Easy Overclocking


Remember the higher you can overclock to the more voltage you are pumping into the CPU and the greater the chance for eventual failure. You would be better off only getting to 176 or 180. If you are getting 200 or more one of these days...POOF!
 
DaveC posted on Apr 23 2005 at 12:16 AM said:
Remember the higher you can overclock to the more voltage you are pumping into the CPU and the greater the chance for eventual failure.  You would be better off only getting to 176 or 180.  If you are getting 200 or more one of these days...POOF!
What, Our GP32`s will turn gay, Nice. :p

I`ll let you know when it does. :D

Trooper
 
Last edited by a moderator:
trooper posted on Apr 23 2005 at 12:46 AM said:
DaveC posted on Apr 23 2005 at 12:16 AM said:
Remember the higher you can overclock to the more voltage you are pumping into the CPU and the greater the chance for eventual failure. You would be better off only getting to 176 or 180. If you are getting 200 or more one of these days...POOF!
What, Our GP32`s will turn gay, Nice. :p

I`ll let you know when it does. :D

Trooper

:lol:
 
Last edited by a moderator:
The_Sheik posted on Apr 22 2005 at 07:51 PM said:
I feel a little newbie/stupid to ask this question, but how can I check it? I' ve got the object but I don' t know how to use it :D


If you read a few posts above, I found that point T7 is GND and T8 vcore. Check the voltage measuring this two points. :)

Greetings and good luck

P.S. Remember that anything above 1.95v is DANGEROUS even if it works fine. The GP32 CPU is 0.18 micron built, like Intel Pentium III 600-1000 MHz (Coppermine) and these models run at 1.7v. Too much voltage will cause electromigration kill your cpu.
 
Last edited by a moderator:
Hi Peeps,

Just a quick update, must be 4-5 days since I drew on my two resitors and I am still stable at 180mhz, must have put in 3-4 hours gaming in that time if not more cos im on my second set of batteries :)

Steve.
 
I tested the voltage with the following settings: wires and : "detector"

The result was quite amazing... 1.20 ! I don' t know if these settings are good, but if they are this could help people who don' t know how to check the voltage!
 
The_Sheik posted on Apr 23 2005 at 01:51 PM said:
I tested the voltage with the following settings: wires and : "detector"

The result was quite amazing... 1.20 ! I don' t know if these settings are good, but if they are this could help people who don' t know how to check the voltage!

Well I'm not sure that's correct.. I'm getting 1.43V there and I can OC up to 220 MHz stably
 
Last edited by a moderator:
day 5 of this trick for me, nothings changed. Still can OC to 180-184mhz stably, was playing OC'd versions of fSMS and fGB for a good 3 hours last night, as well as finally enjoying GP Engine at 166mhz and 60fps running full speed.

If the increased voltage is gonna kill it, it's certainly taking it's time.... Dave C seems to agree with me though... since I can 'only' OC to 184mhz before it will just start locking up, I must assume the voltage is not THAT high, I know there's still that possibility... but oh well... hey nothing lasts forever, if it breaks it breaks, but somehow I doubt it will.

Vimacs - we're not crazy, we're just brave and cheap. Hey if it weren't for guys like us no one would be OC'ing their GP's at all... I'm sure when people first started running theirs faster than 133mhz everyone called them crazy too....
 
I finally got around doing this mod, too, and now my 156mhz nlu is a 172mhz nlu! Hooray! :D :D
It crashes at 176 and higher, maybe some day I`m going to "paint" more out of this machine...
 
I didn' t used the batteries for the test, BTW how could I? It would be cool that someone make a picture or something to show how to check the tension! :)

Thanks to take the time to answer to my newbiness!
 
Hi everyone! Just checking in to see how things are going. I'm glad that this post is getting so much discussion, both Pro and Con. After all, that is what a fourum is for, right?

I'm exited that so many are finding and refining the techniques for this mod. Kudos to all the brave souls who are forging ahead to make this a safe alternative for the masses. At some point I will incorporate all the useful information into the top post so that others may find it all easily.

As for my self, I've been testing out all the advice being given, and pushing my GP32 to the limit. Still haven't gotten beyond 220Mhz, but mostly because I don't have a lot of time to fine tune.

I've been considering ways to make the mod more permanent, and I'm looking to mix ground-up graphite with Elmer's White Glue to essentially make a conductive paste that dries hard. It would still be removeable, but would surely stay in place best.

I guess I'll make several batches with varying ammounts of graphite and test each one out. The hard part will be waiting for the glue to dry :)
 
My guess is glue will screw it up...

edit: Oh and from reading up on graphite "resistors" on websites, what I would do is "draw" several "resistors" on paper, measure their resistances and use that data to calculate how the size of the graphite "resistor" relates to the resistance (the size matters more than how thick it is applied, just apply it really thick). Then once the info is accurate enough (you need to experiment several times) then I would figure out how much resistance I need to add and "draw" a "resistor" accordingly on paper. Then tape that to the real resistor. Its probably the most accurate way to work with graphite, other ways are just guesswork.

edit2: Also, pencils with harder "lead" (a mix of clay and graphite) have more clay, softer ones have more graphite. Meaning that softer pencils will (probably) give more resistance then harder ones, and this should be one of the manipulated variables in such an experiment (lead type, coat thickness?, "resister" size). The responding variable is the resistance of course.
 
ive just done that test by T8 and T7 , i found 2.14v in it(after trying to hold the batterys in and use test probes at same time), i dont know if thats bad or not but so far the gp seems fine, any suggestions ? :D


Steve
 
DistWave posted on Apr 23 2005 at 12:54 AM said:
...Remember that anything above 1.95v is DANGEROUS even if it works fine. The GP32 CPU is 0.18 micron built, like Intel Pentium III 600-1000 MHz (Coppermine) and these models run at 1.7v. Too much voltage will cause electromigration kill your cpu.
 
Last edited by a moderator:
sorry i dint read all the posts before but anyway...

this is(sounds like) a realy risky thing to do!!!! i have done the clockspeed test earlier today and found that my gp32 was stable at 236mhz which sounds like alot!!! is the clock speed test acurate??? cause if it is then im realy happy i guess :rolleyes:

btw i have a FLU. :unsure:
 
Back
Top