Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Power Mac G4 MDD (2002) - From a Kernel Panic galore (!) to power supply troubles (sigh)
- Thread starter Alexis Trinquet
- Start date
- Sort by reaction score
@reddrag0n @Tratkazir_the_1st - I'll do that.
I don't have isopropanol at my disposal but I have a few erasers.
Although, I thought I had a old PC mobo with DDR slots, but apparently it was so broken that I threw it away.. So the MDD is the only DDR machine that I have at the moment.
I don't have isopropanol at my disposal but I have a few erasers.
Although, I thought I had a old PC mobo with DDR slots, but apparently it was so broken that I threw it away.. So the MDD is the only DDR machine that I have at the moment.
@reddrag0n - It's a 167MHz bus MDD, so PC2700 RAM it is then. I don't think I have any DDR2100 modules anyway.
There's no debris in the slots that I can see, though I noticed if I only use single-side modules (like the Samsung module that the MDD came with and a 512MB ProMOS that I had in my parts bin), it boots consistantly, yet with double-sided modules... Not so much.
The Kingston, Elexir and the 256MB ProMOS modules are all double-sided.. Maybe that's where the crux of the issue lies?
There's no debris in the slots that I can see, though I noticed if I only use single-side modules (like the Samsung module that the MDD came with and a 512MB ProMOS that I had in my parts bin), it boots consistantly, yet with double-sided modules... Not so much.
The Kingston, Elexir and the 256MB ProMOS modules are all double-sided.. Maybe that's where the crux of the issue lies?
Have you tried moving the Samsung stick that works with OS 9 to other slots? To see if it is RAM problem or slot problem. Then if you've already tried that. I would boot into Tiger and use Rember 0.3.4b to test all the RAM.
https://www.kelleycomputing.net/rember/
Scroll down the page to find it. I've found the version for Leopard is not as good as the version for Tiger.
https://www.kelleycomputing.net/rember/
Scroll down the page to find it. I've found the version for Leopard is not as good as the version for Tiger.
Last edited:
Yup, of all the sticks, the Samsung and the single-sided ProMOS seem to not cause any trouble, whenever I put any double-sided module, this is where I start getting instabilities.
I'm not at home at the moment, but I'll give Rember a shot.
Okay, I gave Rember a shot, with a ProMOS 256MB double sided stick in slot J21, I started it at 17:10 (CET/UTC+1) approximately, and left it minding its business in the meantime..
Came back at 18:30, saw that it was still plowing away at the RAM... Then I saw that the clock was stuck at 17:31. Yup, it crashed.
I tried to reboot it multiple times, no joy, until I slightly jostled the module around. And now it's back at it again with a 512MB module thrown into the mix for good measure.
I've seen bad RAM messing stuff up but if it's indeed the reason why I was getting all these KPs in a row, I'd be damned!
Came back at 18:30, saw that it was still plowing away at the RAM... Then I saw that the clock was stuck at 17:31. Yup, it crashed.
I tried to reboot it multiple times, no joy, until I slightly jostled the module around. And now it's back at it again with a 512MB module thrown into the mix for good measure.
I've seen bad RAM messing stuff up but if it's indeed the reason why I was getting all these KPs in a row, I'd be damned!
Dodgy RAM can cause all kinds of problems, including data corruption. It‘s a nightmare.
Indeed it's a real nightmare. I've seen it multiple times, it all takes one single (or more) chip failing to render the whole module cactus 
I recall one particularly nasty example I've seen, an HP Compaq desktop (remember these?) that would angrily beep at me and flash a red LED on the front.. Every. Single. DDR3 module I put in would yield the same result...
...until I tested the other modules in different slots it booted, but complained about the RAM not being "optimally installed" (or something like that) then I put some contact cleaner in the first slot and tried the other sticks into that slot, and bam! It worked!
Though the original module, regardless of which slot I put it in would cause the BIOS to hang right when it started counting the avaliable RAM 🤣
Not only it had a bad RAM module, but it also had a dirty slot!
Sorry I diverged a bit from the topic, but I feel like my MDD's doing the same thing, if I slightly push the modules sideways, it doesn't crash or anything!
However I let it run for a while to see if it would crash (burn-in test, while running Rember) and after three hours... Kernel panic again, did the ol' RAM switcharound and it's working again.
I recall one particularly nasty example I've seen, an HP Compaq desktop (remember these?) that would angrily beep at me and flash a red LED on the front.. Every. Single. DDR3 module I put in would yield the same result...
...until I tested the other modules in different slots it booted, but complained about the RAM not being "optimally installed" (or something like that) then I put some contact cleaner in the first slot and tried the other sticks into that slot, and bam! It worked!
Though the original module, regardless of which slot I put it in would cause the BIOS to hang right when it started counting the avaliable RAM 🤣
Not only it had a bad RAM module, but it also had a dirty slot!
Sorry I diverged a bit from the topic, but I feel like my MDD's doing the same thing, if I slightly push the modules sideways, it doesn't crash or anything!
However I let it run for a while to see if it would crash (burn-in test, while running Rember) and after three hours... Kernel panic again, did the ol' RAM switcharound and it's working again.
I hoped Rember would spot bad RAM sticks because the solution would be just buying some new ones. But now the RAM's slots are suspects as well.
The Samsung stick may be immune because it may be a tad thicker than other sticks. RAM modules rarely go bad on their own. I have only had a bad stick so far and it is because a small ceramic capacitor is missing.
I would try cleaning the RAM slots by blowing dust off them using a lens blower or a can of dust spray. Then, I would fold a piece of lint-free tissue (preferably lens tissue) into a suitable shape and soak it with isopropyl or lighter fluid (naphtha) and push it into the slot and rub back and forth, then do it again with a dry piece of the paper to absorb any fluid that is still there, and blow it again to get any lint off.
Then I would dab the isopropyl (or naphtha) onto the RAM module's contacts (given that it has been cleaned) and push it into the slot while it's still wet, pull it out and push it a few times to remove some oxidation.
Hopefully it is not caused by RAM slots' solder joints on the main board.
The Samsung stick may be immune because it may be a tad thicker than other sticks. RAM modules rarely go bad on their own. I have only had a bad stick so far and it is because a small ceramic capacitor is missing.
I would try cleaning the RAM slots by blowing dust off them using a lens blower or a can of dust spray. Then, I would fold a piece of lint-free tissue (preferably lens tissue) into a suitable shape and soak it with isopropyl or lighter fluid (naphtha) and push it into the slot and rub back and forth, then do it again with a dry piece of the paper to absorb any fluid that is still there, and blow it again to get any lint off.
Then I would dab the isopropyl (or naphtha) onto the RAM module's contacts (given that it has been cleaned) and push it into the slot while it's still wet, pull it out and push it a few times to remove some oxidation.
Hopefully it is not caused by RAM slots' solder joints on the main board.
Given the age of these machines, irrespective of brand, capacitors may be past their best. I’d try swap out the PSU if possible, the random nature of the panics you’re getting says cpu power or heat problem to me.
The Samsung PSU has already gone to the great electronics heaven I'm afraid.. And the AcBel is currently waiting for its new caps.
About that particular RAM module, it seems like that that particular 256mb module also feels quite loose in the slot, but if I push them to the side just enough, I get consistant booting, that probably doesn't mean anything. Though the idea that the RAM slots' solder joints might be cracked... I hope that it's not the case.
Although the Kingston module I thought was dead actually seems to be working? I got the MDD to turn on with all sticks in, totaling 2GB of RAM, even though that didn't last long before I let go of one of the modules and had an instant crash.
About that particular RAM module, it seems like that that particular 256mb module also feels quite loose in the slot, but if I push them to the side just enough, I get consistant booting, that probably doesn't mean anything. Though the idea that the RAM slots' solder joints might be cracked... I hope that it's not the case.
Although the Kingston module I thought was dead actually seems to be working? I got the MDD to turn on with all sticks in, totaling 2GB of RAM, even though that didn't last long before I let go of one of the modules and had an instant crash.
Update: Seems like I can reliably cause a crash by using the "Processor" control panel (after having installed dev tools from my 10.4 DVD) and putting one CPU to sleep THEN waking it back up.
Or if I wiggle the RAM modules around... I let Rember run a full cycle and it didn't find anything wrong with the RAM, I didn't try 255 cycles because, I'm using a 250W ATX PSU in the meantime that I fix up the AcBel 360W and I'd rather not run it in that condition for extended periods of time.
I guess I'm going to hunt for two 1GB modules that are sufficiently thick... Or replace the DIMM slots (I feel like using a soldering pot will be mandatory due to the amount of pins... Assuming it's even possible to do so).
This MDD looks like it really is cursed somehow... It's got an attitude but tough luck for it, I can be a bit hardheadedly persistent sometimes. And I'm not going to give up so easily on it.
Or if I wiggle the RAM modules around... I let Rember run a full cycle and it didn't find anything wrong with the RAM, I didn't try 255 cycles because, I'm using a 250W ATX PSU in the meantime that I fix up the AcBel 360W and I'd rather not run it in that condition for extended periods of time.
I guess I'm going to hunt for two 1GB modules that are sufficiently thick... Or replace the DIMM slots (I feel like using a soldering pot will be mandatory due to the amount of pins... Assuming it's even possible to do so).
This MDD looks like it really is cursed somehow... It's got an attitude but tough luck for it, I can be a bit hardheadedly persistent sometimes. And I'm not going to give up so easily on it.
I'd rather not if I can avoid it (but if I have to, I'll do it and set aside the original board), just in case there's a guy selling a 133MHz bus board on the local classifieds, so in case of, I'll probably buy it and do the 167 bus mod.
If it's just the DIMMs being flaky, I'll try and rig up a shim, I can live with that..
If it's just the DIMMs being flaky, I'll try and rig up a shim, I can live with that..
i have an mdd 2003 board (aka 167 mhz with airport extreme slot & FW 800) sitting in my excess parts bin if you want it. it will run os9 with a few software mods but as i said, it's just sitting in a bin. yes it's in an antistatic bag
as for ram, any ddr2700 or 3200 ram i have has been sourced into my G5 which is sitting on the floor not doing much either
as for ram, any ddr2700 or 3200 ram i have has been sourced into my G5 which is sitting on the floor not doing much either
I appreciate the offer Red, but I'm afraid shipping costs will be a real killer, alas. I really gotta fix this board one way or another. Even if I have to spend hours desoldering the RAM slots pin by pin (at least I'd learn how to replace RAM slots along the way)
Now it has decided to start another KP/freeze cycle again... Piece of s....junk.. None of the RAM modules I put it in work now. It booted fine yesterday and now it's back at this.
Frankly, I don't even know what to make of this, I haven't bumped it or disturbed it in any other way.. Tomorrow when I'll have the opportunity I'm gonna buy a can of contact cleaner and blast some into the DIMM slots, I hope that's gonna be it.
Now it has decided to start another KP/freeze cycle again... Piece of s....junk.. None of the RAM modules I put it in work now. It booted fine yesterday and now it's back at this.
Frankly, I don't even know what to make of this, I haven't bumped it or disturbed it in any other way.. Tomorrow when I'll have the opportunity I'm gonna buy a can of contact cleaner and blast some into the DIMM slots, I hope that's gonna be it.