| Click here to select a new forum. |
| Smoke out of the SE/30 Video ROM |
Posted by: chrisrueckert on 2026-04-04 13:09:54
So, this SE/30 had smoke coming out of its video ROM (see cracked/dark spot on the attached picture).
What I have done so far:
- Analog board and PSU recapped and successfully tested with another SE/30 logic board
- Recapped the SE/30 logic board - there are still broken traces around the Sony chip that need to be fixed.
I don't suppose these broken traces cause the video ROM to overheat. However, I'm wondering if this symptom could be related to other broken traces? Could the video ROM just be defective and the Mac would work with a new ROM? I doubt, however, let me know what you tink. |
Posted by: Durosity on 2026-04-04 13:32:54
So, my educated guess is that the soldering around C10 has shorted a 5v line to a 12v line. I have a vague recollection of someone else doing that around here. The problem is if that’s happened it may well have destroyed other components on the board too.
First thing I’d do is check for any shorts in that area. It’s particularly easy to do with tantalum’s due to the big connector they have that wraps around the component. |
Posted by: chrisrueckert on 2026-04-04 14:28:38
Thanks, there was indeed a solderbridge at C9! Now the Mac shows this. Would it be safe to replace the video ROM with one known working? |
Posted by: Durosity on 2026-04-04 14:37:27
I guess it depends on if there’s any other potential sources of an over voltage. Might be worth probing the pins from a known working board and then comparing the voltages to that to the broken one before swapping… though personally if it were me I’d probably just swap and pray to the flying spaghetti monster that it’s not gonna fry its noodle.. I wouldn’t recommend this course of action though 🤣 |
Posted by: Chopsticks on 2026-04-04 18:10:53
good practice might be to check for shorts after pulling that blown rom, so +5 to Ground, +12v to Ground and +5 to +12v.
personally after i do a full recap i normally just quickly check for power rail shorts before i power on. particularly due to the cap goo often eating away solder mask and these macs often run traces under the caps. being that you've changed from electrolytic to tants and they have different pad footprints its probably more so important to check for shorts. |
Posted by: chrisrueckert on 2026-04-05 12:26:28
Yes, checking for shorts seems very reasonable. Here's what I noted:
+5 to Ground: 140 Ohm
+12v to Ground: 53 Ohm
+5 to +12v 140 Ohm
It looks like there are many more shorts. I'll need to investigate further. |
Posted by: Durosity on 2026-04-05 12:55:30
Yes, checking for shorts seems very reasonable. Here's what I noted:
+5 to Ground: 140 Ohm
+12v to Ground: 53 Ohm
+5 to +12v 140 Ohm
It looks like there are many more shorts. I'll need to investigate further. It’s plausible that original bridge has damaged other components on the board and caused the shorts there, so don’t just look for other trace related problems. A thermal camera may come in handy here’s |
Posted by: croissantking on 2026-04-05 13:26:48
Looks like one of the VRAM chips is dead, from the unusual simasimac screen - it’s missing columns of data. |
Posted by: Chopsticks on 2026-04-06 19:55:48
those readings seem very low, id expect to see Kohm values.
wouldnt hurt to pull all the ram, the ROM and the socketed PAL chips out and see if theres any change to those power rails you measured.
maybe youll get lucky...
if memory serves correct C9 if on the negative -12v rail. im not sure what short you had that you fixed there but to start with i would check the schematics and focus on anything using -12v (probably only really the speakers op-amp). after that id move onto the +12v rail as that has the lowest ohm value from what you measured. |
Posted by: MIST on 2026-04-07 09:20:33
You're not alone ...
|
Posted by: croissantking on 2026-04-07 12:06:20
You're not alone ...
Adrian Black made the 12V mishap (in)famous |
Posted by: Trianon on 2026-04-07 12:15:26
Measuring power lines by Ohm will only show when there is a hard short, 1 ohm or below.
It's sometimes a good indication sure, but another chip revision, empty/full cap can give a big different.
Chips are active parts that need full 5V before they take full current.
Better to have a external power supply with current limit, you start at 10mA and go up, should get to the full 5V / 12V before full current is needed. |
Posted by: Chopsticks on 2026-04-07 17:28:45
Most things don't fail with a hard short, a partial short is much more common with digital Logic, 150 ohm to ground on the +5v rail and 53 ohm on the +12v rail are both clear indicators there is a partial short to ground likely inside the internal logic of one of the IC's.
those reading are well bellow what would be expected on each rail on a SE/30. should be kOhms on this particular board
many IC families will draw current both above and below their rated input voltage, their input gates might not switch correctly(or at all) but they'll still draw some power. it really depends on the type of device it is and how its input topology is designed. so i don't really understand what you're meaning there?
i do agree though where possible one should always current limit when working on/repairing something to minimize the risk of blowing a component. |
Posted by: superjer2000 on 2026-04-08 10:40:44
150 ohm to ground on the +5v rail and 53 ohm on the +12v rail are both clear indicators there is a partial short to ground likely inside the internal logic of one of the IC's.
those reading are well bellow what would be expected on each rail on a SE/30. should be kOhms on this particular board I can't speak to whether this is true for the SE/30 or not but when I was troubleshooting a IIci a number of years ago. I thought I had a short between 5V and ground when I was testing between those points on the logic board power plug with my multimeter but it turned out that isn't a valid way to do that measurement and the board was actually fine.
|
| 1 |