Possible P15xHM sandy bridge PLL pin mod overclock idea.

EDIT: If this mod works, someone who has a 2630QM would get a base clock of 2.66ghz instead of 2ghz. And would have a max turbo of 3.85ghz

I was looking at the service manual for the P150hm and I noticed on the block diagram (page 56) that there is a clock generator, it's SLG8SP585:
http://www.silego.com/uploads/Products/product_54/xSLG8SP585r101_10062009.pdf

Now look at pin 30 (CPU_SEL).

If I understand correctly, that pin should have logic 1 to it (100mhz reference clock).

If we ground the pin with a 10K resistor (logic 0) it should boost the reference clock to 133mhz.

The rest of the clocks should keep their original 100mhz ref clock.

So this pin mod should give a stable Bclk overclock.

If the pin mod is too hard for some people, the other option would be to open RWeverything, read the clock gen bytes. Then change bit 7 of byte 0 to 0.
Then maybe you can have 133 with a software mod.
But this is less likely to work from my experience.


Right now this thread is just for ideas, but at some point I will try this mod and maybe turn this thread into a guide.

I still don't have all of the parts I need to be able to put my P150HM together.

 

If this is the case, and it's possible.. then it should work in the newer P150EM's too....

I might try this with my P150EM if that is, as I'm considering buying a new CPU anyways

but good luck I'll report back if I find something

 

I took apart my np8150 today and spent 30mins looking for the PLL on the motherboard, I couldn't find it anywhere.

Anyone know where it is?

Also I noticed that the HM65 chip didn't have any cooling (no heatsink or anything).

When compared to my toshiba notebook, where the HM65 chip has a heatsink.

Shouldn't it have one?


I didn't find anything else interesting, but I was very frustrated that I couldn't find the PLL.

 

should be easy enough to install a low-profile passive heatsink on the chipset interesting thread btw,just saw it only now!

cheers

Sent from my Galaxy Nexus using Tapatalk 2

 

Yeah the passive heatsink idea might come to life.

I don't know what I did yesterday, but after putting my notebook back together, my temps were really bad.

I spent the rest of the night trying to see what the cause was, looks like my fans are spinning too slowly. FN +1 helps but it's still too hot.

The temps start out fine but wont stop climbing even past 95C.

Before I took it apart my temps would top out at 75-80C.
I don't know if that is because I had the P170HM bios and EC. But even now that I flashed back to the sager versions it's still not cool.

I am going to try re installing windows to see if that helps.

 

I thought the PLL was integrated into the CPU?

 

I thought so as well, but the service manual shows a PLL.

But I couldn't find it on the motherboard so I don't know why the PLL is mentioned in the service manual (even in the service manual for the P170HM).

I guess it really must be in the CPU.

Still can't figure out why my temps have suddenly gone so high.

 

did u take off any of ur heatsinks? check if the screws got loose or smth like that and repaste both cpu and gpu. i know these are the obvious things and u probably already checked them but cant hurt to ask

Sent from my Galaxy Nexus using Tapatalk 2

 

Yep checked the paste and the screws.

I think I found out what the problem was.

My fan speed with FN +1 is around 2.5K RPM.

If I connect the fan to 5V, then it's at around 3.5K RPM.

It looks like some bios + EC flash has ruined my fan speeds.

 

ouch, yeah, i read ur post in the other thread. think im gonna hold off from flashing the P170HM BIOS & EC onto my machine

 

Hmm guess I will stay on this BIOS for a while then now that everything is still working fine. I've flashed back and forth between P150HM and P170HM without any issues so far.. I do have to say that the fan policies that SAGER uses are better then the CLEVO one's, with the SAGER BIOS both my CPU and GPU were about 4° cooler when idle and under stress

 

sandy/ivy does not like high bclk...

 

Yes, I was well aware of that, but the reason it doesn't like a high bclk is because all of the other clocks were on the same PLL (sata and so on).
So normally when you mess with the bclk you also mess with the reference clock for the sata signal and other clocks as well (PCI).

But if this notebook had the PLL shown in the OP, then it could mean the clocks were not all using the same PLL.

That could have meant the PLL for the CPU was separate and if that was the case then a high bclk would be tolerated and the system should have been stable with a 33% overclock as long as there was enough voltage.

The problem was that the schematics were not accurate and the PLL is not on the motherboard. More info here:
Clevo P150HM (HM65) with XTU

Sadly it looks like we have hit a dead end.

 

too bad thx for trying tho mate! much appreciated

Sent from my Galaxy Nexus using Tapatalk 2