Easy Overclocking


imageshack?

if not, could you email it to me at Vaulter98c@gmail.com

then, I could email it out to others , and maybe image shack it for you

thx, and either wy if you could email it to me, that would be cool cause then I could print it out today in an hour ( not trying to be pushy, but if you could email it to me within an hour and a half that would be great, but I understand if you can't )

~Octavious
 
adzg posted on Apr 18 2005 at 09:44 AM said:
I have applied the "pencil" technique to countless numbers of athlon processors over the years with 100% success.

However; I am a bit dubious about trying this on my GP32, I think replacing a fried processor in a GP32 might be slightly harder than replacing an athlon chip on a mobo!!

Adz.

The Athlon pencil trick has nothing to do with this. By closing the L2 bridges on the Athlon you could unlock the multiplier menu in your PC´s BIOS.

What jcabrer did is lowering the resistance of the R52 resistor by creating a bypass. But this method will in 90% of all cases either not work or kill the CPU of the GP32 since you cannot tell how high the core- voltage of the GP32 CPU is without switching it on. That can be deadly for the gp32_console .
And replacing the GP32´s CPU could only be done with another GP32 CPU (if you have the skills ;) ). You can´t take just any ARM9 processor, since a processor ins´t just hardware...

jcabrer posted on Apr 18 2005 at 01:58 PM said:
If anyone can host the image, I've got one.  Is there a way to upload one to the forum???
http://www.imageshack.us
 
Last edited by a moderator:
No whizz, Bang or pops, The screen just faded away like it does when you switch it off or the batteries die.

I left it off for a couple of seconds incase of heat, Then switched it back on the red light on the top of the GP lit up, But there was no response from the screen (stays blank with no light).

I`ve just been given an idea from a friend that i should have a go at removing the pencil from the resistor and see if she powers up. I can not try that at the moment as i am at work, But will try it when i get home this evening and post what happens.

Fingers crossed. :p

Trooper
 
trooper posted on Apr 18 2005 at 07:21 AM said:
No whizz, Bang or pops, The screen just faded away like it does when you switch it off or the batteries die.

I left it off for a couple of seconds incase of heat, Then switched it back on the red light on the top of the GP lit up, But there was no response from the screen (stays blank with no light).

I`ve just been given an idea from a friend that i should have a go at removing the pencil from the resistor and see if she powers up. I can not try that at the moment as i am at work, But will try it when i get home this evening and post what happens.

Fingers crossed. :p

Trooper

Oddly enough, the same thing happened to me when I first tried to overclock! Turns out it drained my batteries which were not fully charged :) I hope this is the case. Good luck.
 
Last edited by a moderator:
R52loc.jpg

http://nils.thiele.free.fr/R52Loc.jpg
here it is...
 
I'd say it wasn't the overclocking speed that killed it. Overclocking the processor past it's specified value will normally just make it do daft things like jump to an incorrect address in memory, or calculate a value wrong (which normally ends up causing the device to crash). Removing power and letting it settle will normally bring it back to life.

However, what is guaranteed to kill it is too much heat and overvoltage - apply too much and regardless of speed, it'll eventually pop. It could last for an hour or so before going pop. You can normally tell its going to do so however as it'll get very hot first. Different batches of processors will naturally be able to withstand different amounts of voltages (and of course, different maximum speeds at those voltages).

What I usually do is carry around a can of freezer spray - spray the processor, switch it on and see how fast it melts. Instant = bad :) It also allows you to see what speeds you can get upto if you can cool the processor down better.

I wonder if we could get a heatsink fitted - Nice slab of copper with a bit of artic silver thermal adhesive on it could do quite well.
 
jcabrer posted on Apr 18 2005 at 03:30 PM said:
trooper posted on Apr 18 2005 at 07:21 AM said:
No whizz, Bang or pops, The screen just faded away like it does when you switch it off or the batteries die.

I left it off for a couple of seconds incase of heat, Then switched it back on the red light on the top of the GP lit up, But there was no response from the screen (stays blank with no light).

I`ve just been given an idea from a friend that i should have a go at removing the pencil from the resistor and see if she powers up. I can not try that at the moment as i am at work, But will try it when i get home this evening and post what happens.

Fingers crossed. :p

Trooper

Oddly enough, the same thing happened to me when I first tried to overclock! Turns out it drained my batteries which were not fully charged :) I hope this is the case. Good luck.

The thing is, I wasn`t running it off of batteries, I was using a mains adaptor. This is why i posted my warning of the risks from doing this mod as i was pretty sure the GP was dead. But hopefully (fingers crossed), By removing the pencil from the resistor, It may allow the GP to power up. There could be a chance (although very slim) that the GP could be currently locked at 224mhz (this has happened to someone here before) so not allowing it to power up. Well it`s worth a try anyway.

Trooper
 
Last edited by a moderator:
SO

all I need to do is
1. Open back of case
2. Take my #2 pencil and draw on resistor r52
3. Cover unit
4. Try OC'ing to 176, then try up to 200-210

correct?

~Octavious
 
I'd rephrase that.

1. Open back of case
2. Take a #2 pencil and draw on resistor r52
3. Check resistance with an ohm meter and verify that it's still within limits.
4. Turn on device, and after a few seconds, place back of finger onto processor to check temperature of processor. If it bites you, turn it off before you kill it.
5. Overclock to each required setting and carry out step 4 each time every few minutes.
6. Once happy, Cover unit

If you don't have an ohm meter, then step 4 will just have to do on it's own, but leave it a couple of minutes each time between switch on and overclocking to test temperature.
 
the bad thing about this type of situation is that it needs alot of testing. And there probably aren't that many willing to do it. Once i get my flu, i can try this on my nlu, but I was hoping to sell it and not kill it.

How long have any of you that have tried this been able to run past 166mhz and not killing the GP32?
 
schaap2005 posted on Apr 18 2005 at 05:00 PM said:
Sorry for you trooper :(
How many dead gp32's do you have?
And what was the dumbest action why a gp32 died?

I now own 2 dead GP32`s, Both my fault. Well, You live and don`t learn. :D
It`s all for the good of GP kind. :p

Least i still have a spare working Blu, But for how long. :rolleyes:

Trooper
 
Last edited by a moderator:
Squidge posted on Apr 18 2005 at 02:44 PM said:
However, what is guaranteed to kill it is too much heat and overvoltage - apply too much and regardless of speed, it'll eventually pop. It could last for an hour or so before going pop. You can normally tell its going to do so however as it'll get very hot first. Different batches of processors will naturally be able to withstand different amounts of voltages (and of course, different maximum speeds at those voltages).
Too much heat would kill your ARM9. But it is impossible that the puny voltage of 2 1.5V Batteries heat your CPU up so much. There´s gotta be another reason.
 
Last edited by a moderator:
You guys that are doing this are nuts! Is it worth a few more MHz to take the chance of frying your unit? I can't believe some of you are actually doing this or are even considering it :rolleyes: Even with an ohm meter you are taking a chance as the reading will be inaccurate measuring in-circuit resistance. Also even if you don't go over 200 MHz it won't matter. Once you increase the voltage it will be increased no matter how high you OC to. You will still puff the CPU it just may take a few minutes longer.


If you really want to volt mod the unit do it with resistors of known values as on the Cobbleware site. Anything else is madness!
 
Back
Top