Thread
G3 Lombard issues
I got a G3 lombard that has issues that make me want to smash it all to pieces. but id rather avoid that situation and try to work through the issues.
It was given to me missing the HDD and processor card. basically a stripped machine. it was stripped for a reason, ya know.... well i think i know... hehe.
So i tossed in a processor card with RAM, and I removed PRAM battery plug for safe measure.... Also the main battery is dead. so its removed out of slot.
When you first plug it in, the green light comes on and stays on. Press the reset button in the back, the green light flashes once and goes out. press power, nothing happens.
this cycle is never ending... with the reset button. press once, light flashes once, press again, light comes on stays on. however nothing ever happens from this. no boot.
But sometimes if i remove the power plug and plug it back in, i gotta do this several times, somtimes the green light will come on and then go off. I press power button, the machine will chime and boot. SOMETIMES. sometimes it just chimes and freezes. and i gotta go through the whole process over again.
sometimes it will chime, boot and go to the gray screen and freeze there. Sometimes itll boot all the way up into my OS9 CD and work perfectly fine. no freezing. Sometimes itll boot to the OS9 startup screen and give me a System Error Address Error, reboot it and black screen. gotta screw with it again.
I tossed a jaguar CD in there and itll boot that up. SOMETIMES.
so i put the PRAM batt back in and plugged the AC cord back in. green light came on and stayed on. then pressed reset in back. green light flashed and stayed off, no boot. let it seit for a couple hours and unpugged the AC, plugged it back in, green light came on and went off. press power, booted.
it was booting and rebooting consistently without problems. so when i thought all was said and done, i rebooted the machine again, and black screen. so i unplugged it, and plugged it back in, the LCD flickered and green light stayed on. press reset button, green light went off and nothing, so i got stuck back into this SAME cycle again.
fiddled with the AC cord unplug, plug back in and eventually got it to boot up again, but it went to the gray screen with "?" and froze. it never ends...........
I have learned that the machine will not chime and boot until i see the green light at least stay on for 2 seconds and go out, on its own. then i know its ready to chime and boot. But if the green light stays on solid, or doesnt come on at all when power is plugged in, i know it will NOT boot even when messing with reset button. I have to get the 2 seconds on, then go off condition of the green light.
Then itll boot to the happy mac, then the open firmware screen comes up with a DEFAULT CATCH error and freezes. its going CRAZY
any ideas?
It was given to me missing the HDD and processor card. basically a stripped machine. it was stripped for a reason, ya know.... well i think i know... hehe.
So i tossed in a processor card with RAM, and I removed PRAM battery plug for safe measure.... Also the main battery is dead. so its removed out of slot.
When you first plug it in, the green light comes on and stays on. Press the reset button in the back, the green light flashes once and goes out. press power, nothing happens.
this cycle is never ending... with the reset button. press once, light flashes once, press again, light comes on stays on. however nothing ever happens from this. no boot.
But sometimes if i remove the power plug and plug it back in, i gotta do this several times, somtimes the green light will come on and then go off. I press power button, the machine will chime and boot. SOMETIMES. sometimes it just chimes and freezes. and i gotta go through the whole process over again.
sometimes it will chime, boot and go to the gray screen and freeze there. Sometimes itll boot all the way up into my OS9 CD and work perfectly fine. no freezing. Sometimes itll boot to the OS9 startup screen and give me a System Error Address Error, reboot it and black screen. gotta screw with it again.
I tossed a jaguar CD in there and itll boot that up. SOMETIMES.
so i put the PRAM batt back in and plugged the AC cord back in. green light came on and stayed on. then pressed reset in back. green light flashed and stayed off, no boot. let it seit for a couple hours and unpugged the AC, plugged it back in, green light came on and went off. press power, booted.
it was booting and rebooting consistently without problems. so when i thought all was said and done, i rebooted the machine again, and black screen. so i unplugged it, and plugged it back in, the LCD flickered and green light stayed on. press reset button, green light went off and nothing, so i got stuck back into this SAME cycle again.
fiddled with the AC cord unplug, plug back in and eventually got it to boot up again, but it went to the gray screen with "?" and froze. it never ends...........
I have learned that the machine will not chime and boot until i see the green light at least stay on for 2 seconds and go out, on its own. then i know its ready to chime and boot. But if the green light stays on solid, or doesnt come on at all when power is plugged in, i know it will NOT boot even when messing with reset button. I have to get the 2 seconds on, then go off condition of the green light.
Then itll boot to the happy mac, then the open firmware screen comes up with a DEFAULT CATCH error and freezes. its going CRAZY
any ideas?
It's gotta be the DC board. It also holds the audio jacks. I had this issue on my Pismo maaaannnyyy years ago, and I had the same symptoms. The DC board was replaced and all worked well.
Also, DC boards seem to be a weakness on the G3 models. The only flaw I have found with the G3 laptops is the DC board.
I also had an issue recently similar to this with the PowerBook G4 17" i refurbished. The power cable stayed orange, sometimes turned green, but it would never turn on. I replaced the part and at this time, I am typing on it. Fixed the entire machine with a $26/shipped part.
Also, DC boards seem to be a weakness on the G3 models. The only flaw I have found with the G3 laptops is the DC board.
I also had an issue recently similar to this with the PowerBook G4 17" i refurbished. The power cable stayed orange, sometimes turned green, but it would never turn on. I replaced the part and at this time, I am typing on it. Fixed the entire machine with a $26/shipped part.
that may be. not sure... I am using a Duo AC adapter to power the machine as thats all I have.
if the machine is frozen solid at a screen, it stays that way. it never seems to "shut down" or do a hard reboot. but it WILL hard lockup. frozen mouse and all
And worth mentioning, that sometimes when it chimes, till chime SUPER loud and distorted for about half of the chime. then it gets real low volume for the last half of the boot chime. but when it does this, it dont boot either, just chimes and black screen.
if the machine is frozen solid at a screen, it stays that way. it never seems to "shut down" or do a hard reboot. but it WILL hard lockup. frozen mouse and all
And worth mentioning, that sometimes when it chimes, till chime SUPER loud and distorted for about half of the chime. then it gets real low volume for the last half of the boot chime. but when it does this, it dont boot either, just chimes and black screen.
How many Watts is that Duo AC adapter? A Lombard should use a 45 Watt adapter. Duo adapters come in 18, 25, and 36 Watt varieties. It may just not be getting enough juice.
it doesnt say how many watts it is. it just says 45VA. It runs the powerbook 3400 I have just peachy.
Take the output voltage and current listed on the label and multiply them. That is, "Output 24V" x "1.04A" = 25 Watts. Your Lombard will want 1.7 Amps or more at 24 volts. Otherwise, the power supply may collapse when the hard drive tries to spin up or at some other power-hungry part of the boot process. If that 45VA is output power, it's enough. If it's referring to 50/60 Hz AC input, the adapter is probably too small.
Of course there may be something else wrong as well, but make sure you have the right power adapter first. You can boot a Lombard off of a 36 Watt (1.5A) Duo adapter if the Lombard has a good battery to help out.
Of course there may be something else wrong as well, but make sure you have the right power adapter first. You can boot a Lombard off of a 36 Watt (1.5A) Duo adapter if the Lombard has a good battery to help out.
That sounds like a CPU lockup to me. ISTR there was a known issue with some of the faster (400MHz?) CPU modules for the Lombard locking up like this. Or it could be overheating - worth checking the heat transfer (pad? paste?)
I am not sure about the differences in power requirements between the two but I was running my mildly loaded PDQ(with no batt) off a 25W Duo adapter for a fair while after I got it and only a few times it needed more than the adapter could give. I had to load absolutely everything to draw too much. Now, I would not really recommend running any power supply that close to it's limit but it is good enough for testing I suppose. *shrug*
its a 333 CPU i belive.
It isnt a heat issue, because when the machine is flat cold it still wont startup or bong or spin anything until i monkey with it for a half an hour unplugging, plugging in AC cord, pressing power, reset, blah blah blah.
But its at the point now, it will NOT boot successfully anymore. it always boots up to a System Address Error when it does boot into the OS, every 1/99 tries itll boot to the gray screen and begin to load the OS. but the other 98 tries it never boots or sometimes bongs with black screen, or does all kinds of stupid crap.
It isnt a heat issue, because when the machine is flat cold it still wont startup or bong or spin anything until i monkey with it for a half an hour unplugging, plugging in AC cord, pressing power, reset, blah blah blah.
But its at the point now, it will NOT boot successfully anymore. it always boots up to a System Address Error when it does boot into the OS, every 1/99 tries itll boot to the gray screen and begin to load the OS. but the other 98 tries it never boots or sometimes bongs with black screen, or does all kinds of stupid crap.
Anyone have an apple hardware test ISO for this model? maybe itll tell me what it is.
Edit: I it finally managed to boot OS9 again, but it popped up a dialog that said something along the lines of "the built-in memory detected a problem with cache-memory" then i clicked off the dialog too fast before i could read it all. i got a bad habit of that.
But i rebooted it again, and it system errored out again with illegal instruction.
illegal instruction?? Yeap, sounds like the processor card is toast :-( sucks too.
Edit: I it finally managed to boot OS9 again, but it popped up a dialog that said something along the lines of "the built-in memory detected a problem with cache-memory" then i clicked off the dialog too fast before i could read it all. i got a bad habit of that.
But i rebooted it again, and it system errored out again with illegal instruction.
illegal instruction?? Yeap, sounds like the processor card is toast :-( sucks too.
If it is the processor daughtercard, those are sometimes fairly cheap on eBay. You could even bump it to a 400mhz card.
I think it is. but i dont know for sure though.
Yeah. From memory (sorry) that was the issue with those CPU modules - bad cache. A quick Google for "lombard cpu lockup" turned up loads of hits. Have a look at xlr8yourmac.com. Seriously, it's a known issue. There was a bad batch of CPU modules, and unfortunately, if I remember correctly, it was amongst the 400MHz ones.
|
|
|
V
Considering the problems in reliability the "Professional PowerBooks" have sometimes, that's why I got a G3 iBook clamshell 366MHz graphite SE Tokyo Revision B model. Not as fancy, not as fast, not as easy to fix when something does break, but except when the temperature in our little part of the world gets too hot (at which point it flakes out until it cools off again, then it will resume (as proper as possible) operation.
Strange that NVRAM and PMU reset key combos act like they should but have no effect though, even with the battery removed, but oh well-it still is an awesome baby. And no, I don't mind trading incredible reliability in a considerably used 10 year old laptop for its toilet seat looks...because it's Graphite!
Hope that helps somehow!
Strange that NVRAM and PMU reset key combos act like they should but have no effect though, even with the battery removed, but oh well-it still is an awesome baby. And no, I don't mind trading incredible reliability in a considerably used 10 year old laptop for its toilet seat looks...because it's Graphite!
Hope that helps somehow!
this one is a 333 CPU. but i keep getting the cache error in OS9 when it does boot. so its shot.
ill have to replace it when i get some funds.
ill have to replace it when i get some funds.
its the solder on the CPU card that has cracked/failed.
If i push on the CPU card while the machine is booting, i get an instant system error of either address error, bus error, or illegal instruction.
then it doesnt boot again.
but if i take and put some force on the CPU itself alone, only, it boots everytime, straight into OS9 and no more cache errors.
problem found. once i head into the shop tomorrow, ill toss it on the reflow bench and give it a whirl.
If i push on the CPU card while the machine is booting, i get an instant system error of either address error, bus error, or illegal instruction.
then it doesnt boot again.
but if i take and put some force on the CPU itself alone, only, it boots everytime, straight into OS9 and no more cache errors.
problem found. once i head into the shop tomorrow, ill toss it on the reflow bench and give it a whirl.
It sings happily ever after.....
I reflowed the CPU and the chipset that is on the CPU card, and the machine now runs perfect. and even a bonus to boot: the battery started charging and has developed a full charge. something that it DIDNT do before.
no more random freezes, crashes, and no more cache errors.
I reflowed the CPU and the chipset that is on the CPU card, and the machine now runs perfect. and even a bonus to boot: the battery started charging and has developed a full charge. something that it DIDNT do before.
no more random freezes, crashes, and no more cache errors.
Amazing! <1% of us could have fixed that, I sure couldn't have. I would have bought a new CPU board.
That's awesome techknight! 
Have fun with your working G3 Lombard, great laptop, if only mine had a battery that would charge.
Have fun with your working G3 Lombard, great laptop, if only mine had a battery that would charge.
Excellent news! And potentially a solution to note for any others with the same symptoms.
Well i decided to try this because i fixed an imac G5 isight this way, it had screen and logic board issues, random shutdowns, etc... you could just press on the logic board in a certain spot, and it would just shut off..
I fixed the logic board by reflowing it, but i still havent gotten around to replacing the screen to rid of those annoying lines. However, after reflowing the lombard's CPU card, all problems did go away. However I am going to have to get the PROPER AC adapter for it. as i cant charge the battery and run the machine at the same time. it wont charge and the AC adapter gets bloody hot.
Soon as i shut the lombard down, it charges right up.
I fixed the logic board by reflowing it, but i still havent gotten around to replacing the screen to rid of those annoying lines. However, after reflowing the lombard's CPU card, all problems did go away. However I am going to have to get the PROPER AC adapter for it. as i cant charge the battery and run the machine at the same time. it wont charge and the AC adapter gets bloody hot.
Soon as i shut the lombard down, it charges right up.
What temperature did you use for the Lombard's daughtercard?
I used a basic old heat gun.
I removed the plastic rim that sits around the CPU, and preheated the toaster oven to 200deg F. yea, a little low, but significantly reduces the chance of a board warp/crack during heat.
Make you an aluminum foil heatshield to set over top of the CPU card, and ONLY EXPOSE the CPU. cut out another sheet for the bottom to expose the chipset.
Insert board in to oven and let the board preheat for roughly 10 minutes.
Remove board.
CPU side up, place your newly cut aluminum foil heatshield over top of the board, and align so only the CPU is exposed. this will protect all the vital plastic components and stuff from devastating heat.
Then take a wagner heatgun on low, held it over the IC about 3 inches away, and blasted it on low for about a minute.
Switch it over to high, and move to 4 to 5 inches away. Blast the IC for at least 5 to 15 seconds. This should do it. reseat any cracked balls.
Let it cool down for a little bit, flip it over, apply heatsheild and align to only expose chipset. Repeat the same procedure as above.
This worked for me... your mileage may vary. any time you get flaky motherboards that are newer style, this may indeed fix your issues. Newer junky PC laptops like acer, gateway, etc... all have some sort of motherboard problem, and this reflow on the CPU socket, and chipsets fixes them about 90% of the time.
Remember the ibook epidemic? yea....
Your mileage may very, as i don't have exacting equipment to measure temperatures. YET..... But we have a dedicated toaster oven that I call the reflow oven. this is what I used. I picked it up at a yardsale for $5 bucks, so what the heck.. its big enough to hold most digital boards, and small enough to use on a benchtop.
Once i get some good thermocouples i will eventually automate this. using whats called the "reflow curve" and program a microcontroller to run the oven, and the heatgun, and just make a pipe into the oven where the heatgun connects. and just line the board up with the pipe of what im gonna reflow. then the microcontroller turns the oven on, and warms it up to a proper temperature, and then blasts the heatgun to the proper temperature, and cools everything back off at the proper rate.
But yea, maybe someday when im not so busy.... OH... I use the heatgun to remove all my old SMD caps from mac logic boards. You just have to watch the plastic parts and where you direct the heat.
I removed the plastic rim that sits around the CPU, and preheated the toaster oven to 200deg F. yea, a little low, but significantly reduces the chance of a board warp/crack during heat.
Make you an aluminum foil heatshield to set over top of the CPU card, and ONLY EXPOSE the CPU. cut out another sheet for the bottom to expose the chipset.
Insert board in to oven and let the board preheat for roughly 10 minutes.
Remove board.
CPU side up, place your newly cut aluminum foil heatshield over top of the board, and align so only the CPU is exposed. this will protect all the vital plastic components and stuff from devastating heat.
Then take a wagner heatgun on low, held it over the IC about 3 inches away, and blasted it on low for about a minute.
Switch it over to high, and move to 4 to 5 inches away. Blast the IC for at least 5 to 15 seconds. This should do it. reseat any cracked balls.
Let it cool down for a little bit, flip it over, apply heatsheild and align to only expose chipset. Repeat the same procedure as above.
This worked for me... your mileage may vary. any time you get flaky motherboards that are newer style, this may indeed fix your issues. Newer junky PC laptops like acer, gateway, etc... all have some sort of motherboard problem, and this reflow on the CPU socket, and chipsets fixes them about 90% of the time.
Remember the ibook epidemic? yea....
Your mileage may very, as i don't have exacting equipment to measure temperatures. YET..... But we have a dedicated toaster oven that I call the reflow oven. this is what I used. I picked it up at a yardsale for $5 bucks, so what the heck.. its big enough to hold most digital boards, and small enough to use on a benchtop.
Once i get some good thermocouples i will eventually automate this. using whats called the "reflow curve" and program a microcontroller to run the oven, and the heatgun, and just make a pipe into the oven where the heatgun connects. and just line the board up with the pipe of what im gonna reflow. then the microcontroller turns the oven on, and warms it up to a proper temperature, and then blasts the heatgun to the proper temperature, and cools everything back off at the proper rate.
But yea, maybe someday when im not so busy.... OH... I use the heatgun to remove all my old SMD caps from mac logic boards. You just have to watch the plastic parts and where you direct the heat.
But you know, What happens when you cant buy another CPU card? Thats the way I look at things. Yes when replacing is easier, you always wanna plan for stuff like this, because nothing lasts for ever, and neither does parts. ive learned this the HARD way repairing TVs.
Example: Flybacks... Are you going to walk into a parts store and purchase a flyback for a 1951 emerson tube type tv? Are you going to walk into the apple store and get a new flyback for my mac SE? probably NOT.... Thats the next project that i plan to tackle is rewiding/making new flybacks. I already mastered regular transformers. but not IHVTs yet. There may not be flybacks for certain things anymore, but there will always be a schematic with specifications (in most cases, other cases is just plain testing on working equip), and there will always be magnet wire. hehe.
Yea, I have a lappy with a 'defective' Nvidia GPU. (remember G86 and G84 core issues from 2007?(even macbooks were affected)) and a quick bake fixed it for now.(so then I watercooled it
)A similar bake, however, did not fix my Lombard daughercard. I guess I should try a bit higher temp since I already have nothing to lose.
no, the lombard daughter card needs some crazy heat. as did the imac G5.
but ive seen low heat conditions as you described, same with the xbox360, heat as low as a hairdryer can fix some of them, probably do to hairline fractures in the solder, and the heat causes the expansion. but itll never be correct until a full reflow could be done.
but ive seen low heat conditions as you described, same with the xbox360, heat as low as a hairdryer can fix some of them, probably do to hairline fractures in the solder, and the heat causes the expansion. but itll never be correct until a full reflow could be done.
It looks like the lombard daughtercard is working now. hehe
With nothing to lose I stuck it in at roughly 475C (my asus g2s-a1 bake was 430C and previous lombard bake was 440C) annnnnd one of the cache chips was, er, fully detached from the board.
So, meh, it is clean and I stuck it back in the lombard and it boots! It is able to boot and run Mac OS X; it was unable to before.
At this point I have been unable to verify if it is running on half cache or if it detected a chip missing and disabled the L2. I have no working HDD that I know of to stick in here so I have to operate on CDs running from a 24x TL iMac CD-ROM drive and I am limited to OS 8.6/9 and no higher than 10.3. >_>
OS 9 also did not like my 4GB HFS+ formatted USB flashdrive(that I use with my Beige G3, so it is good). :/
A mostly success! Now I just have to find out what it has detected the cache as. ..there has got to be a way to see that in OF isn't there?
*Edit
Ok, no benches but looks like no cache.. Perhaps that means it will clock up even higher! muahahahahaaa
With nothing to lose I stuck it in at roughly 475C (my asus g2s-a1 bake was 430C and previous lombard bake was 440C) annnnnd one of the cache chips was, er, fully detached from the board.
So, meh, it is clean and I stuck it back in the lombard and it boots! It is able to boot and run Mac OS X; it was unable to before.
At this point I have been unable to verify if it is running on half cache or if it detected a chip missing and disabled the L2. I have no working HDD that I know of to stick in here so I have to operate on CDs running from a 24x TL iMac CD-ROM drive and I am limited to OS 8.6/9 and no higher than 10.3. >_>
OS 9 also did not like my 4GB HFS+ formatted USB flashdrive(that I use with my Beige G3, so it is good). :/
A mostly success! Now I just have to find out what it has detected the cache as. ..there has got to be a way to see that in OF isn't there?
*Edit
Ok, no benches but looks like no cache.. Perhaps that means it will clock up even higher! muahahahahaaa
I had no idea that you could reflow with a toaster oven and a heat gun as you've described. Thanks for the info, it will definitely save me some parts in the future. I do still have my old Pismo CPU board with bad cache, maybe it only needs a reflow.
Ok, I officially need to get some nice big cache chips and attach them! xx( This is silly.
CPU:
http://poopr.org/images/o3dfzhxw25p8mhs70bw.jpg
GFX:
http://poopr.org/images/3aylm1iqqeojj5r3gx2x.jpg
Hmm... perhaps the cache chips from my old 266MHz PDQ card might do? It went down and I suspect the CPU itself was acting up and not the cache. I /should/ bake it.. but meh I already have a 300 for it. I wonder if they might be too slow. I will need to look into that I guess.
CPU:
http://poopr.org/images/o3dfzhxw25p8mhs70bw.jpg
GFX:
http://poopr.org/images/3aylm1iqqeojj5r3gx2x.jpg
Hmm... perhaps the cache chips from my old 266MHz PDQ card might do? It went down and I suspect the CPU itself was acting up and not the cache. I /should/ bake it.. but meh I already have a 300 for it. I wonder if they might be too slow. I will need to look into that I guess.
Silicon Chip magazine did a really good project writeup on setting up an off the shelf toaster oven for this kind of work. There's probably oodles of other writeups around the web.
Short version of theirs: stick a slab of aluminium in there for your board to rest on, attach an accurate and fast temperature probe (like a thermocouple and a multimeter) to the slab. Do a couple of test runs without a board in there and graph the temp vs time. If you're in luck as they were, when you turn the oven on full, it heats up nicely within the "reflow curve xxx mentioned, then when it hits the bottom of the recommended temp range you can turn it off and it'll coast up into the flow zone, peak and then cool down at about the right rate to stay inside the curve. If the delta-t strays outside the curve (or for extra geek points) add a micro to automate it.
This is also a good way to recover useable SMT components from old or dead boards, aka "bake and shake" :lol: