Macintosh IIsi Magic chime sound. What does it mean?

[JC8080]

@zezba9000 ,@9166188 , @elemenoh , did any of you ever solve your death chime? I have the same issue with a board I just recapped. I recapped the board and PSU and it worked fine for a day. I had it running and a breaker tripped, I think because of my 13" CRT. When I restarted the machine I have the chime followed by the death chime. I also get the brief grey screen right after I switch it on. It did boot one time but showed display issues on the screen, it has not booted since. I have swapped PSUs with a known-good unit, removed the RAM SIMMs, and re-flowed the onboard RAM chips. The fact it did boot but had display issues makes me suspect the onboard RAM, since the VRAM is shared with the RAM.

My next step is to swap the RAM chips off another non-working board I have, though unfortunately the condition of those chips is unknown. @imactheknife mentioned in another thread that the onboard RAM/VRAM is often an issue. Swapping the RAM on this board is a bit of a job, so I wanted to check and see if anyone had found a different cause before I jump into the project.

 

[imactheknife]

@zezba9000 ,@9166188 , @elemenoh , did any of you ever solve your death chime? I have the same issue with a board I just recapped. I recapped the board and PSU and it worked fine for a day. I had it running and a breaker tripped, I think because of my 13" CRT. When I restarted the machine I have the chime followed by the death chime. I also get the brief grey screen right after I switch it on. It did boot one time but showed display issues on the screen, it has not booted since. I have swapped PSUs with a known-good unit, removed the RAM SIMMs, and re-flowed the onboard RAM chips. The fact it did boot but had display issues makes me suspect the onboard RAM, since the VRAM is shared with the RAM.

My next step is to swap the RAM chips off another non-working board I have, though unfortunately the condition of those chips is unknown. @imactheknife mentioned in another thread that the onboard RAM/VRAM is often an issue. Swapping the RAM on this board is a bit of a job, so I wanted to check and see if anyone had found a different cause before I jump into the project.

If you have a microscope or a good magnifying glass look at the legs of vram chips near where old caps were, and same for the onboard ram (usually just the one chip) and also check traces around area. The ones closest to the rom chips can have bad traces. Snap some high res pictures of the board around cap areas

 

[JC8080]

If you have a microscope or a good magnifying glass look at the legs of vram chips near where old caps were, and same for the onboard ram (usually just the one chip) and also check traces around area. The ones closest to the rom chips can have bad traces. Snap some high res pictures of the board around cap areas

I thought the IIsi shared it's 1mb onboard RAM with VRAM, and there was no separate VRAM. Is that incorrect, and there are separate VRAM chips? I will take an extra close look all around the caps with my microscope. I'll also use a dental pick to check for movement on the legs of any chips.

I also found some working 80ns 1mb SIMMs in one of my other machines, someone else mentioned successfully using the chips off a 1mb SIMM to replace the onboard RAM. I am planning on doing that since the RAM seems like a potential cause.

 

[imactheknife]

I thought the IIsi shared it's 1mb onboard RAM with VRAM, and there was no separate VRAM. Is that incorrect, and there are separate VRAM chips? I will take an extra close look all around the caps with my microscope. I'll also use a dental pick to check for movement on the legs of any chips.

I also found some working 80ns 1mb SIMMs in one of my other machines, someone else mentioned successfully using the chips off a 1mb SIMM to replace the onboard RAM. I am planning on doing that since the RAM seems like a potential cause.

Will take a picture of chips i had issues with and had to replace. I should not say vram as your correct. They do have something to do with video i am sure. In These pictures i had to replace those chips as some legs had literally disintegrated when touched. If you need spares i have some. If you remove them make sure you check for broken traces etc i had some broken ones arounf those areas on one iisi board.

 

[JC8080]

Will take a picture of chips i had issues with and had to replace. I should not say vram as your correct. They do have something to do with video i am sure. In These pictures i had to replace those chips as some legs had literally disintegrated when touched. If you need spares i have some. If you remove them make sure you check for broken traces etc i had some broken ones arounf those areas on one iisi board.

Thanks for the photos. I checked UE5, UF5, UG5, and UH5 with my microscope and a pick. Some of the legs on UF5 were loose, though none of the legs broke off, they all seemed pretty sturdy. I removed all four chips and cleaned the pads well. I looked for broken traces under the chips with my microscope and everything looked good. I've replaced two of the chips so far and ran out of time, I'll get the other two back on in the next day or two and see what happens.

If re-soldering those chips doesn't fix it, I'll move on to replacing the onboard RAM. I also need to take a close look around all the caps with my microscope to look for broken traces, and take a pick to the legs of any other grungy looking chips near the caps and see if they are loose.

 

[JC8080]

I went over the board with my microscope this morning, I took photos so I can sit down and take a detailed look, however at first glance all the traces on the board look very nice, no signs of damage. That's the good news.

The bad news... Some of the pins on the CPU nearest C34 looked pretty grungy, so I decided I'd to a little scraping and poke at the pins a bit, and the first one I pressed on broke free from the board. Fortunately, it still seems well attached to the CPU. So now I'll look for advice from people with more soldering experience with me.

I know the best thing to do would be to remove the CPU and re-solder it. I can do SMD IC work on chips like the RAM and UE5 chips, but I haven't tried any finer-pitch ICs, and the CPU is pretty fine-pitch. My tools at hand are a nice iron and a hot air station. From a couple videos I've seen, the CPU could probably be removed with ChipQuik, and re-soldered with solder paste, but I have never done this.

The other though I have is to scrape the legs and pads to remove as much corrosion as possible, add a bunch of flux and re-flow. Does this seem like a reasonable thing to try? I would be slightly worried about damaged traces under the CPU, however all the traces in other high-corrosion areas by caps held up extremely well with no visible damage, so I am hoping any traces under the CPU are in good shape too.

Edit: added second zoomed-out photo of CPU

 

[Phipli]

From a couple videos I've seen, the CPU could probably be removed with ChipQuik, and re-soldered with solder paste, but I have never done this.

Don't do this, just solder the free pins in situ, and don't put chipquik near anything that you're not going to remove and clean up.

Best thing to do is reflow the solder that is already on the board with a load of no-clean flux and an iron.

I've had to do this a few times. Latest was this audio card that someone had damaged :



After :

The other though I have is to scrape the legs and pads to remove as much corrosion as possible, add a bunch of flux and re-flow. Does this seem like a reasonable thing to try?

Exactly this, only don't scrape the pins unless flux and heat doesn't clean up the pins, because it is extremely easy to bend QFP pins, and ask @mg.man about his Centris 610. It can be very very difficult to bend the pins back, especially if the pads are proud of the PCB.

 

[Phipli]

Use a scalpel to bend the leg back. You might have to lift it slightly away from the chip to get it to lift over the remaining solder on the pad, then push into the... erm... ankle, of the leg a tiny bit to pre-load the pin so when the solder melts the foot goes and stays down.

Do not put too much heat into the board or the traces come free. Don't hold the iron on for more than 2 or 3 seconds without letting it cool.

You don't need a tiny iron tip, but lots of no clean flux is mandatory.

 

[JC8080]

Thanks Phipli, I will give that a shot. I already gently bent the loose leg back into place, and I have plenty of MG Chemicals no clean flux on standby.

 

[mg.man]

and ask @mg.man about his Centris 610

^ what he said... it's still out of action - too much work / life in the way.

 

[JC8080]

Use a scalpel to bend the leg back. You might have to lift it slightly away from the chip to get it to lift over the remaining solder on the pad, then push into the... erm... ankle, of the leg a tiny bit to pre-load the pin so when the solder melts the foot goes and stays down.

Do not put too much heat into the board or the traces come free. Don't hold the iron on for more than 2 or 3 seconds without letting it cool.

You don't need a tiny iron tip, but lots of no clean flux is mandatory.

Thanks for the advice, the repair seems to have re-connected all legs. I put a bit of side pressure on each leg one with a pick and none were loose any more. I can not currently test the repair since there are other issues with the board, but I checked continuity from the top of each leg to the pad and all are connected.

 

[JC8080]

Death chime fixed! Hopefully this info helps other people with the same issue. Steps taken:

Re-flow all the onboard RAM, the legs of the chips closest to the leaking caps showed corrosion. This did not fix the problem, but these corroded legs might have been a contributing factor.

Next, when I was checking over the board with my microscope, I found a couple legs on the CPU were loose (see posts above). I re-flowed all the grungy looking pins and checked for continuity from the top of the leg to the pad, to ensure my reflow job was solid. I did not try starting the machine at this point.

I also noticed some loose legs on some of the ALS245 chips (UE5, UF5, UG5, and UH5), I removed all four chips, checked for broken traces around/underneath (there were none), and re-soldered. At this point I meticulously checked continuity from every pin of each ALS245 to it's destination (other ALS245s, onboard RAM, RAM SIMM sockets, ROM SIMM socket, CPU, UJ3 chip) and confirmed all were good.

At that point I booted the machine - the death chime was gone, just the normal chime remained, and it seems to be functioning well. Audio works, keyboard/mouse work, SCSI works, video works, NuBus video card works. I still need to go in and poke at the pins of the other grungy looking chips on the board to see if any are about to come loose, and maybe re-flow any that look suspect. But at the moment everything is working!

Thanks to @imactheknife and @Phipli for the guidance.

 

[JC8080]

Unfortunately, shortly after my last post the IIsi went back to the death chime, and I haven't been able to figure it out yet. Steps taken since above:
- replaced onboard RAM with chips from 4x, 2-chip 256kb SIMMs
- replaced all ALS 245s with new parts from DigiKey
- removed and re-installed CPU, ensuring that there are no loose pins, and no bridged pins

I still have the same behavior - normal chime, followed quickly by the death chime (probably a half second delay).

I am at a bit of a loss here, I'm not sure where to go next. I am thinking of removing, cleaning, and re-installing the chips up in the corner by the power switch, those caps leaked and the pins on all those chips look pretty grungy. This is not my photo, it is from a post by @joshc, but that is the same area I am looking at:



Does anyone have any suggestions? From what I have read people always point to the RAM or ALS 245s when there is a death chime. Now that I have ruled those out, I'm not sure what the next steps should be.

 

[imactheknife]

I would focus on areas in circles. Als 245, the ram chip and the one rom chip

Why did you feel like you had to take processor off??

 

[JC8080]

I would focus on areas in circles. Als 245, the ram chip and the one rom chip

Why did you feel like you had to take processor off??

I'll take another look for broken traces and do some continuity checking on the ALS 245s and RAM chips in that corner. I'll take a close look at that ROM chip too.

The CPU had a few pins that were loose in the corner by the caps. I was eventually able to get them all soldered back down, but they were really ugly and I wasn't confident they were making good contact. It was probably unnecessary, it was mostly for peace of mind to eliminate one possible source of problems.

 

[imactheknife]

almost all of the iisi boards i have had and recapped had issues with two of the als245 one at each end near caps, the one memory chip closest to cap and the one rom chip. Also check traces near the cap near processor. Had bad traces there too. These were some of the worst offenders boards i have had

 

[rikerjoe]

Since it appears you had corrosion near the PSU connector on the logic board, check continuity from the connector pins to various locations on the logic board. I had intermittent death chimes with a IIsi caused by broken traces underneath the PSU connector.

 

[JC8080]

Since it appears you had corrosion near the PSU connector on the logic board, check continuity from the connector pins to various locations on the logic board. I had intermittent death chimes with a IIsi caused by broken traces underneath the PSU connector.

Thanks, I hadn't thought of looking there.