VOGONS

Common searches


First post, by athlon_p0wer

User metadata
Rank Newbie
Rank
Newbie

I've been tinkering around with old computers for years now, and I've never had a computer component smoke on me. It startled me quite a bit, and I unplugged the thing as fast as I could but the damage was already done.

The sound card started quietly crackling and shot out a fair bit of black smoke, making the living room smell strongly of ozone. I'm assuming this was my fault, as this was in the 486 system I modified in 2019. Long story short, one of the things I did involved slicing the riser card completely in half, sanding down the cut edge, and covering it with contact adhesive to seal it. I made sure that none of the traces were still sticking out and touching, but I must have made a mistake somewhere.

The Sound Blaster Vibra 16S I used in it for years has some missing pins on the edge connector, and my theory is that most of the traces in the riser card were okay, with just one or more shorted ones on the pins that never got used. The moment I put in the Yamaha YMF-718s, which uses all the pins, and turned it on, the shorted pins were used, and this was the result:

20221227_192214.jpg
Filename
20221227_192214.jpg
File size
1.49 MiB
Views
1430 views
File comment
The front of the card looks fine.
File license
Public domain
20221227_192231.jpg
Filename
20221227_192231.jpg
File size
1.37 MiB
Views
1430 views
File comment
The source of smoke and crackling
File license
Public domain
20221227_192258.jpg
Filename
20221227_192258.jpg
File size
1.15 MiB
Views
1430 views
File comment
Sadly, the riser card bit the dust as well.
File license
Public domain

Oddly enough, there was no damage on the motherboard anywhere. Once I removed the sound card, I double checked and made sure that black was to black on the AT power input, and that everything was hooked up right before I turned it back on, and it worked. It wasn't so smart to turn it right back on after it had just got done smoking, but I was panicking a little and wanted to know just how bad off it was or wasn't.

In retrospect, I should've left the thing alone. It probably wouldn't have done this otherwise, and the only thing wrong with the case it was made for was a broken corner piece of plastic.

Reply 1 of 10, by mkarcher

User metadata
Rank l33t
Rank
l33t

The burnt trace on the ISA riser card seems to be B3, which is the +5V supply. Obviously a very high current (tens of amps?) was conducted through it. On the sound card, it also is the +5V trace that is burnt. You get burn marks both on the trace from B3 to the configuration EEPROM (the 8-pin 93C66 chip) and from B29 to the configuration EEPROM. Both B3 and B29 are +5V pins. Possibly you did nothing wrong with the riser card, and the sound card was shorted already before you inserted it to the riser card. Primary suspect would be the ceramic multilayer capacitor CB14 next to the 93C66 chip in that case. Those caps tend to fail short if they are mechanically stressed too much.

The damage is not as bad as it looks. You burnt some of the solder resist lacquer, you can apply nail polish as replacement to restore insulation. Those traces are really wide, so if the trace is broken, it's again easy to fix it. So a repair plan would go along like this: Check for continuity of +5V on the riser card, check for shorts from +5V to GND. If everything is fine, clean the burn marks and apply nail polish. Check the sound card for shorts from +5V to ground. If the sound card is shorted, try to remove CB14. If you can get the short removed from the sound card, it's again cleaning up and applying nail polish. In case the short on the card originated form a shorted capacitor labelled CB, it should be replaced in the long term by a new 100nF multilayer ceramic capacitor.

Reply 2 of 10, by pentiumspeed

User metadata
Rank l33t
Rank
l33t

Smoked the IDE cable once with my first 386 PC back in the day in full view was connected wrong, and nothing affected anything, replaced the cable and all good. Lucky me.

Blew the dying hard drive's motor in beautiful ring of smoke billowed out around the motor's rotor, as power load on a malfunctioning power supply, I was too inexperienced to diagnose and find the source in 1991-ish.

Cheers,

Great Northern aka Canada.

Reply 3 of 10, by athlon_p0wer

User metadata
Rank Newbie
Rank
Newbie

Thank you mkarcher, I'll see if I can get a multimeter and test that out. Best case scenario is that the riser card is fine, and I can use it to put a sound card back in the machine (likely another Vibra 16s). I've never tried soldering before, so I'm not sure that I could replace the cap on the sound card or patch any traces. All of this comes after me figuring out why the thing hasn't been detecting new in box brand name CR2032 batteries (and this machine is one of those that acts up if it doesn't have one). I wouldn't even care if the PC couldn't ever use a sound card again as long as the base system works. The CR2032 issue has been going on for a while now, and I'm stumped.

I didn't know IDE cables could smoke when plugged in backwards. I've killed a drive and motherboard by doing that though, my first and last Quantum Bigfoot drive (not like the thing would've worked for more than 6 months anyways), and my Intel SE440BX-2, which I only paid US$40 for, and was my favorite Slot 1 board out of the ones I've used. It now sells for US$ 150+. I've become very paranoid about making sure the red stripe corresponds with pin 1 on the drive when using IDE cables without keys.

I suspect there may be something off about the PSU in the 486. Sometimes the machine turns itself on if left plugged in, and this happens even if it's only been plugged in and off for ~10 minutes. The PSU itself is from 1995, and has never been cleaned thoroughly because I don't have a torx security bit to open it. One owner before the guy that sold it to me seems to have been an automotive shop gauging by the amount of chemical scented black dust that was all over the inside of the computer (and is still in the PSU), as well as software to prepare you to take automotive certification tests on the HDD.

Reply 4 of 10, by KCompRoom2000

User metadata
Rank Oldbie
Rank
Oldbie

In 2013, I was moving my Athlon 64 build into another case since the original case broke down a few years prior and the case I was using after that one (a cheap $20 Diablotek case from eBay) was too cramped and was pretty flimsy. I managed to plug the floppy drive power cable in incorrectly and when I turned it on, the graphics card did a continuous beep and smoke came out from where the power connector was on the floppy drive. I had to replace the power supply and floppy drive. Funnily enough, I went through two different cheap power supplies (a 480W Logisys and a 500W Apevia) before I settled on the one I'm using with it now (a 600W EVGA). The Logisys PSU lasted for a year before it died completely, and the Apevia one kept causing the hard drive to restart itself (possibly bad SATA connectors). Finding a suitable replacement that wasn't complete garbage was tough because I needed at least two Molex connectors (for the disc drive and graphics card) and a floppy drive connector.

Reply 5 of 10, by athlon_p0wer

User metadata
Rank Newbie
Rank
Newbie

I've had a PSU die on me while I was playing Garry's Mod on my main rig back in 2015. It stunk of ozone badly, but never smoked and also never damaged any components. Can't remember the brand, some cheap thing that came out of my mom's old gaming PC she bought from a local computer store. Of course they cheaped out on the PSU, why wouldn't they?

I've been lucky enough to have never plugged in a power connector wrong, knock on wood. I remember when I started working on the 486 (my first ever AT system), I was extremely paranoid about making sure black was to black because I had heard too many stories of what would happen if it wasn't.

Reply 6 of 10, by schmatzler

User metadata
Rank Oldbie
Rank
Oldbie

It's been many years, I had a P3B-F motherboard back when it was still state of the start.

It was running a little hot so I decked out my case with 3-4 fans and populated all of the fan headers with those.

The board did not like that. It smoked heavily, all of the fans stopped and yet, it still continued working.

I used it for a few years but the fan controller was toast. Had to power my fans via IDE power adapters directly.

I can't remember any other instances of magic smoke, so I guess I'm pretty lucky.

"Windows 98's natural state is locked up"

Reply 7 of 10, by athlon_p0wer

User metadata
Rank Newbie
Rank
Newbie

Why would they put all those fan headers in there if the controller couldn't handle it? I've had other 440BX boards with three fan headers and ran all three (CPU, intake, exhaust) at the same time and never had any issues, so it must be something specific to that board, maybe the fan controller just wasn't enough for all the fans. You got lucky that when the fan controller smoked, whatever was happening didn't damage the rest of the board.

Reply 8 of 10, by Tetrium

User metadata
Rank l33t++
Rank
l33t++

I remember the first time I let out the magic smoke! I hadn't even gotten into retro computing yet.

An old friend of mine had an old Pentium MMX, something like a 166MHz overclocked to 200MHz or so, with something like 40MB of ram and a PCI Voodoo 3.
He had his system completely outside of the case so he could overclock it better.

At some point as we were overclocking his CPU I got the luminous idea to make the fan spin the other way around, perhaps that would improve the cooling somehow!
And as we didn't want to screw the fan from the heatsink, I thought it was a great idea to simply plug the fan connector to the motherboard backwards to make the fan spin in the different direction! Damn the plug wouldn't fit, oh well I'll just force it in what could possibly go wrong! 😁
Well, magic smoke and our little overclocking adventure ended prematurely.🙄

Thankfully nothing else got damaged. I think he ran his system like this without a CPU fan for a while completely outside of the case until he got an AthlonXP.

schmatzler wrote on 2022-12-31, 02:10:
It's been many years, I had a P3B-F motherboard back when it was still state of the start. […]
Show full quote

It's been many years, I had a P3B-F motherboard back when it was still state of the start.

It was running a little hot so I decked out my case with 3-4 fans and populated all of the fan headers with those.

The board did not like that. It smoked heavily, all of the fans stopped and yet, it still continued working.

I used it for a few years but the fan controller was toast. Had to power my fans via IDE power adapters directly.

I can't remember any other instances of magic smoke, so I guess I'm pretty lucky.

May I ask what kind of fans did you try to use?
It may be possible that you added newer fans with more power draw than your board was able to handle?
Back when these boards were made, 8cm fans were barely a standard yet, let alone larger ones.

Whats missing in your collections?
My retro rigs (old topic)
Interesting Vogons threads (links to Vogonswiki)
Report spammers here!

Reply 9 of 10, by Yoghoo

User metadata
Rank Member
Rank
Member

Haven't had a magic smoke experience for a long time. But had it today with a NOS AOpen IDE CD drive. When I powered it up I immediately smelled something was wrong.

Strange thing it was a chip and not a capacitor or resistor. First time for me but fortunately nothing else was damaged. But I can throw the CD drive away. 🙁

Attachments

  • chip.jpg
    Filename
    chip.jpg
    File size
    77.84 KiB
    Views
    1025 views
    File license
    Fair use/fair dealing exception

Reply 10 of 10, by Shagittarius

User metadata
Rank Oldbie
Rank
Oldbie

I've described this elsewhere on Vogons but my first encounter with magic smoke was troubleshooting my old 386 and I don't even know how it happened but the Soundblaster released its magic smoke. I can't remember but it was probably a SB16 with CD rom interface attached. Anyways I reset the card in the motherboard and it fired up fine. Never had a problem with that PC or the board again, not sure what I smoked but it must have somehow been for something I didnt use.

Oh I should mention I was drinking heavily that day.