NotebookForums.com › Forums › Notebook Manufacturers › Dell Forums › Dell Home (Inspiron, XPS, Studio) › Some Info That might lead to overclocking the 533MHz native notebooks
New Posts  All Forums:Forum Nav:

Some Info That might lead to overclocking the 533MHz native notebooks - Page 2

post #21 of 87
1/20/06 at 2:35pm
_
Quote:
Originally Posted by KraxKill
I have too many un answered questions about this whole thing.

1st-I can't figure out if the pins in question are the right ones
2nd-Do I need to bridge pin 53 and 16 or will pin 16 to pin 1 work? Does it need 3.3v or just ground.
3rd-There is a guy on this fourm that tried to mod a different pll http://www.notebookforums.com/showth...&page=21&pp=15 with not success that I know of however the fact that someone else tried make me more hopefull. I sent' him a PM to try to see if it worked on his and had no reply. Anyone else have any ideas.

Think damn it think!!!!
Looking at the block diagram, if FS_B = FS_A = 1 should change the output CPUfreq from 133 to 166. The question is how well the system react to the change. If you're not sure how to make the pin mod or have the right tools, you might want to bring it to an electronic shop/use their tools soldering/magnifying glass, etc.. Using a multi-meter you can check if the FS_B and FS_A are the same.
What's the chip/PLL ID on yours/what make you think that the chip is the same/compatible with the pdf that you provided? Thanks.
Reply
Reply
post #22 of 87
1/20/06 at 3:14pm
Thread Starter 
I pulled the cover off and checked (that's how I know that is the chip) part number and pin configuration are the same. I also provided 2 pdfs one is the Pericom for which i'm sure is in our systems, the other pdf is for the ics corossreference which I suspect is compatible. Like I said before, pin 16 (FS_B) is terminated if you look closely as in there is no lead to it. The leg is down and a lead runs about an eights of an inch from it and terminates.
When I tried to mod it, I ran pin 16 to pin 1 as it was easier to secure a wire around pin 1 which is 3.3v as well and on the edge, however I don't know if there is some sort of special signal that is needed that is not provided on pin one and is on pin (FS_A) which is infact powered already (not tested just going off of the diagram). When I used the pin 1 to pin 16 method, i did get the computer to not boot 1 time, however i'm not sure if it was the wire or the possibility of not having everything plugged in just right. I was running out of time and knew that connecting pin16 to pin 53 would be nearly impossible given my current tools. I do have a soldering iron a magnifing glass and a volt meter that will messure just about everything. I think I may even have a working ossilascope but the thing is so tini really someone has to look at it to see.

Here are the pdfs again

pericom
PI6C410M
http://www.pericom.com/pdf/datasheets/PI6C410M.pdf

ICS954201
http://www.icst.com/datasheets/ics954201.pdf

P.S fixed link to thread above.
Reply
Reply
post #23 of 87
1/20/06 at 3:37pm
Quote:
Originally Posted by KraxKill
I pulled the cover off and checked (that's how I know that is the chip) part number and pin configuration are the same. I also provided 2 pdfs one is the Pericom for which i'm sure is in our systems, the other pdf is for the ics corossreference which I suspect is compatible. Like I said before, pin 16 (FS_B) is terminated if you look closely as in there is no lead to it. The leg is down and a lead runs about an eights of an inch from it and terminates.
When I tried to mod it, I ran pin 16 to pin 1 as it was easier to secure a wire around pin 1 which is 3.3v as well and on the edge, however I don't know if there is some sort of special signal that is needed that is not provided on pin one and is on pin (FS_A) which is infact powered already (not tested just going off of the diagram). When I used the pin 1 to pin 16 method, i did get the computer to not boot 1 time, however i'm not sure if it was the wire or the possibility of not having everything plugged in just right. I was running out of time and knew that connecting pin16 to pin 53 would be nearly impossible given my current tools. I do have a soldering iron a magnifing glass and a volt meter that will messure just about everything. I think I may even have a working ossilascope but the thing is so tini really someone has to look at it to see.

Here are the pdfs again

pericom
PI6C410M
http://www.pericom.com/pdf/datasheets/PI6C410M.pdf

ICS954201
http://www.icst.com/datasheets/ics954201.pdf

P.S fixed link to thread above.
How did you "secure" the wire to the pin? If not soldered, if might not work, unless you use your omh meter to confirm that connection is good. Soldering pin 16 is about the same the degree of difficulty as any other non-cornered pins. Why would you want to connect pin 16 w/ pin 53 (FS_C) instead of FS_A?

1. FS_C is a three-level input. Please see VIL_FS and VIH_FS specifications in
the Input/Supply/Common Output Parameters Table for correct values.
Also refer to the Test Clarification Table.
2. FS_B and FS_A are low-threshold inputs. Please see the VIL_FS and VIH_FS
specifications in the Input/Supply/Common Output Parameters Table for correct values.

Much thanks.
Reply
Reply
post #24 of 87
1/20/06 at 3:42pm
I took a look at the data sheet. The FS_C and FS_B pins are either tied to ground or open. You need to know which before modding it. With everything turned off, you can check the FS_B pin with an Ohmeter to ground. If it's tied to ground, then you need to disconnect this pin from the circuit board. If you don't, you're basically just dropping the voltage straight to ground. I would tie pin 12 to pin 16 after pin 16 is disconnected from the circuit board. That ties FS_A to FS_B. The reason I'd do it this way is just in case there is some kind of delay from getting the core chip voltage and getting the clock frequency input voltages.

Edit:
After looking at the power up timing diagram, I think you definitely need to tie pin 12 to 16. This will probably disable the test mode off of pin 16. I don't know if this will affect how the chip runs.
Reply
Reply
post #25 of 87
1/20/06 at 3:47pm
Thread Starter 
My bad you are right. I meant connect pin 16 to pin 12(FS_A)

Does anyone want to try a staple i'm at work today a cant.
LL
LL
Reply
Reply
post #26 of 87
1/20/06 at 3:54pm
By looking at that frequency selection table on page 3 of the pericom
PI6C410M pdf, it looks like you need to open FS_B to get the 166 which would mean lifting pin 16. Since at 133 your already set to 001, it looks like 166 is 011, meaning open for pin 16(FS_b). Just speculating based on the what I've seen and could be completely wrong.
Reply
Reply
post #27 of 87
1/20/06 at 3:55pm
Thread Starter 
Thanks for the info ShortBus i'm at work right now but what you said makes sense. Pin 16 seems to not go anywhere thats the thing it just hits the board and stops after short distance.
Reply
Reply
post #28 of 87
1/20/06 at 3:57pm
Quote:
Originally Posted by KraxKill
Thanks for the info ShortBus i'm at work right now but what you said makes sense. Pin 16 seems to not go anywhere thats the thing it just hits the board and stops after short distance.
Does it terminate in a little eyelet? If so, that's just a test point. It's possible that the lead continues on the other side of the circuit board or through the circuit board if it's a multilayer board.
Reply
Reply
post #29 of 87
1/20/06 at 3:59pm
Thread Starter 
Now i'm confused does this mean that all we might need to do is lift pin 16 and be done with it or is there still a reason that we would need to connect pin 16 to pin 12?
Reply
Reply
post #30 of 87
1/20/06 at 4:00pm
You still need to connect pin 12 to pin 16 to supply the 3.3 VDC.
Reply
Reply
post #31 of 87
1/20/06 at 4:00pm
Quote:
Originally Posted by ShortBus
The reason I'd do it this way is just in case there is some kind of delay from getting the core chip voltage and getting the clock frequency input voltages.

Edit:
After looking at the power up timing diagram, I think you definitely need to tie pin 12 to 16. This will probably disable the test mode off of pin 16. I don't know if this will affect how the chip runs.
Do you think there's that much of a delay to have to tie pin 12 and 16 together? All these timing are in the nanosecond range, right? Do you think there'd be a noticeable delay or are you talking an electronic delay that would cause this not to work?

Also, what about voltage for the system? Would the ram still run at 533 while the FSB was at 667? Would there be enough voltage for the cpu?
Reply
Reply
post #32 of 87
1/20/06 at 4:02pm
Quote:
Originally Posted by ShortBus
You still need to connect pin 12 to pin 16 to supply the 3.3 VDC.
Answered while I was typing. What will the 3.3v supply? Is that for the FSB or the chipset or what?
Reply
Reply
post #33 of 87
1/20/06 at 4:05pm
Quote:
Originally Posted by ShortBus
I took a look at the data sheet. The FS_C and FS_B pins are either tied to ground or open. You need to know which before modding it. With everything turned off, you can check the FS_B pin with an Ohmeter to ground. If it's tied to ground, then you need to disconnect this pin from the circuit board. If you don't, you're basically just dropping the voltage straight to ground. I would tie pin 12 to pin 16 after pin 16 is disconnected from the circuit board. That ties FS_A to FS_B. The reason I'd do it this way is just in case there is some kind of delay from getting the core chip voltage and getting the clock frequency input voltages.

Edit:
After looking at the power up timing diagram, I think you definitely need to tie pin 12 to 16. This will probably disable the test mode off of pin 16. I don't know if this will affect how the chip runs.
KraxKill meant FS_B is open above, but again need to confirm w/ an omh meter.

From the Pericom datasheet:
Note: 1. Maximum 10" trace length for CPU @ 200 MHz, 16" trace for SRC @ 100 MHz.

Sounds like sufficient thoughts and design/consideration to get this chip to output to 200Mhz in the realworld.

KraxKill, you didn't answer how you secure the wire to the pin? Is your "stable" comment meant you use the paper stable non-soldered?
Much thanks.
Reply
Reply
post #34 of 87
1/20/06 at 4:07pm
After looking at the timing diagrams, the IC expects the FS_A, FS_B, and FS_C pins to be either 0 or 1 at a particular point in time. From looking at the diagrams, this is in the millisecond range and I would expect late timing to make a difference.

For example, in the power up timing diagram, the IC doesn't expect the VCC core voltage to come on until about 0.2 milliseconds after pins FS_A and FS_B are already set.

As for the system voltage, I don't know how to answer your question. Tying pin FS_B (16) to 3.3 volts isn't going to hog up much power, but the chip itself probably will.
Reply
Reply
post #35 of 87
1/20/06 at 4:09pm
Quote:
Originally Posted by flick
Answered while I was typing. What will the 3.3v supply? Is that for the FSB or the chipset or what?
The 3.3 Volts will supply the FS_B pin. Doing this will make this input a logic 1 for the circuit, therefore making it output at 166MHz.
Reply
Reply
post #36 of 87
1/20/06 at 4:58pm
Quote:
Originally Posted by ShortBus
The 3.3 Volts will supply the FS_B pin. Doing this will make this input a logic 1 for the circuit, therefore making it output at 166MHz.
Doh, that was actually a dumb question on my part. I should have known that already. Thanks for the answer.
Reply
Reply
post #37 of 87
1/20/06 at 8:23pm
Thread Starter 
No i didn't permanantly secure the clip as I wanted to make sure that the thing works first so I used a small wire tini even. But the thing is too small for my hands to work with. If this is even possible the project would take some doing
Reply
Reply
post #38 of 87
1/20/06 at 8:32pm
Remember the good old days of changing jumpers and no heat sinks?

-Doc
Reply
Reply
post #39 of 87
1/21/06 at 4:42pm
ok, I'm still reading this thread, but i wanna get this outta my head:

He DID connect something with a tiny wire, he used a new driver, he changed the BIOS, and ALSO he is using 667mhz RAM. (Said it can't work without the RAM)
Reply
Reply
post #40 of 87
1/21/06 at 5:03pm
Thread Starter 
Well i did get the wire connected, however, I cant get the thing to boot. Looks like a new bios is in order here or upgraded ram. I don't konw this is as far as I want to tinker with this thing for now. I don't want to go upgrading ram on a hunch. And if I do, why not get a faster chip in my case I only have a 740 at 1.73.
Reply
Reply
New Posts  All Forums:Forum Nav:
  Return Home
NotebookForums.com › Forums › Notebook Manufacturers › Dell Forums › Dell Home (Inspiron, XPS, Studio) › Some Info That might lead to overclocking the 533MHz native notebooks