GAMECUBE COOLING INFO!!!!!!!!!!!!

Tchay

Tchay

Frequent Poster
Okay! The gamecube auto-shuts off at 270 degrees Fahrenheit = 132 degrees Celsius. I just tested it using this. The moment my multimeter hit 2.70 volts the gamecube shut off. This was tested on a REV A DOL-001 CPU 11. I will do more testing on other Rev A and C boards.

UPDATE: When the motherboard temp reaches 217ºF = 102.77ºC (2.17v on multimeter) the gamecube will shutoff (meaning the chips have reached 270ºF). The difference in temperature is because naturally the actual Gecko and Flipper will be MUCH hotter than the motherboard itself. So if you are placing your thermal sensor on the motherboard, then by the time the Gecko gets to 270ºF, the board itself will be 217ºF. Kapeesh?

Finally, I can begin testing to see which heating methods are the best. In this thread I will test the stock heatsink against numerous other types of heating methods like copper sheats, ram heatsinks, macbook heatsinks, and fan sucking versus blowing.

I strongly encourage serious modders to buy the LM34 to do testing of your own. And if any of you want to help me out and contribute, I'd appreciate. I'm doing this science fair style :cool:

For each test I will be placing the sensor directly below the Gecko processor. I will also put a make shift plastic cover over the whole board and heatsink method to simulate a portable case. Temperature will be recorded with and without fan usage.

STOCK GC HEATSINK
- After 15 minutes, no fan and covered with plastic shell - 135ºF = 57.22ºC not leveled out yet
- 25 minutes after using stock GC fan to suck air off aluminum - 110ºF = 43.33ºC leveled out temp


COPPER - 11 gauge .094” x 3.5” x 4.25”
- After 15 minutes, no fan and covered with plastic shell - 176ºF = 80ºC not leveled out yet
- After using stock GC fan to suck air off the copper - 142ºF = 61.11ºC leveled out temp
^ that really surprises me actually (in a good way).


THE CCUBE (Trimmed GC heatsink fins cut to 2mm protrusion (about 7mm thick off the chips) with tons of fans)
- 40 minutes in playing Metroid Prime 2 - 120ºF = 48.88ºC leveled out temp
:neutral2: I was always paranoid about the Ccube getting too hot. Guess I overreacted? Maybe I can sell it after all.


THINKPAD HEATSINK + FAN - credits to zenloc for showing this option (approximately 8.73125mm thick)
http://www.ebay.com/itm/05K5494-IBM-Coo ... 2ec4d1335b (credits to AteOeight for finding them on ebay)

Testing during Zelda: Twilight Princess gameplay. Note, the motherboard is OMGWTF trimmed. I tested this inside of my Gamecube Envision. The mobo is screwed into the case and I am using Arctic Silver 5 on all the chips.

Thinkpad heatsink (using 3.43v for fan)
131ºF - motherboard temperature (backside of CPU - Gekko)


Thinkpad heatsink (using 5v for fan)
126ºF - motherboard temperature (backside of CPU - Gekko)


Thinkpad heatsink (using 7.4v for fan)
120ºF - motherboard temperature (backside of CPU - Gekko)


For each of these I have to admit that the "leveled out temp" could increase after hours of use and in hotter climates.




Stickies:
Heatsink links + methods
 
Love you, let's make a custom heatsink :lol:
EDIT : a few seconds later... that is freaking hot :eek: a pc shuts off around 90°C
 
Σigma said:
Suggestion. Also try a stock CPU heatsink if you have one.
Click to expand...

Most stock CPU heatsinks are huge (according to ebay and google images) and will probably be BETTER than the GC heatsink. I'll see if I find a small one. But I'm interested in methods that are smaller than the stock GC heatsink. And of course anyone can simply buy the sensor chip and try it out for themselves ;)
 
Here you go. Bit pricey, but tiny. It's the kind I used on my PGPC. I thought my motherboard didn't come with one so I bought that one. Turned out it did come with one. It ended up being a good thing since my onboard GPU got really hot, so I put it on there.
 
Kyosho said:
Here you go. Bit pricey, but tiny. It's the kind I used on my PGPC. I thought my motherboard didn't come with one so I bought that one. Turned out it did come with one. It ended up being a good thing since my onboard GPU got really hot, so I put it on there.
Click to expand...

WAAAAAAY to thick, but thx for the suggestion. I can get heatsinks and fans for much cheaper (check out my sticky in the first post for links)
 
Good info you are collecting here Tchay and nice to see someone seriously putting in some time and crunching numbers on it. Once I get some real free time I may join you in your quest for cooling prowess.
 
Ashen said:
Good info you are collecting here Tchay and nice to see someone seriously putting in some time and crunching numbers on it. Once I get some real free time I may join you in your quest for cooling prowess.
Click to expand...

Please do! you would probably be way better at documentation and accuracy than me. :p
 
When you say directly below the Gecko I assume you mean on the other side of the board, yes? Or do you mean below as in beside it? Stupid question, but just thought I should double-check. I might try some stuff, but it'll be oldschool thermometer stylie.
 
Kyosho said:
When you say directly below the Gecko I assume you mean on the other side of the board, yes? Or do you mean below as in beside it? Stupid question, but just thought I should double-check. I might try some stuff, but it'll be oldschool thermometer stylie.
Click to expand...

Good question actually! I place the chip below as in beside it, same side of board as gecko. I also put a small dab of thermal clay under it and then squeeze it to the motherboard with a paperclip. Prolly not the best method but it seems fairly accurate.

Now, when I tested the shutoff temp, I had placed the chip directly on the Gecko. So for a true comparison, I need to see what the shutoff temp reads with the chip being beside the Gecko (since that is how I have been testing all the heatsinks - no room to put the temp chip right on the Gecko or Flipper).
 
:trollface:
Problem?

*I had to do that, but is it 750$ without the motor!?! If it is, that's utter insanity.
 
I'm bumping this because I plan to do some more testing pretty soon.

I fried my temp sensor chip because i was too lazy to lookup the pinouts :dah:

Gotta buy a new one. I'm going to reveal the "????" setup as its not much of a secret anymore.

And I'm also going to do testing for Zenloc's Thinkpad heatsink setup. After that, we will truly know what the best heatsink alternative is for the gamecube :D
 
Hey, I can contribute to this now!

IMG_7457.jpg


I've learned that using the inlaid fan method, the board actually gets hotter than the heatsink! :U That makes sense though, I guess. The board itself doesn't have a heatsink. Maybe it should? :dah: It makes me want to try to fit another fan in and see if I can cool down the board too, just for the heat of the unit in your hands.

Lets see some pictures and get some numbers, eh? I used Twilight Princess's title screen to get my numbers, after having played SSBM on it for about 30 minutes for gigglies.

First, the actual heatsink itself! The stock GC heatsink, shortened by Ashen, and then inlaid fan by ShockSlayer and his mighty all powerful oscillating dremel! Thermal paste is a thin layer of Arctic Silver 5, seems that I had just enough left over from the SLGC to do the job.

It's temperature?

heatsink.jpg


Max I got was about 90°F.

Oh, that? That's my laser infrared temperature gun, great for looking cool, tricking animals, and confounding your neighbors as you tell them exactly what temperature a 1 foot area is from 12 feet away! It is accurate to within 1.5 percent, displays temperature to 1/10th of a degree, and has a response time of less than 100 milliseconds.

In short?
It's good enough for the GodDang ShockSlayer.

Now, onto the other side of the board:

sweetspot.jpg


^This area(around the GPU) got the hottest, the highest temperature I was able to get out of it was around 137°F. Not bad really.

cpu.jpg


^The CPU was not as hot as the GPU, coming out at around 110°F.

ram.jpg


^And, as expected, the ram even less than that, the highest I got was about 90°F, but that was probably more due to the proximity of the CPU than to the RAM's actual temperature.

So, final writeup:

Trimmed Stock GC Heatsink with inlaid Tchay fan and Arctic Silver 5:

Heatsink: ~88ºF = 31.1ºC
GPU(board): ~135ºF = 57.2ºC
CPU(board): ~110ºF = 43.3ºC
RAM(board): ~90ºF = 32.2ºC

Any questions?

SS
 
Thats pretty Dang good! How thick in mm is that heatsink?

The setup I had tried a while back was like 5mm (as thick as the tchay fan). But even blasting the fan with voltage, I still was confident with the setup as the heatsink got pretty hot, bad sign.

But this setup seems great.

The stock heatsink gets the board to like 135 degrees F. So this one is close enough.

When I'm not feeling lazy, I'll add your info to the first post :dah:
 
You must log in or register to reply here.

Similar threads

Mr. Hooker
$400 GBA Project
Replies
0
Views
634
Mr. Hooker
Mr. Hooker
topherpharmer
Trouble Recording Inputs on ATmega328P
Replies
0
Views
748
topherpharmer
topherpharmer
S
Help GameCube portable
Replies
0
Views
843
SLOWPOKE
S
T
Any Gamecube controller's joystick schematics?
Replies
0
Views
2K
Tsukimaru
T
S
DOL-101 GCVideo (Semi)Working!
Replies
1
Views
4K
Sci_Karate
Sci_Karate
Back
Top