I really wonder how the owner made that happen, maybe left hot air unattended on the side of the board for a while or something, but yeah clearly all those caps are dead/dying!
People tend to stick hardware in the oven, without thinking for themselves what might be some issues that could occur. I’d definitely replace those capacitors. Usually, they’re resistant to heat, anywhere from 85°C to 105°C. The molten connector tells me the temperature went higher than that. Evidently, those capacitors will dry out and/or pop. Nothing is unfixable. :)
I'd say the majority of the caps are probably cooked, I see a fair few that look suspicious. Even some of the ceramic capacitors look suspicious, look at C16 to the right of the CT1745A chip, it's really dark compared to the ones next to it, like it's getting smoking hot from being shorted. This card may work again, but it's going to need some serious low level diagnostics.
If the computer isn't posting, then the sound card must be doing something pretty catastrophic signalling on the ISA bus. Maybe it's time to resort to an oscilloscope and/or logic analyser to see how it's effecting signalling on the ISA bus. That will help to narrow down where the fault might be.
use a thermal camera (if you have one) to see if any more caps or other components are shorting. also i think it's pretty hard to damage ICs with heat. unless the card was stored in a humid environment and the person heated it very quickly or went above 300-350C. This is called popcorn effect and the bigger the IC the bigger the risk. However I would still recheck for a short somewhere since it stops the whole system from booting..
have you tested for shorts? it could be any one of them surface mount caps and resitors have a hairline fracture from the heat. but i say its that audio chip the pins probably sent too much voltage to the wrong place. dame shame. would be nice to see you save the card somehow.
Recap the whole board, just, all of it, because the more caps I look at, the worse it gets. Some of them look like they'd explode of you sneezed in their vicinity. After that I'd check for bridging with any of the other chips/connectors, because there's no telling what might have gotten messed by by that heat. Lastly, I'd test to see if that crystal is still okay.
I would check resistance to ground on all power lines into the card, as well as the data lines. I wouldn't be surprised if the onboard power supply of the card is shorting out, or some line is being pulled down. If you have one, maybe a thermal camera will be a quick way to see if a component is shorting, but you can get away without one, it will just take more time and effort. I think there is a lot more that can be done in terms of diagnosing before completely condemning this card. Obviously, commercially it makes no sense to spend this much time and effort on the card, but for greater justice and youtube content I think it will be more than worth it!
Exactly my thought. Although, i would not waste time for cheeking data lines at this stage, until the system is at least able to boot up. Any IC malfunction would not prevent system boot, this has to be a short on a card. So, i would start with checking power lines on the SB's ISA connector.
check for shorted ceramic caps; heat stresses can cause the ceramic to fracture and short the layers together. since they're usually used as bypass caps, the thermal gradient and stresses they endure are much higher due to the ground planes sinking all the heat away on one leg.
These cards had a serious problem with the PNP EEPROM corrupting. When that happens it stops detecting on the creative PNP utility TSR and Windows drivers can't find the card. Copying EEPROM data from a working card will save it if that is the problem. Maybe this card had that problem before it pop'd.
Everyone commenting about shorts -- if there were shorts between a power rail and ground, the power supply would not have stayed on. The overcurrent protection would definitely have kicked in and immediately shut the supply down. So I highly doubt there's a short between power and ground anywhere.
You're absolutely correct. And that overcurrent protection works like a charm. When I built my first -5V Adapter, one of the tantalum caps was bad and immediately caused the power supply to turn off.
There's definitely not a short between power input and ground. Shorts between two signal lines or between a signal and power/ground are much harder to detect.
So, turn of events: there was a short between two signal lines. The card posts now and also plays digitised music. Unfortunately, no FM/MIDI sound. Will try the card in an older system with pure DOS since the drivers have issues with win98 DOS
@@bitsundbolts Are you aware that on some/most AWE32 cards, FM is routed through the EMU8K synthesizer/effect processor and *requires* initialization using AWEUTIL /S in DOS (or installation of an AWE32 wavetable driver in Windows 95)? "No FM" might be just due to missing initialization.
In the electronics assembly industry, there is a term called "bake". It refers to some components needing to be "baked" before assembly. Baking removes moisture from inside of IC chips that could expand during reflow soldering and destroy the parts. I suspect the "pop" that was heard during hot air rework was from just that. During the baking procedure components are slowly heated to much lower temperatures than reflow soldering to gently remove the moisture. This is also why you sometimes receive parts from DigiKey etc. in sealed plastic sleeves with silica gel pouches inside. They are moisture sensitive (to some MSL level).
Excited to see my old card even if it's still not working. If you look at my first post on Vogons, the computer was only able to POST and detect the card one time, when it gave me the error. After that, the computer wouldn't POST unless I changed a jumper but it wouldn't detect the card.
What were you hoping to achieve by blasting the thing with hot air? Genuinely curious. Hot air isn't going to unbend pins that are touching, for example.
Oh well, you can't save them all. Even with your perfect restoration methods, some parts are just beyond repair. I fear the same is also the case with the crispy Monster 3D II, but there's only one way to find out. 😄 It is unfortunate that the AWE32 probably met its final demise due to the ill-advised layman repair attempt, but the previous owner probably took the "it's broken anyway" approach and didn't want to sink any cost into a repair, which we can't really fault him for either. It is a good candidate for the next level of diagnosis though. Knowing you, you are probably alrady itching to dive into EE level diagnosis. 😉 Who knows, maybe you'll get lucky and the remaining fault(s) are limited to generic parts. 🤞🏻 Happy Easter everyone!
Great video! Thank you for sharing your adventure attempting to repair this sound card. :D Hmmmm I need to see if I can find RAM for mine so I can upgrade it.
I would use a thermal camera to figure out which components are getting too hot. Also, it might be worth checking out the 74 TTL chips, such as the 74LS245. It is better that it is one of those chips that is damaged rather than the proprietary ones. And, as others have already suggested, check for shorts and the rest of the capacitors.
It could just be me having just done a full recap on a motherboard and PSU, but to me a lot of the rest of the caps on that card look like they're bulging as well. If the previous owner just air gunned the whole card, he may have blown or come close to blowing some of those other caps as well. I'd just try a full recap before giving up on it; caps are cheap. And maybe test for continuity/shorts around the components near where the hot air went.
Lots of el caps seems to be bulging, so I'd go for a full re-cap. If that won't work check if the oscillator generates anything. Then look for possible ceramic caps short.
Decades ago we were told not to get parts too hot or they will die, but if you watch Sorin blasting parts with very high temps, ie 380ºC, it doesn't seem to matter. Me personally I would keep trying, unless I found an unreplaceable part burring hot when powered, ie, shorted. You could check for shorts, diode check data lines, check if power is on the card, pull the IC totally and resolder, etc
Maby measure all caps (ceramic and electrolythic) for shorts and replace all the electrolythic ones anyways for good panasonic, nichicon or rubycon ones in case the card does work again. They are afterall almost 30 years old anyway.
Yeah I see other bulging caps and I'd check that crystal oscillator since it's close to the main chip where heat would of been applied. You're going to have to check the ceramics for shorts and see if there any off value or open resistors.
so sad! Probably a shot in the dark but I might see some bulging on the caps on the left hand side of the melted connector. I doubt they would prevent the whole system from posting but worth a shot. Actually, I'm wondering how is it possible that the card is causing that issue. If you have a scope you could scope all the address/data lines as I feel that at that stage a shorted Address/Data line is the only way a SB could prevent a Motherboard from posting. From the video, it seems that the ISA bus is directly connected to the main IC so that's unfortunate. Other things I can think of: - Check that everything has voltage - Check the clock - Check for shorts - Check that nothing is scorching hot :)
Pro repair tips: First check for a short between power and ground. There are several common logic chips on the board being used to isolate several buses (74LS244/74LS245), plus a few other smaller logic chips and a DRAM next to the large Creative chip. I would remove and test all them in a Retro Chip Tester Pro. If you don't have a RCT just remove and replace them with new parts or known good used parts. The melted connector is in the audio output section so that won't stop the computer booting but a bad logic chip connecting the card to the ISA bus certainly will cause an issue if the chip(s) are bad. Due to the custom chips the only thing you can do is check/replace any common off-the-shelf chips. If that doesn't work then it's scrap and a lesson to all noobs watching... waving a hot air station over a board like a magic wand won't do anything except break it. You can't fix it with a hammer. Do not mess with expensive and complicated electronics that you do not have the required skill levels and equipment to handle.... PRACTICE ON JUNK BOARDS UNTIL YOU GET THAT SKILL LEVEL!!!
What i think is that everybody who tried or learn to repair electronics is that sometime and somewhere have broken something that was broken but could be fixed :). I hope at least he learned!
Im always impressed by your skills working on such a tiny scale! Bummer that it didn’t work out. Maybe some of the logic ICs that are straight on the ISA bus have failed and that’s why the computer won’t post? I would assume damage on the analog section wouldn’t actually prevent the machine from posting would it?
I had one of these SB AWE 32 cards in my socket 7 K5/K6 based PC as well as in my first slot 1 based PC. I always wanted to added extra RAM to mine but I never did get around to doing so. Judging by the number of capacitors that appear to damaged I would bet there is a short in at least one of them that's causing the system not to POST. It's a shame that hot air killed this card. However, it is understandable that the first couple of attempts may not go well until a person has had more experience.
i really look forward to your repair videos. i like how you approach things and your breakdown of the methods you use. I always look forward to one of you videos. I hope to see lots more in the future. SIde note. My wife PC PSU was doing some odd things like shutting off randomly and then boot cycling until you turned it off for a while. so i grabbed a spare psu to see if the problem persisted and nope the extra PSU worked like a champ. I figured bad component or heat related. SO I popped open the old PSU and blew the dust out. put it back together and it worked better random restarts happen less often. but i noticed inside this PSU the Fan had a Plastic sheet covering almost 2/3 of the fan. blocking most of the air flow. well since the PSU is giving me issues i figured remove the plastic sheet to improve air flow. Its been 2 weeks now and the old PSU hasn't shut off once now. It was a Thermaltake TR2 600w PSU. I know they claim the plastic sheet is there to direct the flow of air but the one in this PSU was so stiff that if it was meant to flex and let air flow well that would have never happened.
To me it sounds like there's a short somewhere, maybe a fried capacitor. If the chip was fried it would probably still allow the system to boot, it would just not work properly or at all. I'd say capacitors are the most likely culprits. They don't like being overheated, and a shorted electrolytic cap across a power rail can certainly prevent the system from booting.
Before you try anything else, lightly sand the pads with 3000 grit sandpaper ( I just revived what I thought was a dead sound card doing that, a standard eraser didn't work )
I once had the EEPROM which stores the settings on my GUS PnP corrupted somehow which prevented the machine from booting. I fixed it by hotplugging the card to a live system and reprogramming the EEPROM with the standard tool from Gravis. So a non-booting ISA card does not necessarily mean a hardware fault. Though in this case the liberal use of hot air does seem like a more likely cause for the failure.
Are the electrolic capacitors around the melted connector in good shape? As far as I can see on the video, a handful has a nice bulge on their top. I can't tell if changing them would revive the card though.
Failing to post. I'd check for shorts on the ISA bus connector. A shorted address, data or irq line could do that, random dead components should not cause a post failure unless they are shorting things. As an isa pnp card the awe chip will likely be directly connected to the isa bus on a number of the pins so I might suspect solder bridges from the hot air. I would also consider taping over the power contacts on the isa bus connector to see if the card still causes failure to post when unpowered. Taping lots of contacts might also help narrow down which isa bus line is getting messed up by the card and you might be able to trace where it goes on the card and the failed component
Looks like this "sound" was from pcb itself, it was "popped" from overheating, and delaminated. You can see it in oscillator area, near resistor R68(7:54) - damaged vias
I checked the crystal. Seems fine to me. I need to look up a datasheet though because I could measure the correct frequency only on one pin. Not sure if there should be a frequency on any of the other three pins.
Flex check the isa port contact pins with a continuity’s with sound on your multi meter then ground and each pin while applying minor pressure if a short occurs in the meter solder is required for perfect operation nes cartridge repairs
I've had one of these from my childhood falling on me, lol. It just died. I replaced all of the large power-related capacitors only to find out that the main IC is dead :/.
i'd be recapping all those electrolytics. the heat did them no favors, and probably finished off the aging/drying of the fluid. after that, visual inspection of the soldering across the whole board, testing for shorts on the fingers, etc. once all thats checked/done, osc probing the signaling contacts to see whats going on when trying to boot, if it still doesn't allow post.
It's quite unlikely that any of the chips were damaged by hot air unless there is visible signs of scorching on them or the PCB close by. More likely dead capacitors or shorted ceramic capacitors. With the level of effort you put in to the AWE chip you might as well have removed it with chip-quick. Straightened all the pins under the microscope then put it back. This gives you the opportunity to inspect the PCB pads and glue down any loose ones using UV solder mask.
Wenn man bedenkt wie die Anschluss aussieht, sind mit Sicherheit alle Kondensatoren hin. Die sehen auch nicht mehr gut aus. C140 ist bereits ausgelaufen und fängt schon an mit Kontakten an U23 zu reagieren und wenn ich mir die Pinbelegung für die Kontakte von U23 ansehe, kommen da die Spannungen an. An der Stelle vielleicht auch auf Erdungsschluss prüfen.
I hadn't realized how insane the owner went with the hot air, I wonder if they got a soldering hot air gun or a woodworking one for paint removal... I suggest recapping the whole board, electrolytics hate heat and it was applied very globally and for ages, and even on the video the other caps don't look great IMHO, I hope it's just a cap that failed shorted or something like that!
I think it may have some issues in it's power circuitry. The bigger chips don't look damaged but maybe small resistors, power regulators and other capacitors might be affected. If you can, maybe try to measure things. I like the idea of @phirenz to use an oscilloscope too, but it depends on if you have one. There can be some affordable ones, I think I've seen some on other TH-camrs' channels.
I may be on a wild goose chase here so take this information with a lot of salt. However the C0 at least for Phoenix BIOSes relates to the ISA standard of Interrupt 19... Further digging lead me down a rabbit hole of information related to a software requested halt that is usually done by IDE / RAID controllers to initialize their drives to be booted by the computer. I am no ISA expert or even knowledgeable, but there are interrupt request pins on the ISA standard pinout... Maybe it is being pulled high or low (Pending on what it wants) due to some factor elsewhere on the card.
This is tested on an Asus p3b-f, it has an award bios. Iirc at this point in time during the post it's testing the memory or trying to init every components on the motherboard, so yeah it's probably throwing out the ISA bus in some way. Needs more analysis :)
Has this card been reverse engineered or are schematics available? Could possibly be one of the smaller 74 logic chips ... but without a schematic/block diagram it's just guessing. I know on a GUS PnP (since I was the first person to reverse engineer those) a faulty 74F32 would prevent the system from booting.
I actually found a small solder bridge between two signal pins. That was very odd - thin as a hair. The PC boots now, and I could get sound from it. The FM synthesizer didn't work, but that may have been an issue with drivers and Windows 98. I'll try the card in another system and also play around with the flash chip on this card.
I hope to learn a lot more about electronics. It's such an interesting subject. I'm a software engineer by education, but lost touch with the field. It's rare to find interesting projects to work on in a corporate environment and I'm tired of maintaining a 20+ years old codebase.
With the correct technique fixing badly bent legs on the edge of a QFP chip is very easy. First heat the leg with soldering iron and gently push the leg up from behind so it is above the pcb. Add flux and wick away all the solder so the legs are free to move easily. Bend the legs to the correct shape so the leg is perfectly straight and sits flat on the pcb then re-solder the legs with perfect alignment. I made some special tools so I can use the lifting technique even in the middle of a large QFP chip since this kind of damage is common on arcade pcbs... lots of clueless cowboys in that scene lol
Original owner mentions that the chip gets hot to the touch. There is a good chance the card had a faulty chip even before his hot air attempt. He probably just created more issues now. My guess, the card it is a waste of time but then again, where is the fun if you don't investigate further... hope you find a donor card to transfer over an AWE chip. That would have been my next move.
yeah I have a hot air station. I got it for 3d printing related stuff. I have a launch wii console. I promise you it needs recapped. I think I'm just going to use twist and normal solder gun method. I don't trust myself with hot air on circuits.
I see a few other caps that dont look healthy. I doubt the chips died from heat but capacitors like to pop under heat ive done that to my playstation the first time i worked on retro hardware.
Oh wei. Wenn man doch keine Ahnung von Hot Air hat, warum spielt man dann nicht erst mit einem echten Totholz umher bevor man die Karte grillt. Aber wieder tolle Arbeit und Frohe Ostern. BTW: Kontrolliere bitte den Bereich um den CD In Anschluss und PC Speaker. Das sieht alles sehr knusprig aus. Vorallem die Caps.
You did a whole lot better job straightening those tiny legs. However the card should have died long ago. A misplaced effort on a bad card. But it is soothing to watch you work on the card at least. 😁
first of all replace all electrolytic caps as those die when exposed to heat.. and to see how that person has baked the whole card the caps or likely all dead. once you done that you may try it / test it and id it wont boot don't tuen off the pc but grab a thermal camera and look if something gets hot as long non of the creative branded chips are defect then the card can be saved
I'm all for people learning new skills, but this card looks like it was treated by a careless butcher. Watch any video on using hot air and the first thing they teach is protecting plastic from melting. I have little respect for such work, it's not amateurish, it's sloppy and careless.
Capacitor c120 is also bulging.
I'd consider looking at all those capacitors between the caps between the audigy chip and the 4 pin connector
I really wonder how the owner made that happen, maybe left hot air unattended on the side of the board for a while or something, but yeah clearly all those caps are dead/dying!
People tend to stick hardware in the oven, without thinking for themselves what might be some issues that could occur.
I’d definitely replace those capacitors. Usually, they’re resistant to heat, anywhere from 85°C to 105°C. The molten connector tells me the temperature went higher than that. Evidently, those capacitors will dry out and/or pop.
Nothing is unfixable. :)
I'd say the majority of the caps are probably cooked, I see a fair few that look suspicious. Even some of the ceramic capacitors look suspicious, look at C16 to the right of the CT1745A chip, it's really dark compared to the ones next to it, like it's getting smoking hot from being shorted.
This card may work again, but it's going to need some serious low level diagnostics.
If the computer isn't posting, then the sound card must be doing something pretty catastrophic signalling on the ISA bus.
Maybe it's time to resort to an oscilloscope and/or logic analyser to see how it's effecting signalling on the ISA bus. That will help to narrow down where the fault might be.
I'm always amazed by the magic you do with bent pins.
Thank you!
use a thermal camera (if you have one) to see if any more caps or other components are shorting.
also i think it's pretty hard to damage ICs with heat. unless the card was stored in a humid environment and the person heated it very quickly or went above 300-350C. This is called popcorn effect and the bigger the IC the bigger the risk. However I would still recheck for a short somewhere since it stops the whole system from booting..
have you tested for shorts?
it could be any one of them surface mount caps and resitors have a hairline fracture from the heat. but i say its that audio chip the pins probably sent too much voltage to the wrong place. dame shame. would be nice to see you save the card somehow.
So many damaged capacitors...a fair few appear to have been killed by the heat.
Recap the whole board, just, all of it, because the more caps I look at, the worse it gets. Some of them look like they'd explode of you sneezed in their vicinity. After that I'd check for bridging with any of the other chips/connectors, because there's no telling what might have gotten messed by by that heat. Lastly, I'd test to see if that crystal is still okay.
I would check resistance to ground on all power lines into the card, as well as the data lines. I wouldn't be surprised if the onboard power supply of the card is shorting out, or some line is being pulled down. If you have one, maybe a thermal camera will be a quick way to see if a component is shorting, but you can get away without one, it will just take more time and effort. I think there is a lot more that can be done in terms of diagnosing before completely condemning this card. Obviously, commercially it makes no sense to spend this much time and effort on the card, but for greater justice and youtube content I think it will be more than worth it!
Exactly my thought. Although, i would not waste time for cheeking data lines at this stage, until the system is at least able to boot up. Any IC malfunction would not prevent system boot, this has to be a short on a card. So, i would start with checking power lines on the SB's ISA connector.
check for shorted ceramic caps; heat stresses can cause the ceramic to fracture and short the layers together. since they're usually used as bypass caps, the thermal gradient and stresses they endure are much higher due to the ground planes sinking all the heat away on one leg.
These cards had a serious problem with the PNP EEPROM corrupting. When that happens it stops detecting on the creative PNP utility TSR and Windows drivers can't find the card. Copying EEPROM data from a working card will save it if that is the problem. Maybe this card had that problem before it pop'd.
Everyone commenting about shorts -- if there were shorts between a power rail and ground, the power supply would not have stayed on. The overcurrent protection would definitely have kicked in and immediately shut the supply down. So I highly doubt there's a short between power and ground anywhere.
You're absolutely correct. And that overcurrent protection works like a charm. When I built my first -5V Adapter, one of the tantalum caps was bad and immediately caused the power supply to turn off.
There's definitely not a short between power input and ground. Shorts between two signal lines or between a signal and power/ground are much harder to detect.
So, turn of events: there was a short between two signal lines. The card posts now and also plays digitised music. Unfortunately, no FM/MIDI sound. Will try the card in an older system with pure DOS since the drivers have issues with win98 DOS
@@bitsundbolts Are you aware that on some/most AWE32 cards, FM is routed through the EMU8K synthesizer/effect processor and *requires* initialization using AWEUTIL /S in DOS (or installation of an AWE32 wavetable driver in Windows 95)? "No FM" might be just due to missing initialization.
In the electronics assembly industry, there is a term called "bake". It refers to some components needing to be "baked" before assembly. Baking removes moisture from inside of IC chips that could expand during reflow soldering and destroy the parts. I suspect the "pop" that was heard during hot air rework was from just that. During the baking procedure components are slowly heated to much lower temperatures than reflow soldering to gently remove the moisture. This is also why you sometimes receive parts from DigiKey etc. in sealed plastic sleeves with silica gel pouches inside. They are moisture sensitive (to some MSL level).
Excited to see my old card even if it's still not working. If you look at my first post on Vogons, the computer was only able to POST and detect the card one time, when it gave me the error. After that, the computer wouldn't POST unless I changed a jumper but it wouldn't detect the card.
I did change that jumper, but there was no change. I haven't given up on the card yet, I just didn't have enough time to look for other faults.
What were you hoping to achieve by blasting the thing with hot air? Genuinely curious. Hot air isn't going to unbend pins that are touching, for example.
I suspect he confused a hot air station with a hot air gun from the hardware store, plus didn't know how to shield other sensitive parts
Fight in the corner was fabulous! Check the card against shorts.
Check all caps, check for short on 3.3v 5v and data bus. Check for solder bridges.
Oh well, you can't save them all. Even with your perfect restoration methods, some parts are just beyond repair. I fear the same is also the case with the crispy Monster 3D II, but there's only one way to find out. 😄 It is unfortunate that the AWE32 probably met its final demise due to the ill-advised layman repair attempt, but the previous owner probably took the "it's broken anyway" approach and didn't want to sink any cost into a repair, which we can't really fault him for either.
It is a good candidate for the next level of diagnosis though. Knowing you, you are probably alrady itching to dive into EE level diagnosis. 😉 Who knows, maybe you'll get lucky and the remaining fault(s) are limited to generic parts. 🤞🏻
Happy Easter everyone!
Great video!
Thank you for sharing your adventure attempting to repair this sound card. :D
Hmmmm I need to see if I can find RAM for mine so I can upgrade it.
Try resoldering all the ferrite beads. They may crack under physical stress but show no visual sign at all.
Thanks for the suggestion! I'll try that before recapping the entire card.
I would use a thermal camera to figure out which components are getting too hot.
Also, it might be worth checking out the 74 TTL chips, such as the 74LS245.
It is better that it is one of those chips that is damaged rather than the proprietary ones.
And, as others have already suggested, check for shorts and the rest of the capacitors.
It could just be me having just done a full recap on a motherboard and PSU, but to me a lot of the rest of the caps on that card look like they're bulging as well. If the previous owner just air gunned the whole card, he may have blown or come close to blowing some of those other caps as well. I'd just try a full recap before giving up on it; caps are cheap. And maybe test for continuity/shorts around the components near where the hot air went.
Lots of el caps seems to be bulging, so I'd go for a full re-cap. If that won't work check if the oscillator generates anything. Then look for possible ceramic caps short.
Decades ago we were told not to get parts too hot or they will die, but if you watch Sorin blasting parts with very high temps, ie 380ºC, it doesn't seem to matter.
Me personally I would keep trying, unless I found an unreplaceable part burring hot when powered, ie, shorted.
You could check for shorts, diode check data lines, check if power is on the card, pull the IC totally and resolder, etc
Maybe pull parts off, plug it in, and see if it boots, ie, it might tell you what part is the problem.
Maby measure all caps (ceramic and electrolythic) for shorts and replace all the electrolythic ones anyways for good panasonic, nichicon or rubycon ones in case the card does work again. They are afterall almost 30 years old anyway.
Yeah I see other bulging caps and I'd check that crystal oscillator since it's close to the main chip where heat would of been applied. You're going to have to check the ceramics for shorts and see if there any off value or open resistors.
so sad! Probably a shot in the dark but I might see some bulging on the caps on the left hand side of the melted connector. I doubt they would prevent the whole system from posting but worth a shot.
Actually, I'm wondering how is it possible that the card is causing that issue. If you have a scope you could scope all the address/data lines as I feel that at that stage a shorted Address/Data line is the only way a SB could prevent a Motherboard from posting. From the video, it seems that the ISA bus is directly connected to the main IC so that's unfortunate.
Other things I can think of:
- Check that everything has voltage
- Check the clock
- Check for shorts
- Check that nothing is scorching hot :)
Pro repair tips: First check for a short between power and ground. There are several common logic chips on the board being used to isolate several buses (74LS244/74LS245), plus a few other smaller logic chips and a DRAM next to the large Creative chip. I would remove and test all them in a Retro Chip Tester Pro. If you don't have a RCT just remove and replace them with new parts or known good used parts. The melted connector is in the audio output section so that won't stop the computer booting but a bad logic chip connecting the card to the ISA bus certainly will cause an issue if the chip(s) are bad. Due to the custom chips the only thing you can do is check/replace any common off-the-shelf chips. If that doesn't work then it's scrap and a lesson to all noobs watching... waving a hot air station over a board like a magic wand won't do anything except break it. You can't fix it with a hammer. Do not mess with expensive and complicated electronics that you do not have the required skill levels and equipment to handle.... PRACTICE ON JUNK BOARDS UNTIL YOU GET THAT SKILL LEVEL!!!
What i think is that everybody who tried or learn to repair electronics is that sometime and somewhere have broken something that was broken but could be fixed :). I hope at least he learned!
Im always impressed by your skills working on such a tiny scale! Bummer that it didn’t work out. Maybe some of the logic ICs that are straight on the ISA bus have failed and that’s why the computer won’t post? I would assume damage on the analog section wouldn’t actually prevent the machine from posting would it?
Ah, the original owner got a hammer and saw a bunch of nails.
Check all capacitors and also use a thermal camera to see (after replacing any defective capacitors) if any shorts are visible on thermals
I had one of these SB AWE 32 cards in my socket 7 K5/K6 based PC as well as in my first slot 1 based PC. I always wanted to added extra RAM to mine but I never did get around to doing so.
Judging by the number of capacitors that appear to damaged I would bet there is a short in at least one of them that's causing the system not to POST. It's a shame that hot air killed this card. However, it is understandable that the first couple of attempts may not go well until a person has had more experience.
i really look forward to your repair videos. i like how you approach things and your breakdown of the methods you use. I always look forward to one of you videos. I hope to see lots more in the future.
SIde note. My wife PC PSU was doing some odd things like shutting off randomly and then boot cycling until you turned it off for a while. so i grabbed a spare psu to see if the problem persisted and nope the extra PSU worked like a champ. I figured bad component or heat related. SO I popped open the old PSU and blew the dust out. put it back together and it worked better random restarts happen less often. but i noticed inside this PSU the Fan had a Plastic sheet covering almost 2/3 of the fan. blocking most of the air flow. well since the PSU is giving me issues i figured remove the plastic sheet to improve air flow. Its been 2 weeks now and the old PSU hasn't shut off once now. It was a Thermaltake TR2 600w PSU. I know they claim the plastic sheet is there to direct the flow of air but the one in this PSU was so stiff that if it was meant to flex and let air flow well that would have never happened.
Voltage injection can be handy for finding unseen damage
To me it sounds like there's a short somewhere, maybe a fried capacitor. If the chip was fried it would probably still allow the system to boot, it would just not work properly or at all. I'd say capacitors are the most likely culprits. They don't like being overheated, and a shorted electrolytic cap across a power rail can certainly prevent the system from booting.
If there was a short, the power supply would have tripped on overcurrent and it wouldn't have stayed powered as long as it did.
Before you try anything else, lightly sand the pads with 3000 grit sandpaper ( I just revived what I thought was a dead sound card doing that, a standard eraser didn't work )
That boot lock I have seen it before. Check the firmware in the eeprom is the little 8 pin chip next to the awe chip
Abnormal currents, shorts, heat?
Maybe yeet every single chip off the card except the ISA controller and see whether it still makes the bus unusable?
I was gonna suggest removing the main chip to see if that still results in bus shenannigans too, so yeah.
I once had the EEPROM which stores the settings on my GUS PnP corrupted somehow which prevented the machine from booting. I fixed it by hotplugging the card to a live system and reprogramming the EEPROM with the standard tool from Gravis. So a non-booting ISA card does not necessarily mean a hardware fault. Though in this case the liberal use of hot air does seem like a more likely cause for the failure.
Super repair attempt. :)
Clock crystal was near area of excessive heat - check that it is working.
I could measure the correct frequency on one of the pins. Not sure if there should be something on the other three pins
Please keep trying!
I will try my best to resurrect this card. All suggestions from the comments are helpful and we might find the flaw on this card.
Are the electrolic capacitors around the melted connector in good shape? As far as I can see on the video, a handful has a nice bulge on their top. I can't tell if changing them would revive the card though.
Failing to post. I'd check for shorts on the ISA bus connector. A shorted address, data or irq line could do that, random dead components should not cause a post failure unless they are shorting things. As an isa pnp card the awe chip will likely be directly connected to the isa bus on a number of the pins so I might suspect solder bridges from the hot air. I would also consider taping over the power contacts on the isa bus connector to see if the card still causes failure to post when unpowered. Taping lots of contacts might also help narrow down which isa bus line is getting messed up by the card and you might be able to trace where it goes on the card and the failed component
Looks like this "sound" was from pcb itself, it was "popped" from overheating, and delaminated. You can see it in oscillator area, near resistor R68(7:54) - damaged vias
I think that at 1:31 you can even see the PCB bulging next to the oscillator 😳
Maybe check the rails for shorts and see if that big crystal is oscillating
I checked the crystal. Seems fine to me. I need to look up a datasheet though because I could measure the correct frequency only on one pin. Not sure if there should be a frequency on any of the other three pins.
If it's of any help, I do have one of these that I'm pretty sure is functional, or at least was the last time I tested it some number of years ago.
Flex check the isa port contact pins with a continuity’s with sound on your multi meter then ground and each pin while applying minor pressure if a short occurs in the meter solder is required for perfect operation nes cartridge repairs
I've had one of these from my childhood falling on me, lol. It just died. I replaced all of the large power-related capacitors only to find out that the main IC is dead :/.
Did I see "SPDIF" on that silk screen? Bizarre, I didn't think computer sound cards were including that in the ISA era.
i'd be recapping all those electrolytics. the heat did them no favors, and probably finished off the aging/drying of the fluid.
after that, visual inspection of the soldering across the whole board, testing for shorts on the fingers, etc.
once all thats checked/done, osc probing the signaling contacts to see whats going on when trying to boot, if it still doesn't allow post.
I'm a step closer. There was indeed a short between two data lines. There'll definitely be a follow-up
It's quite unlikely that any of the chips were damaged by hot air unless there is visible signs of scorching on them or the PCB close by. More likely dead capacitors or shorted ceramic capacitors. With the level of effort you put in to the AWE chip you might as well have removed it with chip-quick. Straightened all the pins under the microscope then put it back. This gives you the opportunity to inspect the PCB pads and glue down any loose ones using UV solder mask.
Fried like crispy bacon…RIP sound card 😂
Did you check if the quarz oscillates? I can imagine the heat cracking it...
I got the card half way to work. There was a tiny solder bridge I didn't see. I'll get back to this card after I fix all those P2B motherboards...
I would've recommended to replace all the capacitors. Possibly many more has been damaged.
Thermal camera?
Measure live voltages?
Wish I could give advise . Unfortunately I don’t have a clue at all
Wenn man bedenkt wie die Anschluss aussieht, sind mit Sicherheit alle Kondensatoren hin. Die sehen auch nicht mehr gut aus. C140 ist bereits ausgelaufen und fängt schon an mit Kontakten an U23 zu reagieren und wenn ich mir die Pinbelegung für die Kontakte von U23 ansehe, kommen da die Spannungen an. An der Stelle vielleicht auch auf Erdungsschluss prüfen.
I hadn't realized how insane the owner went with the hot air, I wonder if they got a soldering hot air gun or a woodworking one for paint removal... I suggest recapping the whole board, electrolytics hate heat and it was applied very globally and for ages, and even on the video the other caps don't look great IMHO, I hope it's just a cap that failed shorted or something like that!
Rework 10/10
look for unusual resistance in the isa connector, and all ics..
Maybe the crystal (xtal was also cooked) because its very close to the main sound chip.
How to replace all the capacitors heat really messes them up especially from a hot air workstation
Check if the crystal oscillator is running.
I think it may have some issues in it's power circuitry. The bigger chips don't look damaged but maybe small resistors, power regulators and other capacitors might be affected. If you can, maybe try to measure things.
I like the idea of @phirenz to use an oscilloscope too, but it depends on if you have one. There can be some affordable ones, I think I've seen some on other TH-camrs' channels.
Maybe the hot air destroyed more of the electrolytic capacitors
not sure if its from the camera angle but c120 up by the melted connector looks swollen
I may be on a wild goose chase here so take this information with a lot of salt.
However the C0 at least for Phoenix BIOSes relates to the ISA standard of Interrupt 19...
Further digging lead me down a rabbit hole of information related to a software requested halt that is usually done by IDE / RAID controllers to initialize their drives to be booted by the computer.
I am no ISA expert or even knowledgeable, but there are interrupt request pins on the ISA standard pinout... Maybe it is being pulled high or low (Pending on what it wants) due to some factor elsewhere on the card.
This is tested on an Asus p3b-f, it has an award bios.
Iirc at this point in time during the post it's testing the memory or trying to init every components on the motherboard, so yeah it's probably throwing out the ISA bus in some way. Needs more analysis :)
@@DxDeksor For Award BIOS it was even worse... C0 is "Turn off chipset cache" which I found to be really really strange indeed.
Has this card been reverse engineered or are schematics available? Could possibly be one of the smaller 74 logic chips ... but without a schematic/block diagram it's just guessing. I know on a GUS PnP (since I was the first person to reverse engineer those) a faulty 74F32 would prevent the system from booting.
I actually found a small solder bridge between two signal pins. That was very odd - thin as a hair. The PC boots now, and I could get sound from it. The FM synthesizer didn't work, but that may have been an issue with drivers and Windows 98. I'll try the card in another system and also play around with the flash chip on this card.
Wow, that is some serious skill to reverse engineer a GUS! Congratulations. Are those GUS clones the result of your work?
@@bitsundbolts Yeah, definitely check the drivers. I've had the same issue with an AWE32 in the past. Glad you managed to fix the card.
@@bitsundbolts I'm the person behind the ARGUS, yup ... luckily there have been many more clones since :)
I hope to learn a lot more about electronics. It's such an interesting subject. I'm a software engineer by education, but lost touch with the field. It's rare to find interesting projects to work on in a corporate environment and I'm tired of maintaining a 20+ years old codebase.
With the correct technique fixing badly bent legs on the edge of a QFP chip is very easy. First heat the leg with soldering iron and gently push the leg up from behind so it is above the pcb. Add flux and wick away all the solder so the legs are free to move easily. Bend the legs to the correct shape so the leg is perfectly straight and sits flat on the pcb then re-solder the legs with perfect alignment. I made some special tools so I can use the lifting technique even in the middle of a large QFP chip since this kind of damage is common on arcade pcbs... lots of clueless cowboys in that scene lol
Original owner mentions that the chip gets hot to the touch. There is a good chance the card had a faulty chip even before his hot air attempt. He probably just created more issues now. My guess, the card it is a waste of time but then again, where is the fun if you don't investigate further... hope you find a donor card to transfer over an AWE chip. That would have been my next move.
fine surgery you did there
There is a short visible between 4 and 5 leg at 6:42
yeah I have a hot air station. I got it for 3d printing related stuff. I have a launch wii console. I promise you it needs recapped. I think I'm just going to use twist and normal solder gun method. I don't trust myself with hot air on circuits.
I see a few other caps that dont look healthy. I doubt the chips died from heat but capacitors like to pop under heat ive done that to my playstation the first time i worked on retro hardware.
What flux and where to buy one you are using? Thanks
Links to the flux are in the video description.
Oh wei. Wenn man doch keine Ahnung von Hot Air hat, warum spielt man dann nicht erst mit einem echten Totholz umher bevor man die Karte grillt. Aber wieder tolle Arbeit und Frohe Ostern. BTW: Kontrolliere bitte den Bereich um den CD In Anschluss und PC Speaker. Das sieht alles sehr knusprig aus. Vorallem die Caps.
And For AWE32 Check All Capitor
Check oscillator and voltages on all chips
You did a whole lot better job straightening those tiny legs. However the card should have died long ago. A misplaced effort on a bad card. But it is soothing to watch you work on the card at least. 😁
Check for dead shorts?
3:13 looks like there is another bridged pins
That might just be a reflection of the flux I'm using. Will double-check though.
9:49 those caps don't look good. Maybe these are the cause for the system not posting? I doubt the soundchip is the culprate.
Dont know if the card could work in an 8 bit slot instead of 16 bit..
Thermal camera, or recap the entire board
It looks like dram chip has a little dot...
Pull chips and get new PCB?
get oscilloscope and take a look at isa signalling lines. I bet you'll find the culprit of that c0 there.
recap the whole thing
I had some of these... all in garbage by now... :(
first of all replace all electrolytic caps as those die when exposed to heat.. and to see how that person has baked the whole card the caps or likely all dead. once you done that you may try it / test it and id it wont boot don't tuen off the pc but grab a thermal camera and look if something gets hot as long non of the creative branded chips are defect then the card can be saved
Aye Double-you Eee
Recap the card
I'm all for people learning new skills, but this card looks like it was treated by a careless butcher.
Watch any video on using hot air and the first thing they teach is protecting plastic from melting.
I have little respect for such work, it's not amateurish, it's sloppy and careless.
Pls pls HELP me to repair my 2x PC-XT motherboards. PLS
I Really Want To Give My ga-P41-es3g And Asus CUSI-FX To You For Repair
get oscilloscope and take a look at isa signalling lines. I bet you'll find the culprit of that c0 there.