Here it's
Printable View
Here it's
Conguratulations, Malves;-)
> BTW, what's VTT and why should I worry about it?
--
VTT is Bus Termination Voltage.
It's ideal that Vtt=Vddr/2, in my recognition.
I could check it in BIOS hardware status monitor of IC7.
See the attached images, please:
It shows Vddr=3.23V and VTT=1.37V, so VTT is lower than
Vddr/2, but it seems that small difference doesn't make
troubles. It may causes a little decrease of maximum memory
clock, I suspect.
Thanks
Hey Kyosen,
I did the Vnorth mod on my IS7. But the voltage won't go up. I have everything setup just like yours, except for the ground where I picked a screw hole. Maybe this doesn't work on the IS7.
What happens is. At 0ohm pin 6 of LM358 reads 1.4v and right FET leg reads 1.4v. When I increase resistance on the pot, pin 6 reads less voltage - up to 0v, but FET leg keeps at 1.4v constant.
HiPro looking at this diagram, does that mean that the rightmostQuote:
pad is NOT already tied to ground - that it is tied elsewhere?
If the right pad is already at ground, then all this would do is
put a trimpot in parallel with the 10k resistor :confused:
Right most pad is actuall VDD and not ground. Left most pad goes to winbond chip.
Hello Malves,
> I did the Vnorth mod on my IS7. But the voltage won't go up.
> I have everything setup just like yours, except for the ground
> where I picked a screw hole. Maybe this doesn't work on the IS7.
--
Hmm...
Can you check the resistance/connection between Pin#7 of
LM358 and left(=Gate) leg of FET?
If it is 0ohm, I have no idea.
If it is not 0ohm, Pin#6 may not be a target pin.
Find the LM358's pin which shows 0ohm with left(=Gate) leg of FET, please.
LM358 has 2 operational amplifiers. See the attached image.
If the found pin is Pin#1, the target may be Pin#2.
Anyway, I don't have IS7, so please check and try at your own risk.
Regards
If you mean the right pad of the R50 resistance,then NO........It's not grounded.......It's the Vdimm(Vddr)......and it's NOT 10Kohms but 10ohms...... ;)Quote:
Originally posted by dexx
HiPro looking at this diagram, does that mean that the rightmost
pad is NOT already tied to ground - that it is tied elsewhere?
If the right pad is already at ground, then all this would do is
put a trimpot in parallel with the 10k resistor :confused:
Hey Kyosen,
I tried pin 6 down, trimmer grounded to FET leg. I started at 1k and less resistance gave me more voltage. But I end up frying my board. Something went wrong and my guess is because when checking for the voltage on the FET leg, the ground was removed by the multimeter lead. Solder wasn't strong enough.
Well, time for an IC7. :)
Hello Malves,
> But I end up frying my board.
--
Oh, me too;-)
Now I'm playing with AI7.
AI7's circuit looks more simple than IC7.
Vnorth of AI7 may be common with Vagp, I suspect...
Thanks
how did u fry it? too high vnorth? if so how high did u go. i wanna know whats safe
Safe is not trying to do the Vnorth mod. ;)
Kyosen,
I'd like to get an AI7, but I'm getting a Prommie hopefuly by the end of month and that 45ยบ angle socket may be a problem.
> how did u fry it? too high vnorth? if so how high did u go. i wanna know whats safe
--
Yellow wire came off by accident...Poor soldering caused it.
Method itself has no problem on IC7, I believe.
Maybe using some hot glue, after soldering on the FET leg, is a good idea. :)
I'm not to sure how much abuse the IS7 can take but the IC7 can take a ton of abuse... I did the VDimm mod and it was working for a good while... I had some DT moments while I was doing the mod and got some solder on the winbond chips right legs... Well I cleaned it off but not good enough... I lost all my temps and voltage reading fromt he MB... Then when doing some other stuff in my case I got my desolder braid out and tried to clean off the winbond chip some more... this time My board wouldn't post now... I thought I was screwed... At one point in the 5 hours I was messing with it (the end btw) My board powered up for a split second and then wouldn't power up any more... After messing with it for another hour or 2 I figured I fried something... I had given my Mem like 3.38volts at one time while I was messing with it... So I was all set on getting me a AI7 to replace my IC7....
A few days later I said to myself.. Hell it's already broke so lets really abuse this thing and see what happens... I couldn't find or smell anything that had fried anywhere on the board... I took my scribe out and scraped the hell out of the winbond chip.... Ended up even exposeing a lot of copper traces in the process... I resoldered it all back up and low and behold.... It POSTS again... Actually had some temps and voltages for a bit but it was all really wacky like... My CPU was reading like 1.35 when I had set it at 1.675.... They disappeard all together again after tring to clean some more solder away from the winbond chip...
Now I have my system stable at 288 FSB... My 2.4C D1 will not go anyhigher... I'm giving my RAM 3.15v as measured on the left mosefet leg... The only quircky thing now is that it won't post on inital boot.... I have to either power off and on again or reset it to get it to post... Also if I change anything in the BIOS I have to reset it to POST.... I'm just amazed at how much punishment I've given this thing and that it's working again... Thought I would share my results... Oh and My memory is at 2-2-2-5 DDR460... Now I can wait for PIC-Express and DDR2 before I get another board.... So if you think your board is toast give it hell.... Maybe you can raze it from the dead....
Yup my semi dead ic7 i decided to use for VDD/VTT/Vnorth mods. And low and behold it's running better that 99% of them out there. I can't believe this old arse mobo kicks so much arse. Luck sometimes does help. :D
I am a nOOb when it comes to voltmodding & want to vDimm mod my IC&-G mobo.
I went to my loacal electronics engineer & gave him the blueprints of hipro5's method of how to mod the IC7-G with all relevent pics of the mod.
The thing is, my soldering guy has done the mod a little different from the methods described in this forum.
Instead of solering the 10 Ohm fixed resistor to both of the "R50" pads, he's actually solder only one of the fied resistors to the "R50"pad closest to the "Winbond Chip" & the other wire to left leg of the mosfet closest to the "R50" closest to the "Mosfet".
I've borrowed "johnrr6's" schem2 (hipro5 method 2) & edited it to show what my electronic tech guy has done.
I haven't fired my board up yet as I would like to ask one of the senior members if the way I described is possible. Too afraid that I might blow up my mobo if I turn my PC on & my method is wrong.
hipro5, I have been following your method (that's how I wanted it done in the first place) but my tech guy says that the "R50" pad closest to the "Mosfet" actually goes from the left leg of the "Mosfet" goes to the "R50" pad closest to it so it makes no difference whether I solder onto the "R50" pad closest to the ?Mosfet" or if I solder directly onto the left leg of the "Mosfet".
Could one of my peers please clarify this for me? I'm kinda shaky as my mod did not turn out like how hipro5 described.
Cheers:toast:
I think he has just saved himself some trouble and moving the end of the 10k resistor from the end of R50 too the mosfet for ease of soldering.
You should be able to check this with a multimeter if your unsure, set it to continuity beep and check that there is a direct connection between the end of R50 and the mosfet pin he has soldered too.
jeez fried my ic7 as well with vnorth mod...Quote:
Originally posted by kyosen
Hello Malves,
> But I end up frying my board.
--
Oh, me too;-)
Now I'm playing with AI7.
AI7's circuit looks more simple than IC7.
Vnorth of AI7 may be common with Vagp, I suspect...
Thanks
everything was fine but i suspect altering vnorth 'on the fly' caused my mobo to crack...
Holst,
The mod works perfectly. When set to 2.8 I get 3.24 (PSU max). In a way, it seems simpler to do it the way I posted up b4.
Got my 9800pro & IC7-G modded, now for the Vapo to come back in action. hehehe
Thanks for clarifying it for me.
edit:
I noticed that after I vmodded my 9800pro, if I set the V-AGP to any higher than 1.55, I get these artifacts when my PC posts. A reboot fixes it but it's kinda annoying. I'm only using one of those Arctic Cooling VGA Silencers til I have time to set my WC again.
My next mission? VTT mod. My IC7-G behaves very much like my old Max3. 3.1v vDimm, 1.35 VTT.
Anyways, if it wasn't for you guys at XS, I woulda went out & wasted money on another Max3 & hope that I'd get 3.1 vDimm.
Many Thanks To The XS Guys:toast:
Hmm one question !
Why does VAGP increase in BIOS increases overclock capabilities on my IC7 ??
I don't notice any Vnorth change by doing that (increasing AGP from BIOS) !
Yup, keeping VTT at 50% or Vddr is a must on certain boards. Its a must on IC7 too for high RAM clocks. Using only Vddr mod at 3.3V on IC7 my system would reboot and give bluescreens even at 240 1:1. With the Vtt mod its SuperPi 1M stable at 255 1:1 2-2-2-6 now :)Quote:
It shows Vddr=3.23V and VTT=1.37V, so VTT is lower than
Vddr/2, but it seems that small difference doesn't make
troubles. It may causes a little decrease of maximum memory
clock, I suspect.
Also on Gigabyte K8N (uses winbond voltcontroller too) the Vtt (Vbt) mod is required for running 3DMark2001.
It would pass superpi, 3dmark03 and pifast at 260+ 1:1 (w/ 3.5V Vddr) without the Vbt mod but 3DMark2001 would always quit to desktop!? Funny board eh? :D
Vddr and Vbt voltmods enabled (@Vddr 3.52V/Vbt@1.76) it can pass 3DMark2001 at 327FSB:268MEM 2-2-2-6
if i do the mod below will i have to do another mod for vtt to be 1/2 of vdimm or should it be stable as is?
http://www.frontpage2000.family-net....is/Schem1a.jpg
Vdimm is gonna be stable and high...Quote:
Originally posted by shortbus
if i do the mod below will i have to do another mod for vtt to be 1/2 of vdimm or should it be stable as is?
http://www.frontpage2000.family-net....is/Schem1a.jpg
Vtt will not be altered by doing this
This pic shows my IC7(#2) Vddr & Vtt(Vbt) mods. It says all;)
Very nice 'all in one' IC7 Vmem guide :)
I've been pushing my HyperX today again on IC7 and can run Superpi 1M at 257.1MHz 2-2-2-6. This is with Vddr@3.52V/Vtt@1.76V/Vagp@1.79V/PSU 3.3V@3.6V.
I removed the stock NB cooler and installed bigger heatsink with delta fan on it. Also removed all the original heattransfer stuff and replaced it with normal white paste. Open-window mod gave some 10C ambient temps so the NB was rather cool.
The same RAM has done 258-259 on P4C800 which had only Vdimm mod at 3.46V.
Any ideas on how to make the IC7 go even higher? :)
Vnorth mod? Something else?
Where is hipro with his special mods huh? :D