PS3 Fan Fix V4.2

KillerBug, I was just watching a youtube video of a fan modification where the fan stays on even after you turned the PS3 off, is this a good idea for the slow cool down? Does yours do that as well, he sells the diagrams for $2 I think in case you wanted to see it.

 

Here's a link http://www.youtube.com/watch?v=VuH__HDo0B0

 

He sells diagrams for $2? I suddenly feel like a chump for giving away my diagrams (and full assembly instructions) and selling PCBs for $4.

No, mine does not stay on after the system. You do not want this. YLODs are caused by solder cracking, and solder cracking has two main factors involved: the amount of heating and cooling, and the speed of heating and cooling. If the fan stays on after the CPU and GPU stop producing heat, then both will cool very quickly...certainly not a good thing.

 

ANY 1 HELP WHEN I TURN PS3 YELLOW LIGHT COMES ON SOLID NOTHIN HAPPENS

 

In another post you give advise to floppie90 to leave the PS3 in the XMB for a while so it can cool down before turning it off, doesn't that mean that the fan will still be going at whatever speed you have it set for and isn't that the same thing the guy is doing in the video? Or will the heat be less because the PS3 is no longer playing a game or movie and the heat can be reduced in a slower pace and if we do the same thing that the guy did in the video, the PS3 is no longer producing any heat but is being cooled like if it was and thus cooling it down faster, something we don't want to do?

 

Yeah...you get the idea I think...as long as the system is on, the chips are making heat...if you idle at the menu, the chips are making less heat than full throttle gaming, but they are still making a good deal of heat (temperatures actually go down a few C while the games are loading; the XMB uses THAT much power). Depending on how fast you set the fan, the difference between gaming temperature and menu temperature can be very small or rather large. The faster the fan spins, the smaller the variance.

 

thank you very musch for this technical info

it was very helpful i wouldn manage it:|

 

Why so sophisticate? It's enough to cut 3.3V wire, and connect it to PWM fan wire through rheostat and resistor. I am using 4.7kOm potentiometr and 27+34 kOm resistors and have very smooth speed control.

 

I have trouble believing that the fan control is smooth...that is how this project initially started, and such controls always had a spot that was about 1/20 of a turn where all the adjustment took place. On top of that, it did not work with most models.

 

Well, if you don't believe words - watch the video.
http://www.youtube.com/watch?v=qmwogM5CRQ0
As you can see there is no 1/20 spot of adjustment - it controls the speed on almost every point of rotation. I can't say about slim models, but I'm sure it would work with all the FAT models. Difference only in location of 3.3V wire. And I think slim models have not very different way of controlling fan speed. Maybe it just need another value of tuning resistors.

 

That does seem to work well...the top speed it limited and there doesn't seem to be a minimum, but that isn't a big deal I guess. Is your fan a Delta?

I know for a 100% fact that your method would not work with the slim because the 3.0V wire does not have the current for the job...I tried it without a pot and the fan didn't even spin.

 

Top speed is limited by maximum speed 3.3V could do (I mean, even if there would be no resistor and pot, maximum speed would be the same), and the absolute minimum is stop of the fan. That is because controlling of the fan speed is located approximately between 60 (maximum) and 65 (fullstop) kOm, that's why I'm using resistors, and that's why I get smooth controlling.
What do you mean by Delta?

Hmm.. If the fan didn't even spin when PWM wire is plugged to that voltage, maybe there is really another way of controlling speed.

The thing is, using resistor and pot is waaaay much easier than using microcontroller.
Maybe if elaborate on slim's fan controlling system, there will be too easy method of controlling speed.

 

The problem with doing what you describe with the slim is that the 3.3v wire has so little amperage on it that it actually drops to 3.0v just from triggering the power supply to turn on.

When I asked about the Delta fan, I was asking who makes the fan; Delta makes most PS3 fans (including mine) but some of the 20GB and 60GB models got other brands of fans that are more friendly to work with.

I think I know what you are doing now; you found the threshold between 0 and 1, and you are setting the voltage so that it sits in that threshold...is that about right? If it is, it is a nice idea...assuming that it works cross-platform.

By cross-platform, I mean there are at least 4 different major revisions of fan on the fat and at least one on the slim (not counting the new model). Among each revision are multiple part numbers, and even amongst the same part number, the 0/1 threshold is probably going to be a bit different between fans.

Still, a neat option that I think will work well for some people. I'm even half-tempted to offer something like what you describe, but with a trimmer instead of the two resistors, so it can be dialed in. It wouldn't be as easy to use as the PS3 E-Z Chill Hidden Control Model, but it would be about the same as the Set-N-Forget model and the cost would be lower.

 

I don't know if my fan is Delta and at the moment I don't have a possibility to check it. But I agree, that different fat consoles will need different resistance for controlling.

Truly, I don't know how's that actually works. I knew, that fan is controlled by PWM wire so I tried to make it run high, but not be too noisy. So I tried to apply different resistance, and I noticed, that changing in speed is located in a small range of resistance. At first I just applied ~63kOm for running about 70% all the time, but some games would'nt need such air flow, but in FFXIII console overheated in 15 minutes. So I just installed potentiometer in that range. But I don't know how really that resistance influence on speed.

So, everybody that has multimeter, couple of resistors and couple of hours can do it with his/her own PS3, without a deep knowledge of electrotechnics )

 

Your method exploits a weakness in PWM. Basically, PWM is all about the ratio between "on" and "off". "off" might be rated at 0V-1V, "on" might be rated at 2v-5V. If you put the voltage between 1v and 2v (just example numbers), then the circuitry inside the fan detects both "off" and "on" from a single voltage, with the ratio of "off" and "on" changing based on the voltage. It works...but getting it exactly correct is a real bother.

 

It seems you're right. But I having here another problem. The thing is, I bought PS3 with non-working FAN and BD-Drive, but for very low price. After repairing (I restored BD drive and set FAN for ~70% of speed) I've played Resistance for ~15 hours without any freezes. But some games, like Marvel Ultimate Alliance, Dark Kingdom had freezes ~after 15 minutes of gaming.

And now I'm playing FFXIII. It works smoothly only at 100% of fan speed, but still, in some locations I have freezes ~after 5-10 minutes of playing.

It seems, that I have FAT ps3, that is extremely easy overheating. Maybe one or another component is faulty in some way.
I've changed thermal paste, added extra thermal heatsinks on memory, and fan at fullspeed - and still freezes.

If I play FF on my old SD TV - it would'nt overheat even at 70% of fan speed (low resolution), but HDTV causes console to freeze.

 

It may be this...

http://killerbug666.wordpress.com/2011/06/14/poor-contact-areas/

...or it could be the thermal grease under your RSX chip cover.

 

I hope that it's because of poor contact. I'll check this in one of upcoming days (I think I've spend ten times more time on technical improvement, than games ). Thx for the link! )

P.S. Have you opened these chips? Have you reapplied thermal paste on them?

 

I don't think the cell can be opened; it seems to be bonded with solder. The RSX does open, but I have not done it.

http://forums.afterdawn.com/t.cfm/f-153/ps3_rsx_chip_came_off_the_motherboard-809486

 

Yeah, thanx for the info! This morning I've opened my ps3, opened RSX chip and scratched off thermal compound and paste. I applied new paste (not adhensive), and on the heat spreader itself. I've cheked heatsink bases, but I wouldn't say that they're warped, they looked well enough. And I added washers on heatsink screws (before I only bend them slightly).

I assembled console, and cheked FFXIII on fullspeed fan for hour, and Gran Turismo 5 at 70% of speed of fan for an hour, and no freezes! (Last time I boot up console - It lasts only 10 minutes) fusk yeah!

Thank you again for info, it seems that problem appeared because of low quality thermal compound, that is used in RSX chip. That's why no matter what quality of thermal paste I used on heat spreader, and no matter what speed of fan it has. CELL chip seems to have more reliable conjunction, and as a result - more reliable heat withdrawal.