start resistance at 200k. lower resistance for more volts.
sounds good
200k trimmer is huge though :o
aslrihgt thanks for your help
I just like to be on the safe side. Besides, that's what i'd do.:)
hehe alright
i bet ill end up using a 10k trimmer
O_O!
time to start looking for a huge pot
Well, if you wanna take the risk i say a 50k trimmer will work on that.
10k shouldn't be a problem either, if the vdimm is fed by the 3.3v rail. You'll just have a higher increase on the voltage from default in BIOS.
will i need to vmod my 3.3 rail to say get 3.5 volts?
If the vdimm on this board is from the 3.3v, which i think it's, then yes.
Heya Malves & Solo,
I was dissapointed to see that this thread had no further discussion with no confirmation to the mod.
Anyway, I performed the mod on my brand new 4PCA3+ rev 2.2 board today. As instructed by Malves I connected a shunt POT (10k) BTW from pin 6 to 11. I had previously set my BIOS to 2.6V (vdimm) for safety. On POST I had 2.75V.
Heres the kicker, as I decreased resistance the voltage did indeed rise to top @ ~2.97V. By increasing the setting in the BIOS for voltage the voltage does not rise above ~2.97V or so. If I set the POT to correct for say 2.7V then raise the BIOS setting then it rises as expected as well, but not above 2.97V. SO I proceeded to measure the voltage myself thinking possible EPOX Magic Health would not read above 3V. I measured the V from Ground to the back of the mosfet identified by Solo and confirmed by me. Yet again the voltage will not go above 2.97V.
I use a very new Antec Tru Power 430Watt PS unit. Its 3.3V rail is solid and I would expect to top around 3.18V or so for vdimm. As my memory is GEIL PC4200 and my CPU is Prommy cooled I need that little gain for max FSB obviously.
Any thoughts as to getting more out of this mod or board for that matter?
Thx so much.
P4 2.4C M0 @ 3.5Ghz -60degrees F
2X256 GEIL PC4200 DDR good to 280Mhz FSB 1:1
Epox 4PCA3+ rev 2.2
WD Raptor
sorry tactical, i jut haven't had time
how low did you set resistance to?
in the 12/25 bios, it still had the 3.3vdimm option, and i think the max was around 3.11 volts
ill pick up a 10k pot soon, i only have a 1kpot right now
thanks for the update
Hey,
I set the resistance all the way down till the POT would not turn any more. It would seem then that 0 Ohms will give about 3V. I saw the options for up to +.7V in the release bios. I did not test to see if that BIOS actually did raise the voltage any. I will test that myself in the next few days possible tonight.
Your 1K pot set to 1K Ohms across the two used pins would be sufficient presuming you have some beefy memory such as Corsair XMS, or GEIL Ultra series that is rated for around 3V anyway. But then if you didnt have good memory you wouldnt need the mod :)
Keep me posted.
actually, ill test out that 1k pot then :)
bRB!
very interesting, i put the 1k trimmer @ 1kohms, and right when i booted (set at 2.6v) it was already at 3.09! :o
i was like uhhhhh, so i played around wiht bios options, being 2.7 and 2.8, but no change
i started adjusting the trimmer, and i dont get a change eitehr :o
That is very interesting. All you did I presume is connect that in parallel from pin 6 to 11 right. Also what rev is your board and what BIOS are you using?
yeah parallel
im running the latest bios, and rev 2.2
Im going to do some testing with the release BIOS over the next 15 min or so.
You might have interest in this thread as well. http://www.ocforums.com/showthread.php?threadid=266442
I need to get at least 3.1V on my memory. BTW what processor and what max FSB are you able to achieve? I can get 290 stable and 292 is not stable, anything above that wont POST. My spec is stated previously above in another POST also.
Ok,
I went back to the release BIOS which allows for 3.3V in the BIOS. At 3.3V with the mod turned down I only get 2.97V as read by the BIOS and 2.99 as read on the mosfet by multi-meter. Then at 3.3 in the BIOS with mod all the way on *0 Ohms I still only get 2.99 - 3.0V. I'm just stuck here.
I need some help with connecting the 3.3V rail directly somewhere to the mosfet or something then dropping it with some diodes or something down to get better vdimm. Perhaps Malves has some input there or hipro5 if he can be found.
lets PM malves again i guess
my mushkin black lvl2 will crash boot into windows at 250fsb at 3.09volts...
im running a 2.4c currrently at 3.3ghz stock cooling/volts (5:4)
Have you guys tried checking vdimm on other FETs?
no i haven't, but both pchealth stats, and the onboard monitoring program display it as 3.04, multimeter as 3.09
Tactical,
Can you take readings on your board for the point #3 shown in the pic on the first page of this thread post #3?
my #3 is now 3.11
Yes, gimme a few minutes
3.02V
Like I thought. The mod works, but something is limiting the voltage to the FET. That's why you got 3v and and solo about 3.09-3.1v. Gotta find out what.
I have spent several hours looking at this and have come to some conclusions. The mod we performed raised the emitter voltage on the second mosfet. The last mosfet emitter is tied to the 12V rail. The second mosfet with the mod defaults to about 5.5V or so on the emitter as controlled by the LM324. The first mosfet is also fed by the LM324 and has about 5.4V as well. With the mod activated and the POT at 0 Ohms it has about 11V or so on it. I believe that we need to mod the 3rd amp on the LM324 as well to achieve the remaining voltage. This pin would be pin 9 on the LM324.
I will test this within a few hours or so.
Whoops,
I was very tired last night when I was staring at my board and measuring. We need to mod pin 13 for the 4th amp in the LM324. I will do this now. Hopefully it will work to return to say what happened. If you guys dont hear from me from here then you know what heppened. :) rma baby
lol
ok,
Im still here, I connected pin 13 in parallel with the other mod to the same POT which did not let my system POST. I dont know if it wouldnt work in parallel like that 2 mods off 1 POT or just it wont work. I done for a little while. I have messed up two traces near my LM324 and I dont want to risk permanent damage till I see what you all have to say.
My system BTW is attached to a Prommy MachII and I have been soldering with it standing up vertically. LOL its a pain to turn it over but I guess I need to so I dont mess it up further.
I have a suggestion, but not sure if it will work. You have 4 options here:
1. Raise pin 13 of LM324 and solder it to that FET leg where the trace goes, via a 200 ohms trimmer set at 0 ohms. Raise resistance.
2. Cut the trace, scrape 2 spots on each side of the cut, and solder a 200ohms trimmer set at 0 omhs. Raise resistance.
3. Cut the trace, scrape one side of the cut facing the IC and solder a 200 ohms trimmer set at 0 ohms, to that spot and FET leg.
4. Cut the trace and solder the same trimmer to pin 13 and FET leg.
which FET leg is this
follow the trace from pin 13...same leg I just asked you guys to measure voltage.
will any of these kill my board?
i wiould like to attempt to cut the trace, but dont want to bother with an RMA
If you know how to cut the trace, you are fine. We are working on this mod, so something may go wrong.
Please, do it at your own risk.
i am going to attemp to do #4 suggestion...
so which one will be ground, pin 13 or FET leg
Solder middle leg to pin 13 and other leg to FET.
sounds good...
with a 1k trimmer i guess
im gonna unsolder it and reset it to 1k resistance
sounds good?
you have to start at 0 ohms. Also, it would be a good idea to have the vdimm mod done, so you can check if vdimm will go up, too.
oh, so i should leave the previous mod i just did on it?
yes, that's the vdimm mod. I'm thinking that the FET voltage is limited, that's why you guys can't go past 3.09 and 3 vdimm. What's weird is why yours is higher than his.
Before attempting this, could you check the readings on all legs on the vdimm FET?
when my dad leaves, i will be remodding my board and setting original trimmer back to 1k just in case
then i will procede with the pin13/fet mod
Bro, if you have msn, add me to your list. We need to talk.
Heya Malves,
I will attempt the mod most likely today, I would like to get your msn handle also. I guess your profile may have it. I will upload a pic of a diagram of the circuit so we can all look at it with certaintanty. I will mark the appropiate voltages as seen on my board with the vdimm mod in place as well. Also, what time zone are you in Malves and Solo?
Here is my schematic drawing of the circuit as it applies to the Epox 4PCA3+ Rev 2.2 as seen by me. If you find any errors let me know but this is it. I thought this may help us see what we need to do. Background in electronics may be needed to read it properly. As far as the 3 pin MOSFETS I have them all oriented the same way on the drawing.
Also if you were looking at a MOSFET with the two legs down and the tab up then pin 1 is left pin 2 is right and the middle pin or back tab is pin 3 per my drawing.
Image found here:
http://www.eitscorp.com/Tactical64/Board%20Layout.jpg
i just pm'ed you my msn.
yeahi just did the #4 idea
cut trace and soldered a 1k trimmer onto pin13 and FET leg
NO EFFECT WHATSOEVER :(
so i took the mod off and my board won't post
so i resoldered it on and everything works fine..except the vdimm is nonadjustable hehe
I am going to try the Northbridge volt mod today and see if I can get more Mhz or FSB with that. I have the vmod on hold. One thing is that for the rev 1.0 board the vmod was not on the LM324 but teh small IC to the upper left side of it with smaller pins. This chip is different on the rev 2.2 boards however it may be the chip we need to mod. I found this information on http://www.vr-zone.com/?i=242&s=1 and it perhaps sheds some light on our troubles.
i couldn't find anything similar on the rev2.2 boards that was used on the rev1.2 boards (
Not similiar how,? the part number is different and manufacturer different but the circuit design is basically the same. It cannot be coincedence that the chip we are modding was in place on the 1.1 boards and not being modded. the chip in question U36 was however modded. So now we have a replacement part with the same basic circuitry and we are modding the wrong chip i say by messing with the LM324. Its just a comparator amp.
Anyway, I think it should be looked into.
The northbridge mod from 1.65V to 2.6V made no discernable difference. However my VR was to small and I was worried to run it at those high volts for too long. I tried once to boot 296FSB 5:4 and froze on POST. My system will only boot win at 290FSB 5:4 currently. SO either it was to high volts and froze or just makes no difference yet till mem voltage is there.
2.6volts is a lot..
we'll see
id really like to push my ram :)
I SAY SCREW THIS $$%%##@ EPOX 4PCA3+ .
Ive got an Abiy AI7 in the closet. Ima bout to say later Epox. Screw Epox on P4 systems.
Someone make me a good offer on this fully working Epox 4PCA3+ mainboard. ITS YOURS. I need at least 80 dollars US for it. I accept Paypal or Money Order. I will use my Ebay Seller Rating for trustworthiness to the interested buyer.
My Ebay ID is Tactical64 check my ratings out.
I am a verified Premier Business Member on Paypal as well with over 50 transactions.
what a good price.
i have dimm-modded my 4pca3i rev2.0 here.