Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Problem with overclocking PowerBook G4 Hi-Res

I got the CPU Director working and its disapointing to see that even this nice program cant see the new clock but the old stock...

5F88C1C6-62AC-41BD-BB48-817894791F8D.jpeg


But just for pure fun I disabled the L2 cache in this program and the Xbench score was 15. So now I know how bad it would be without the L2 cache.

EDIT: I mentioned the full load and TFF using a lot of resources. For two days I was testing the TFF and the fans are kicking in less often than with stock clock. It seems that the overclock is beneficial for the TFF.
 
Last edited:
I have an extra G5 Quad that needs a dual-pump fix which I was keeping around to some day experiment on replacing with modern air-cooling, as I'm not a fan of liquid cooling. I haven't turned in on in ages as I've had so much else going on. I've been trying to find someone with access to machining to make custom heatsink for the quad to send them this one as a way to create/test the heatsink so I can replace the liquid cooling system in my main G5 and possibly anyone else that wants to convert their quad to air-only cooling.
882d3bfda81cc0f3b9da69ad94268622.jpg
 
  • Like
Reactions: z970
reukiodo said:
I have an extra G5 Quad that needs a dual-pump fix which I was keeping around to some day experiment on replacing with modern air-cooling, as I'm not a fan of liquid cooling. I haven't turned in on in ages as I've had so much else going on. I've been trying to find someone with access to machining to make custom heatsink for the quad to send them this one as a way to create/test the heatsink so I can replace the liquid cooling system in my main G5 and possibly anyone else that wants to convert their quad to air-only cooling.
882d3bfda81cc0f3b9da69ad94268622.jpg
Click to expand...

How much do you want for it?
 
Okay, yesterday I disassembled the DLSD and cleaned it and replaced the thermal paste and the difference are 3-4 degrees of Celsius. So thats good. The fans are kicking in about the same but not to such high RPM as before and also when the CPU load drops the temp quickly drops too so as the fan RPM. So now its more quieter experience and whats more important its a bit cooler. I cant seems to find any software that can allow me to change the second fan kick in temps. The G4FanControl sets the curves just for the first fan and I believe if they were working in parallel and not one and then the other it would dissipate more heat more quietly. Also I am adding some photos that the CPU in DLSD is not the crippled G5 but the 7447B. The original thermal stuff was gummy and sticky like aquarium silicone and hard to remove even with brush and alcohol.

image1.jpeg image2.jpeg image3.jpeg image4.jpeg image5.jpeg image6.jpeg
 
  • Like
Reactions: juanstdio and dextructor
Don't mind me bringing this post back from the graves, but I am trying this on a 15-inch DLSD. Have some 10kOhm resistors on the way. Apparently you can't simply bridge the BOM solder pads. My goal is to get to the 189MHz bus. I did try bridging those before I read that 10k resistors were required, which simply led to the fans spinning up at full blast. Genuinely hope everything will work as soon as I get the resistors.

I keep seeing that post on HardForums every time I look up the topic of overclocking a PowerBook. No, I am not replacing it with a Chromebook.
 
  • Like
Reactions: JoyBed
k24a1 said:
Don't mind me bringing this post back from the graves, but I am trying this on a 15-inch DLSD. Have some 10kOhm resistors on the way. Apparently you can't simply bridge the BOM solder pads. My goal is to get to the 189MHz bus. I did try bridging those before I read that 10k resistors were required, which simply led to the fans spinning up at full blast. Genuinely hope everything will work as soon as I get the resistors.

I keep seeing that post on HardForums every time I look up the topic of overclocking a PowerBook. No, I am not replacing it with a Chromebook.
Click to expand...
First of all, dont tryvjust ramping up your bus without prior changes to every other settings like the core voltage or core multiplier. Also you actually can just bridge the pads, I was doing that while experimenting with MDD daughterboard. The reason your fans ramped up and you got no boot is that your core clock was raised as well with the bus, because the core clock is just bus clock multiplied by the set BOM. First of all, lower the core clock multiplier and then raise the bus. But dont jump to 189MHz straight away, try the next higher setting(with the lowered core multiplier!) and then you can go even higher if it boots. You need to go step by step, not jumps cuz if your big jump in setting the BOMs results in no boot then you cant properly figure out if its the higher bus clock or higher core clock that resulted in no boot situation. In the end if you find your highest possible bus clock then you can start upping the core multiplier. In that situation you will find the highest core clock possible. Also I forgot to mention… In all of these steps, DONT CHANGE YOUR CORE VOLTAGE! If you get your highest bus and core clocks and they are stable, JUST THEN you can up the core voltage a bit. The reason Im saying this is because when you will start increasing the clock the thermal output will also raise. Giving it also more voltage would increase the thermal output SIGNIFICANTLY. So do all experiments with stock voltages. If you encounter any problems just reach out to me. I would also be glad for any feedback how is it going. PS: get the best RAM in that PowerBook as you can find, bus clock will raise their clock too so be sure that they are good and they can handle it.
 
  • Like
Reactions: alectrona6400
JoyBed said:
First of all, dont tryvjust ramping up your bus without prior changes to every other settings like the core voltage or core multiplier. Also you actually can just bridge the pads, I was doing that while experimenting with MDD daughterboard. The reason your fans ramped up and you got no boot is that your core clock was raised as well with the bus, because the core clock is just bus clock multiplied by the set BOM. First of all, lower the core clock multiplier and then raise the bus. But dont jump to 189MHz straight away, try the next higher setting(with the lowered core multiplier!) and then you can go even higher if it boots. You need to go step by step, not jumps cuz if your big jump in setting the BOMs results in no boot then you cant properly figure out if its the higher bus clock or higher core clock that resulted in no boot situation. In the end if you find your highest possible bus clock then you can start upping the core multiplier. In that situation you will find the highest core clock possible. Also I forgot to mention… In all of these steps, DONT CHANGE YOUR CORE VOLTAGE! If you get your highest bus and core clocks and they are stable, JUST THEN you can up the core voltage a bit. The reason Im saying this is because when you will start increasing the clock the thermal output will also raise. Giving it also more voltage would increase the thermal output SIGNIFICANTLY. So do all experiments with stock voltages. If you encounter any problems just reach out to me. I would also be glad for any feedback how is it going. PS: get the best RAM in that PowerBook as you can find, bus clock will raise their clock too so be sure that they are good and they can handle it.
Click to expand...
Thanks for the advice.
I also bridged the BOM back to 167MHz and I did hear the CD drive spin up but then the fans went up again. Tried resetting and I didn't hear the CD drive and the fans spun up. Could be a bad bridge?
So from my understanding you run the system at 1.67GHz yet you have the bus at 189MHz after stepping up a few times?
 
JoyBed said:
lower the core clock multiplier and then raise the bus. But dont jump to 189MHz straight away, try the next higher setting(with the lowered core multiplier!) and then you can go even higher if it boots.
Click to expand...
Hmm, where would the core clock resistors be on a 15" HiRes PowerBook G4 motherboard?
 
k24a1 said:
Thanks for the advice.
I also bridged the BOM back to 167MHz and I did hear the CD drive spin up but then the fans went up again. Tried resetting and I didn't hear the CD drive and the fans spun up. Could be a bad bridge?
So from my understanding you run the system at 1.67GHz yet you have the bus at 189MHz after stepping up a few times?
Click to expand...
Double check your connections, check if you resoldered it correctly, use flux so the connection are good. And no, My DLSD runs at 189MHz bus and 1,89GHz core. I couldnt go any higher without increasing voltage and that always meant quick overheating. On stock voltage it can be like that at max.
k24a1 said:
Hmm, where would the core clock resistors be on a 15" HiRes PowerBook G4 motherboard?
Click to expand...
I think I found all info I needed here: https://www.overclock.net/threads/ultimate-mac-overclocking.1445164/ But I also have all the schematics and stuff for the 17” DLSD so I was just double checking the info on the site and comparing it to the schematics. Also when you will complete all the laboring and everything is stable and at maximum as it can get clean up the old thermal goop and replace it with a good thermal paste. Let me know how its going!
 
What did you have to do before you got it to boot successfully? Could you perhaps label the resistor locations corresponding to the BOM settings?
 
k24a1 said:
What did you have to do before you got it to boot successfully? Could you perhaps label the resistor locations corresponding to the BOM settings?
Click to expand...
To get a successful boot you just need to correctly set the BOMs to a setting the CPU can handle. Also having RAM sticks that can handle overclocked bus are required, so in my case I just plugged there RAM sticks that had higher specs than the PowerBook runs at stock PC2-4200 and I tossed there a PC2-5300.
 
  • Like
Reactions: lepidotós
Alright @JoyBed I have managed to boot at 189MHz FSB but the system does kernel panic even with better memory (PC2-5300). Do I adjust vcore and how would I step back to 183MHz if needed?
 
k24a1 said:
Alright @JoyBed I have managed to boot at 189MHz FSB but the system does kernel panic even with better memory (PC2-5300). Do I adjust vcore and how would I step back to 183MHz if needed?
Click to expand...
Okay, post me the kernel panic. Also what is the core multiplier when you have the 189MHz bus ?
 
Unfortunately it isn't in much detail and it just says your computer needs to shut down. The core clock was the default 10x

Also can you guide me through how the resistors are laid out?
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back