How to pinmod and further overclock the M11x R1 (SLG= Hardmod only)

Yep

I can't get above 550 or 275 FSB while in OC mod so I'll need a volt mod.

In non OC mode I can get 740 or 370 FSB but my ram is now the issue being that it's PC3-8500. I'm half tempted to get some bigger memory modules and faster ram. I though someone already tried this and it didn't work. I'll need to reread or search where I read this from. Did you guys figure out if the ram with run with higher CAS timings?

Newegg.com - Kingston HyperX 8GB (2 x 4GB) 204-Pin DDR3 SO-DIMM DDR3 1600 (PC3 12800) Laptop Memory Model KHX1600C9S3K2/8GX

 

Still, can you give more info on that one? On ICS we need to solter between ground and pin 10k resistor. And what should we solter on FSLx mod?

 

See Overclocking SLG8SP556 equipped M11xR1 using FSLx PLL pinmod.

 

Just wanted to jump on here and say I did this mod like a friggin' neanderthal, worried that I had just screwed up my laptop forever, but it did work. I used a bulletin board pushpin to essentially smush the proper pin down, then pulled up a bit to permanently disable it. (Stupidly, I didn't get any pictures.)

Once I reassembled and got back into Windows, I was able to use SetFSB (free version) as described in this thread to overclock. Unfortunately I wasn't able to get it stable at anything beyond about 1.9GHz. My VID in overclock mode is 0.9v. Attached is a screenshot (in CPU-Z and SetFSB) of the highest overclock I can get without fiddling with anything else. This is the RAM that I'm using.

I suspect the FSB is too high, possibly causing the northbridge or the RAM to fail, but I've no idea how to change the memory divider, and suspiciously, the 2GB of cheaper RAM that came in this laptop fails at the exact same overclock point (0Eh byte changing to 7Dh, overclocking the CPU to ~1950MHz). The failure usually comes really, really fast when I go over, either with an instant total lockup or applications starting crashing immediately, so that leads me to believe it's the RAM, but if both sets of sticks were crashing at the same place, that makes it seem like it's something else that's failing. Ideas?

edit: added a second screenshot of the memory tab in CPU-Z

 

Try turning off Overclock mode, it should let you relax the FSB abit and push up the RAM. We cannot change the RAM divider- we either get 1:2 from OC mode off or 2:3 with OC mode on.

If that doesn't work you may need to consider a VID mod, which will give you 1.1v through the use of a conductive pen.

Otherwise, write it off as a CPU limit. What is your VID with Overclock mode off?

I tried not to give info on something not tested but Nando4 already posted it.

370 FSB? Pics or it didn't happen.

I went through using Thaiphoon Burner but it didn't work with anything higher then CAS6 timings on boot. You could try Thaiphoon Burner and we can confirm if it's the memory or the motherboard not accepting post CAS6 timings.
Try the 4GB 1333 kit instead, it has CAS6 timings for 1142mhz and has been tested to boot fine by DR650SE.

 

Right

I'm not actually 370 stable as under intel burn test it would fail after numerous cycles. I also confirmed this when I tried playing bad company 2. I ended up backing it down to 360 and haven't had it crash yet.


I'll see about getting some benchmarks up later, but I have to get back to studying as I've been procrastinating and need to get back to my Calc II homework.

Pictures of the 360 at least. I'll probably end up doing the volt mod next week friday when I'm done with class work for the week.



Uploaded with ImageShack.us

 

im scared i might not be able to confirm the SLG FSLx mod.

i had my M11x downloading steam updates when i left for work today. i came home just now and it was froze... it was on Intel gfx and in OC mode. not other tweaks or anything.

im scared that the stock OC is failing... that would really F$%&ing suck! its hard to believe that this chip cant do at least a 300mhz OC. it could be my HyperX ram as well but idk at this point. this is kinda odd cuz before work today i made at least 12 3Dmark06 runs trying to find a stable 335M overclock (which seems to be around 600/1440/875). anyway...

i guess i am probably going to need to do some kind of volt mod. to even be able to test the FSLx mod. cuz if im not stable on the stock OC i sure as hell wont be much help seeing if the FSLx mod even works.

do you guys know how i would do this? i read on how ppl were doing it on the ICS chip but not the SLG.

and also i asked DavyGT if i need to lift any pins for the SLG FSLx mod and he said no but when i went back to read the instructions i swear it said i did need to disconnect pin 7 from the circuit. please advise.


thanks.

edit:

just ran intel burn 2.5 at the defalt settings and passed. temps never went over 52c. in fact they stayed at 50c 90% of the time.

 

I got those freezes as well at stock CPU and GPU, in "random" circumstances- listening to music, in a multiplayer game, heavy multitasking and even idling. I tried stability tests that were extreme in my description and couldn't repeat it- Dell's preboot diagnostics test, 100 runs of maximum stress IntelBurnTest and 72 hours of Prime95, nothing. It seems to be fixed with a new motherboard...

A VID mod is possbile, it does not require soldering and involves connecting 3 adjacent pins together. I better check with my friends' SLG to see if it features the same voltage regulator.

Don't use 3DMark06 as a 3D stability test, it is too lax. Try OCCT GPU test or ATiTool, they both feature an error scanner so you know if you get artifacts. Plus, OCCt has a VRAM memory test, for testing graphics memory stability.

I do not understand what do you mean by "lifting" pins so I just explained it the literal way here and here. What I said and what Nando4 instructs are the same in practice however- solder a 10k resistor to PIN 7 then route it to 3.3v (pin 12).

 

so in your opinion should i get a mobo replacement from dell?

thanks for the stability tip.

here is nando4's quote:

Then disconnect the FS_C pin7 from the circuit and route it to a 3.3V point (eg: pin6 of the PLL provides 3.3V) via a 10k resistor. Best done if there is a resistor attached to pin7 which you can lift a leg on.

am i not understanding something? sorry not trying to argue. it just seems nando4 is saying to disconnect pin 7 from the mobo and soldier a 10k resistor to it and then soldier the other side of the 10k resistor to a 3.3v pin (pin 6 / pin 12).

 

Ok, I made a mistake. According to Nando4, you disconnect pin 7 from the circuit to connect to your own.

I just tried out a M11x with a 1.0125 stock VID, it doesn't seem to be able to reach 1.9Ghz stable. This chip is .05v higher in stock VID then my M11x (.9625v). I have been trying to collect as many stable overclocks with the stock VID but I haven't had any luck. (I define stable as 20 runs of IntelBurnTest, maximum stress.)

If VID could be used to "grade" chips, then the FSlx modders will know if they will likely need a VID mod or not depending on their stock VID. But I don't want to make any judgement yet, as correlation may not be causation.

 

DavyGT, what do you mean by stock VID? It is important to note that the bios OC mode is effectively doing a 200->266Mhz BSEL mod on the CPU. As you'd be aware, Intel chipsets compare the CPU BSEL set and the chipset BSEL sampled and seeing they are different, the chipset then locks the CPU to the lowest VID and lowest multiplier. I believe the lowest VID is 0.9V on the SU7300. The locked VID means need a VID mod to get more overclock headroom.

<strike> @some guy - if you find there is already a highish resistor value of say 2k-10k leading off from pin7 (FS_C) then can just run a parallel 33ohm resistor to 3.3V. That would create a voltage divider pulling the signal at FS_C high to get your 333Mhz PLL mode when in bios-OC mode.</strike>

Striked above info out. This won't work as the chipset will be given a 333Mhz BCLK but only supports 200 and 266Mhz, so will shutdown. Must actually separate the chipset and PLL signals. Disconnect the FS_C signal by cutting the track or lifting a resistor leg of a track that runs to it.

 

I mean the VID before Overclock mode is applied.

 

hello everyone... i'ved read the previous posts and seen the images, and im guessing this is just for SU7300 and not for SU4100? sorry for the noob question.. i really wanna try this out but i'ved never seen any SU4100s... thanks!

 

The SU4100 can do this mod, it just happens that the SU7300 configurations doing this mod are more common.

 

Thanks! i think i will do this mod this weekend with pictures and hopefully will turn out ok... in the mean time, i will back read again just to make sure i understand it well. any pointers would much appreciated... thanks again, DavvyGt

 

Everything is pretty much covered in the original post. I guess the only real pointer I can give is that you read and do this part of the OP now: "How can I tell which PLL I have?" to avoid hyping up the easymod all through the week if you have an SLG. We're in a situation where it looks like SLG have a majority of the R1 PLLs .

 

@ nando4 & DavyGT

thanks for clearing that up. i plan on trying this mod possibly later this week. i actually want to do it tonight but I've. been trying to see if my system will freeze up on me again before i get myself into opening up this lil guy.

im not trying to mod a defective board.

 

i disassembled my m11x just find and take a look whether i solder or pry it.. its too small to solder it without touching other pins! i think i would go to pry procedure. 1 question, do i have to purchase/donate to setfsb in order to unlock the TME? tia!

 

Since you didn't do any mods, you still have warranty. Try and get it to freeze again, it's a matter of time and application. I seen a thread about this and it blames the hard drive or wireless card, but I'm rock solid since replacing my board. I'll have a look for that thread here.
UPDATE: Found one.

You do not have to purchase SetFSB to unlock TME. We use SetFSB to program the PLL to overclock our M11x.
There are 2 versions of SetFSB: free and shareware. Both versions can do the same job, however, the shareware version of SetFSB is much more convenient to use as a) it's possible to have SetFSB apply the overclock automatically on boot and b) the PLL is officially supported, meaning we can use the sliders.

In saying that, that means you have an ICS PLL right?

We need more ICS PLL users to help the VID experiment, where we investigate if there is a relationship between stock, unoverclocked VID and the limit of their overclock.

 

You're not on the Intel 6200 wireless card are you? As soon as i switched the cards, i had nothing but random crashes. I switched it back to the dell card as of yesterday morning and have been crash free for 24hrs now.

 

i honestly don't know. i just have the stock WiFi card in there.

I've been trying to leave it on as much as possible to see if it will freeze up again on me. its at home now downloading steam games. just the way it was the last time it locked up on me.

really sucks cuz i want to do this mod pretty bad but as I said before i want to rule out a bad board.

 

Hmm, I just paid to extend my warranty (due to my faulty screen hinges before. So I'm thinking of holding of on this. I may do it anyway though, because I know my system operates fine, the only issue I forsee is a screen replacement eentually. That said I'm on the fence of cutting the PLL pin. In all honesty, the hinge is the only issue I've ever had with the system. Never any freezing or anything.

 

They've just announced that hinge failures will be handled outside of warranty and that there is in fact a re-designed LCD assembly that is supposed to fix the design/materials defects in the original.

 

To discuss the hinge related update from Chris, move to this thread -

http://forum.notebookreview.com/ali...eased-update-hinge-issue-m11x-01-24-11-a.html

Back on topic please

 

surprisingly yes i have ICS PLL. i am very tempted to do the mod as soon as possible, i cant sleep thinking about it. heheheh, but i need read more before i proceed.

 

so it works!

im not stable at all tho. at first i had crazy issues with windows starting the video driver correctly. it wouldn't recognize the 335gt at all.

reinstalled the driver and restarted, came back into windows and all seemed fine so i ran 3dmark06. was getting like 1/3 more fps till i crashed...

not sure if i need more Vcore or what.

oh, is my ram really running 1500mhz? or am i just dumb?

 

Congratulations, you are the first SLG modder.

What is your VID at stock, before the mod if you can remember? Don't try to find out now though by turning Overclock mode off, as you won't be able to boot.

I think you need a VID mod.

Your RAM is running at 500mhz, which is 1000mhz effective.

 

thanks! could not have done it with out you and nando4

so i tried reading up on the volt mod but i get lost as to what chip im actually looking for.

my vid in cpu-z was 0.975 before (in non OC mode). its still the same in OC mode as before.


just FYI this isn't considered 100% working. plz don't try this until i get the volt mod to see if the funky GPU issues are just driver/unstable CPU related. right now my laptop is basically useless until i volt mod or unsolder the mod i just did.

 

Here is where the VID modding stuff begins in the old thread.

Can you check if you have the same voltage regulator?

 

ISL6261
ACRZ
H013BDM

plz tell me its the same as yours?

 

Yes, it is the same.

Since we both need a VID mod, I should write up this mod. Can you write up your SLG FSLx mod?

EDIT: The quick VID involves connecting pins VID 3 and 4 together, either with wire or conductive pen. This will give 1.1v with OC mode on.

 

nice!

ill be more than happy to do a lil write up. or you can it doesnt matter to me. i took some pics to you can use them for any thing you need.

brb... must volt mod!

 

Well done.
All hale some guy.

All us SLG guys can now feel a bit better about ourselves.

Maybe DaveGT can modify the thread title to reflect the new status of the SLG support.

 

I tried to edit the title. It doesn't work so I'll have to contact a mod when I'm done with the VID mod.

I think I caused some misunderstanding in my VID mod- you are supposed to connect pins 31 and 32 together (VID 3 and 4 in datasheet) NOT pins 3 and 4. :S

 

SLG here...

Bah.

 

No luck and no boot with both my VID mod attempts.

 

You may be shorting the board connecting VID4+VID3. The board would detect that and fail to boot.

Instead try to connect VID4 to GND via a 33ohm resistor (or say < 100ohm) to get a 1.1V overvolt. If want 1.0V instead then also connect VID3 to 3.3V via a 33ohm resistor. Those are strong pull-down/pull-ups which assume those pins are connected back to a resistor network with weak pull-up/pull-down resistors.

To be absolutely sure that VID4/VID3 is getting the logic you want would mean isolating them from the rest of the circuit and then running a 10k to GND/3.3V respectively.

 

Then that means Maddi2k could not have done the VID mod properly, remaining at .9v.

It would also explain why Tassadar98 can't do his VID mod.

 

VID4 to 33ohm to gnd seems to be working but not detected in cpu-z

going to stress test brb.



well. was going fast as hell then froze up on me. i rebooted just for it to retun back into stupid mode where it doent see the 335gt.....

well i gotta be up in 4 hrs so i guess this is a loss for now....

 

CPU-Z reports the VID pins that the CPU sets which is transmitted to the voltage regulator. SO it will still say 0.9V. WHat is actually happening is you are hijaaking that VID logic so instead of the voltage regulator sending 0.9V to the CPU it sends 1.1V.

 

ok cool. well in my case I think itit may be the chipset voltage that may be the issue. maybe that's why I'm having issues with getting the 335gt to switch.

just a theory...

is there any higher vcore I can set to try to prove this?

 

sorry for double post....

ive been trying to use some logic with the data sheets...

to get 1.2v would i ground Vid5 through a 33ohm resistor and run Vid3 to a 3.3v through a 10k resistor?

FYI: i saw no raise in temps after the 1.1vmod but i would say that cuz i replaced the TIM. under Intel Burn i saw temps 1-3c lower than i did when testing on the stock voltage. i think this chip can handle some more juice. but im not sure if this will make me stable. as im afraid it may be a RAM or Chipset voltage issue. i guess we will see....

regarding the FSLx mod:

im using PIN19 for my 3.3v. is this ok? or do i need to only use PIN6/PIN12? i used PIN19 just because it was easier to get to being on the very end and all.

thanks again for any help.

 

Last time I tried a hard mod, I botched it big time... I won't ever try it again. But if I could find someone that lives near me I would probably pay them $40 to do it for me.

 

i consider myself pretty good at this kind of stuff and i have to say this is the hardest mod I've ever done.

 

It could very well mean that running VID4 to GND via a 33ohm resistor was insufficient. To be 100% sure that have a logic 0 on the VID4 pin would need you to isolate VID4 (disconnect if from the rest of the circuit) and then run it to GND via a 10k resistor. I'd expect temps to go up with a 0.2V increase in Vcore.

Otherwise you may be seeing 2.0Ghz overclocked instability due to RAM.

 

trying this out now.

hoping im not hitting a RAM wall. as i don't think there is any better ram to buy... and going down to 300mhz FSB isnt worth it IMO. thats to many PINs to solder for 10% increase...

 

You shouldn't be hitting a RAM wall- you are well within limits of the RAM, at CAS6, 1000mhz. Try setting the latencies to 6-6-6-15 using MemSet, as I see you have set 6-5-5-12 timings in a previous CPU Z screenshot.

 

Well, the question is: why did my maximum cpu-clock raised from 1800 to 2100 mhz? Definitely just bridged both pins via a small cable. I also feeling that the left side of my m11x is warmer than before. Any idea why this worked for me ?

maddi

 

im throwing n the towel.... 1.1v mod didn't help.

i never been so frustrated with a mod before.

having to "lift" these pins is near impossible with out destroying the pin. ive probably come close to bricking my less than a week old latop 10 times.

i even made a super tiny soldering tip for this mod but its just to f-ing tiny to keep trying these different configs of soldering with out destroying the chips them selves. ive even went as far as trying to use a single strand of wire to hit theses small points on the chips and then link them them to the resistor...

i guess i would be in a better mood if i had a stable 2.0ghz laptop right now but all i got i a voided warranty and the same laptop i had 4 days ago

i hope the work i put in can help someone else with the SLG chip but i feel any further messing around with mine may result in a brick. maybe if i had a super awesome soldering station or something id try all the configs i could but i know im just going to break something...

IDK.. sorry for being such a downer but this mod is no walk in the park for the SLG. if all i had to do was tap the top of these pins with a tinned resistor id be in heaven but this is just too hard for me to continue trying different stuff.

now... i hope i have enough material to work with so i can get this thing back to stock...

 

Some guy, So you havn't successfully OCed the SLG systme?