[PLOG] Mac G5 --> PC Conversion

[angelkiller]

Cool stuff ib_prime! And thanks for the info about the power switch ground. I'm going to connect the FW and USB shield grounds to a FW and USB ground, respectively. I know Cuzza said to run a ground straight from the PSU, but all the cables going to the front panel have already been sleeved and removing that to add one cable would be a major PITA. And running a single bare cable to the front panel would look... well.. you know.

I don't have time right now to do this, but I'm gonna try really hard to get it done tonight. I also wanna get Win7 installed also.

Oh, I don't know if I ever talked about the software for this thing. Originally, I wanted to dual boot OSX and some kind of Windows. I picked OSX mostly because I wanted to use Final Cut Pro for my video class. That's not gonna happen. Mostly because the teacher doesn't have one of those volume academic licenses where you can install it on any academic related computer. There's only one copy of FCP and it's in use. Second, the other Macs run some version of OSX that is not compatible with Intel Macs (so pre 10.4). So I would have to get my own copy of OSX. So no OSX here. So which version of Windows? I have a spare copy of XP, XP x64 and Win7. I chose Win7 only because it's newer and cooler. Edit: I've have just recieved word from the 'HelpDesk' people at my school that they've recieved their copies of Win7!! The head guy said that they'll be giving them out to a few select technically-savy students and I'm one of them! Sweet! So I'm definately gonna get the real Win7 on this Mac at some point in the near future (2 weeks). For now, I'll just stick with XP just to get it working.

I've linked the other wire diagrams that I've collected. Unfortunately, I don't have their original sources, but I bet they're from Insanely Mac's forums.

http://img.techpowerup.org/091018/G5 Front Panel Wiring Diagram 1.jpg
http://img.techpowerup.org/091018/G5 Front Panel Wiring Diagram 2.png

Eventually, I'll make a post with all the links and images/diagrams and such that I've used for this project.

 

[Cuzza]

Man I wish I was still at school - Free Win7! Not to mention the schoolgirls. *sigh*

 

[angelkiller]

Not to mention the schoolgirls.

And they're all legal for me.


I soldered the USB shield to the USB ground, so both pins share the same cable. Makes a Y shape. I did the same with the FW Shield and a FW ground.

Unfortunately, the power button still doesn't work.

On the bright side, now, when I plug in the plug, the machine turns on for about four seconds then shuts off. I mean the thing came on! The lights and everything! But it was a bit short lived.

Anyone wanna take a guess as to why simply plugging the cord in turns the thing on? Why it shuts of so quickly? Why the power button still doesn't work?

In the meantime, I'll try to get access to the power switch pins so I can turn it on manually. BTW, the red cathodes look sick!!

 

[p_o_s_pc]

And they're all legal for me.


I soldered the USB shield to the USB ground, so both pins share the same cable. Makes a Y shape. I did the same with the FW Shield and a FW ground.

Unfortunately, the power button still doesn't work.

On the bright side, now, when I plug in the plug, the machine turns on for about four seconds then shuts off. I mean the thing came on! The lights and everything! But it was a bit short lived.

Anyone wanna take a guess as to why simply plugging the cord in turns the thing on? Why it shuts of so quickly? Why the power button still doesn't work?

In the meantime, I'll try to get access to the power switch pins so I can turn it on manually. BTW, the red cathodes look sick!!

I had that happen before to me. I had the power switch on the +pin and the other side on the pin next to it(pwr. LED i think) so it shorted it out and kept it shorted.

 

[Cuzza]

yeah that's what it sounds like to me. somehow the power switch is being permanently grounded, just like you were holding the button in.

 

[troyrae360]

try unpluging the power swtich jumper and starting it up with the end of a screw driver (just short the pwr + an -) that should start it up then see if it keeps going.

PS you just need to touch the screwdriver on the pins and take it away, dont leave it there!

 

[angelkiller]

This is such a PITA. In order to plug the front panel connector in, I have to remove both fans on the rad. The only way to do that without taking the mounts out is to stick a screwdriver through the holes on the front of the Mac so I can unscrew the fans. When the fans are loose, I can slide them away so I can access the area where the cable plugs in. Then I gotta put everything back.

Ok. It makes sense that the power button is permantly set on. BUT HOW? I'm looking through the three wiring diagrams (Diagram 1, Diagram 2, Diagram 3) and I'm seeing some inconsistencies. Just look at all three. They don't agree on the same way to get the power button working.

I'm going to try to explain the differences, but it's hard because I don't know which is right; right is only relative to what the other say. If the little bump on the connector is on the right side, then I'm going to call the top right pin #1, the top left #2, the one right under #1 will be #3, and across from that is #4 and so on. Diagram 1 and Diagram 3 are already numbered like this. So Diagram 1 calls pin 13 the negative power switch. However, Diagram 2 calls pin 13 the positive LED and diagram 3 calls it the positive power switch. Diagram 1 calls pin 14 the positive power switch, Diagram 2 calls it just the 'power button', I assume it's positive. Diagram 3 calls pin 14 the positive LED. WTF??

My biggest issue here is that none of them agree on anything. It would be different if all three called pin 13 the power switch, but two said it was positive and one said it was negative or something. But they disagree on the function of the pin and they are so inconsistent that I can't draw any conclusions. You would think when 3 diagrams show the same damn thing that they be pretty close? Guess not.

Edit: I just had a brillent idea! Why not just check the damn Service Manual? I forgot I had it. I whip it out and I get this:

Pin 13 is the LED (positive I assume) and pin 14 is the power button. And that 's from Apple's own service manual. I don't think you can get much more official than that. Unfortunately it doesn't say where either of those ground. I just hope it's not the Chassis grounds because those are the only 2 pins out of 18 that I haven't connected. (and wiring them up properly isn't a small task) Ok, now that I have this info and I know it's right, what does everyone recommend?

@troyrae360
I would do that, but this is a pic of where the pins are. I can only get to the pins if I remove the optical drive. While that sounds seemingly simple (alliteration!) the opticial drive has an IR LED and detector sticking out from the front, and taking the optical drive out requires removing those, which I currently have carefully positioned. I would like to avoid having to reposition those. If necessary I will, but you get what I'm saying.

 

[Cuzza]

I thought we had already established that it IS the chassis grounds.

 

[angelkiller]

Ok. I got confused because I connected the chassis grounds to the FW and USB grounds and it's not working. So I guess I'll have to properly wire the chassis grounds then.

 

[ib_prime]

Hey, if its only the power button thats giving you the problem now, you can just use a multimeter and test the with lead completes the switch. I did it this way for the switch and found it was completed by Pin6, I can't verify now as I've completely destroyed my front connect IC board ... haha I was having similar headaches with the front connect as Angelkiller... so i just scratched up the board and connected everything directly by wires to the bottom lead points... which I guess angelkiller can do too.

 

[angelkiller]

Hey, if its only the power button thats giving you the problem now, you can just use a multimeter and test the with lead completes the switch. I did it this way for the switch and found it was completed by Pin6, I can't verify now as I've completely destroyed my front connect IC board ... haha I was having similar headaches with the front connect as Angelkiller... so i just scratched up the board and connected everything directly by wires to the bottom lead points... which I guess angelkiller can do too.

Great idea! I didn't think of that. But first, I need some more wire. Getting 18 wires from the mobo headers to the front panel used up all my supply. Hopefully, I can pick some up tomorrow. Again, awesome idea.

 

[angelkiller]

 

[erocker]

Man I wish I was still at school - Free Win7! Not to mention the schoolgirls. *sigh*

There's nothing stopping you from pretending. I've been known to hang around the student union.

 

[angelkiller]

Notes:

- CPU stabilizes around 25C at stock speeds & voltage
- Res needs more water, right now it's really turbulent and air gets in the tubes.
- Optical drive is fucked up. Opens & closes ok, but when it goes to spin the disk, you hear this noise like the disk is wobbling while spinnging then it stops completely. Luckiliy, I have on on hand. There goes the IR LED mod. I'll have to retrofit that mod into the new drive sometime in the future.
- Power button works! But only in 1 direction. If you pres it again, nothing happens. Hold it in, still nothing. Only way to power down is a hard unplug.
- Speaking of plug, the AC power recepticle on the actual computer is REALLY fragile. If you plug the AC cord into it, it comes out of its socket. Temp solution is to not unplug it. Will have to come up with a solution later.
- It's pretty damn loud.

I'm gonna add some more water, change the ODD now. Tomorrow I'll take it to school and I can get Win7 installed and get this OC'ed in my free period and at lunch.

 

[phanbuey]

put the fans in pull - youll get better temps and it will be quieter.

great job btw - looks bootiful.

 

[angelkiller]

What do you mean in pull? Every fan sucks from the left, blows to the right.

And thanks. I'm soooo happy right now. I'm just relieved that it's finally done.

Win7 RC is installing it now. I'm gonna have a go at OCing tonight! Can't wait!

All the other macs have a sign above them with a little info on them. So I decided to make my own sign. It's a parody of the original sign. It's funnier if you could see the original signs.

 

[p_o_s_pc]

looks cool. nice to see you got it working
you should have it crunching(WCG) since you don't pay the electric bill and i'm sure its left on all the time

 

[angelkiller]

looks cool. nice to see you got it working
you should have it crunching(WCG) since you don't pay the electric bill and i'm sure its left on all the time

Thanks. And I definitely will get this crunching. In fact, I should get my main rig back crunching again too.

More notes:

- After filling up the res, it's alot quieter. The main two sources of noise are the 2 Yate Loon intakes and the 2 stock Apple exhausts. The exhausts are the worst. They have this really grainy sound to them and they make up like 70% of the perceived noise. The 2 80mm fans for the PSU run at 6v and the GPU fan are not noticable over the intake and exhausts. The Gigabyte mobo doesn't seem to be controlling the fans, even though I have it set in the BIOS. Whatever.

- Google Chrome is fast no kidding. I'd switch in a heartbeat if it had extensions.

 

[Cuzza]

Nice one buddy, awesome to see this done! Any thoughts of WCing the GPU?

@ erocker, I've been known to stop by the student union building and use the showers

 

[brandonwh64]

So what was the verdict on overclocking? what speed did you get it too?

 

[angelkiller]

Nice one buddy, awesome to see this done! Any thoughts of WCing the GPU?

Thanks. I had considered it, but I decided against it because I thought it would add unnecessary heat to the loop. I bet it would look cool though.

So what was the verdict on overclocking? what speed did you get it too?

I fell asleep.

But I'll definitely get around to doing it though. I can't just leave it a 2.13GHz.


It idles at 32C in Windows and loads at 50C. That's at stock. 2.13GHz @ 1.23v. I expect to get at least 3GHz and I hope to get ~3.5GHz.

 

[angelkiller]

Ok, here's what I got.

Notes

- Power switch works fine now
- I'm on bios version F5 and F9 is out now. I'll try flashing it soon.
- Anything higher than 3.4GHz wouldn't make it to Windows.
- Extra voltage doesn't help. I tried 3.5GHz at 1.5v and that didn't work. If it needs more than .05v to go 100MHz then I think it's not worth it.
- My definition for 'stable' for this OC session was that it passed Linpack for around 6 minutes. I'm going to do leave Linpack running overnight at 3.4GHz.
- And I just realized that my CPU has a G0 revision. Whoa. I was told that it was a B3.

 

[Cuzza]

awesome! looks like your watercooler is doing the job. I see what you mean about not adding the gpu to the loop, it might knock back your overclock.

 

[angelkiller]

awesome! looks like your watercooler is doing the job. I see what you mean about not adding the gpu to the loop, it might knock back your overclock.

Yeah. You really think the temps are ok? I thought they were a bit on the hot side. I mean I would think that an air cooler could keep a 3.4GHz quad at 70C. :\

 

[angelkiller]

Some final pics. Unfortunately, I'm no photographer, and the camera wasn't mine, so they came out really grainy.

The Video Editing Lab: And yes, that's a PS3 on the cart. Best. Class. Ever.

Sometimes I have a little trouble telling which one is mine.

Maybe I'll get some better quality pics, but for now, these'll do. I think it's safe to say that this project is finished for the time being. There are a few things I'd like to do to it, but for now, I'm done. Any questions or comments?