gerry wrote on 2024-09-16, 14:25:

PD2JK wrote on 2024-09-16, 09:19:

Feel free to comment of course.

does this mean it might not have been the caps? its still a nice future proofing job either way.

Yes that is possible. I've tested a few old ones and they were still good. But I was on the job anyway, so I figured to replace them anyway.

Daniël Oosterhuis wrote on 2024-09-16, 08:05:

This weekend, I reballed the PCI Creative Blaster 3DFX Voodoo Banshee a colleague gave me to try and fix. The card had artifacti […]
Show full quote

This weekend, I reballed the PCI Creative Blaster 3DFX Voodoo Banshee a colleague gave me to try and fix.
The card had artifacting issues on cold boots, that would slowly fade as the chip got hot.

Once properly hot, Glide games worked just fine (starting them while the card was still artifacting would result in garbage output and crashing), but to me that indicated poor solder connections under the chip.
With how poorly these chips were cooled with piddly little heatsinks (the PCB had significant discoloration on the area of the PCB under the chip), I guess the PCB has been bent from the large temperature swings, causing cracked solder balls that expand and reconnect due to thermal expansion when the IC gets hot.

I did attempt a simple reflow before this, but that resulted in fully corrupted graphics in the BIOS, so that didn't do the trick.
I also tried a cheap Chinese replacement chip, but once soldered on, it didn't even get hot and had no output, so it's either a fake or already dead.

So, I pulled that chip back off, and practiced reballing it as the chips I've been reballing before were all ceramic BGA chips (PowerPC processors) rather than plastic BGA chips, which are susceptible to popcorning if not handled correctly.
After that went well, I finally went ahead and reballed the original IC, and soldered it back onto the card.

The attachment bansheereball1.jpg is no longer available

The attachment bansheereball2.jpg is no longer available

The attachment bansheereball3.jpg is no longer available

The attachment bansheereball4.jpg is no longer available

An initial test, after giving the PCI edge connector a good clean, results in a clean picure in the BIOS, free of artifacts.
I've still got to test Glide games, but this is a very good sign already.

The attachment bansheereball5.jpg is no longer available

Woah, nicely done. I've got to do the same at some point with one of my banshee cards that wasn't helped by a reflow.
I also have the same card you have but it went from no VGA output to 100% working with a reflow so I was quite lucky there: Re: What retro activity did you get up to today?

What does the potentially fake banshee chip look like? You could potentially check its ground and vcc pins against a good chip but otherwise sadly I don't know the pinout. I find checking power and ground pins is a really good way to root out fake chips since those will usually not match up. I had to do that for some YM2612 chips and complaining about fake chips that were really SRAM on ebay got me banned by a big group of the chinese sellers 🙁

Could you say where was it from so that I can avoid that place? I'm still trying to find a Voodoo 1 TMU chip to fix my orchid righteous 3d 😒

Thermalwrong wrote on 2024-09-16, 15:46:

Woah, nicely done. I've got to do the same at some point with one of my banshee cards that wasn't helped by a reflow. I also hav […]
Show full quote

Woah, nicely done. I've got to do the same at some point with one of my banshee cards that wasn't helped by a reflow.
I also have the same card you have but it went from no VGA output to 100% working with a reflow so I was quite lucky there: Re: What retro activity did you get up to today?

What does the potentially fake banshee chip look like? You could potentially check its ground and vcc pins against a good chip but otherwise sadly I don't know the pinout. I find checking power and ground pins is a really good way to root out fake chips since those will usually not match up. I had to do that for some YM2612 chips and complaining about fake chips that were really SRAM on ebay got me banned by a big group of the chinese sellers 🙁

Could you say where was it from so that I can avoid that place? I'm still trying to find a Voodoo 1 TMU chip to fix my orchid righteous 3d 😒

It was this ad on AliExpress, it's multiple chips so you need to select 500-0013-03 as "color" to see it as an option.
Cheap enough to gamble on, and it was clearly a gamble lost.

The chip looks identical to the real thing, on the top and the bottom, so I don't think it's a fake chip, just a real one pulled off a Banshee sent off to China for ewaste.
Just unfortunate this one is dead, and even with how cheap it is, I don't recommend trying one yourself of course:

IDK if it's retro, but it's a 15 year old piece of junk...

CR600 laptop, MSI, previous symptoms, turn on, freezes in POST screen while extremely hot... had another socket P CPU, so thought I'd swap it over, T4300 to T2370, annnnd, it's different... doesn't show post screen, seems to do more CD activity, but screen blank... hmmm. IDK if it's wanting a BIOS because it doesn't have microcode for the T2370, so is tryna load something from CD drive... Either that or nothing gonna happen at all. Anyway, not getting super hot while it's sitting there blank, so dunno if that's better or not. Maybe after I play "sheep and goats" or "towers of hanoi" type game with CPUs among other laptops I'll have a T7100 to try instead, but for now it didn't seem to advance the frontiers of computing or anything.

Daniël Oosterhuis wrote on 2024-09-16, 20:55:

Just unfortunate this one is dead

What made you think the chip was dead? Have you tried installing the chip on the board? Or using a tester to examine the contacts? Do you know that the chip Banshee has unused contacts?

If you ever seen how those chip are salvaged you would have less doubts 😀 Open fire pit with PCBs roasting and some child banging them on the ground so the chips fall off.

rasz_pl wrote on 2024-09-18, 14:18:

If you ever seen how those chip are salvaged you would have less doubts 😀 Open fire pit with PCBs roasting and some child banging them on the ground so the chips fall off.

https://www.youtube.com/watch?v=WF6Vz5p4FUI

Futurama did it! Its really really sad.

dormcat wrote on 2024-09-15, 19:58:

GigAHerZ wrote on 2024-09-15, 18:20:

Have you ever thought using GameBoy cartridge cases as CPU containers? Check this out!

Are those pieces of black paper/tape meant for preventing pins scratching cases?

Yes, this is foam. Pretty dense variant and i recommend something lighter / more "airy", but works well.

Today I researched and wrote an article about Altered Beast. Some things I learned:

• It was Makoto Uchida's first game. He would go on to create two of my favorite beat 'em ups, Golden Axe and Die Hard Arcade.
• The game was supposed to use pressure-sensitive controls, but Sega couldn't get the company that owned the patent on it to create a small enough batch of them to go in the cabinets.
• The flimsy story of the game was loosely based on a gory story from Greek mythology.
• The System 16 had two variants (a and b). I had assumed is was just one.
• More player controls were removed at the end of the game's development and were never properly replaced, which is why the gameplay is so monotonous.

Postman5 wrote on 2024-09-18, 06:22:

Daniël Oosterhuis wrote on 2024-09-16, 20:55:

Just unfortunate this one is dead

What made you think the chip was dead? Have you tried installing the chip on the board? Or using a tester to examine the contacts? Do you know that the chip Banshee has unused contacts?

I did install it on the board, and it didn't even get slightly warm, whereas these chips usually get real hot, real fast.
It might look like I didn't use it, but that's because I reballed that chip to practice before reballing the original IC.

I used the exact same methods to solder the old one back on after reballing, and it did get hot right away, though I initially didn't get video output.
But that was just a case of cleaning the PCI pins and reinserting the card, after that all was fine.

rasz_pl wrote on 2024-09-18, 14:18:

If you ever seen how those chip are salvaged you would have less doubts 😀 Open fire pit with PCBs roasting and some child banging them on the ground so the chips fall off.

Oh, I'm fully aware that chips are often salvaged like that in China and other Asian countries that we dump all our ewaste to, so chips from sources like this are a total gamble and I treat it as such.
It's also why I prefer it when Chinese IC sellers just give you the chip still attached to a cut off piece of PCB, at least no one's given it a baptism by fire then 😁

Started to disassemble my new Cordata SX20 laptop.
Found that there were previous attempts to fix it and two missed screw inside the laptop 😀 CMOS battery is absent, so it would be tricky exercise to find the spare.
External battery leaked too and didn't looks original inside. FDD was stuck, but I hope that after the cleaning and new lube apply it will works fine.

This should qualify as mostly retro:

I made a Test Box!
The computer I was using is my M2N-E/AX2-4600+/GT9600 (codename AMDemon) using Autodesk Inventor 2012.
I designed it to house the standard testing set to make checking boards easier (and portable!) The test print went ok... It came out alright, but I never went that high before, so there's some things I'm going to do differently for V2.
First thing I'm doing is putting a front panel in there. I initially wanted one, but I didn't know if printing it was going to be possible. Second thing I did was make the cables come out the side. I kinda knew I was going to have cable issues. I also added a fan hole(80mm), mostly for aesthetics (and a Lian-Li grill was sitting on my desk, unused) The final thing I added was a female molex with the two molex and floppy (made a SATA one, too) on the other side, to make it easy to power on. A PC speaker hole is also a MUST, so slapped one of those on there. I had to put some bars across the face, too, so it won't come off the printer surface.

Looking good so far, just need to buy some more printer fuel to check out v2. Approximated print time on the main housing is 58 hours and 220meters. Could do a faster one, but with that much material, might as well go for it.

Daniël Oosterhuis wrote on 2024-09-19, 07:54:

I did install it on the board, and it didn't even get slightly warm, whereas these chips usually get real hot, real fast. It mig […]
Show full quote

Postman5 wrote on 2024-09-18, 06:22:

Daniël Oosterhuis wrote on 2024-09-16, 20:55:

Just unfortunate this one is dead

What made you think the chip was dead? Have you tried installing the chip on the board? Or using a tester to examine the contacts? Do you know that the chip Banshee has unused contacts?

I did install it on the board, and it didn't even get slightly warm, whereas these chips usually get real hot, real fast.
It might look like I didn't use it, but that's because I reballed that chip to practice before reballing the original IC.

I used the exact same methods to solder the old one back on after reballing, and it did get hot right away, though I initially didn't get video output.
But that was just a case of cleaning the PCI pins and reinserting the card, after that all was fine.

rasz_pl wrote on 2024-09-18, 14:18:

If you ever seen how those chip are salvaged you would have less doubts 😀 Open fire pit with PCBs roasting and some child banging them on the ground so the chips fall off.

Oh, I'm fully aware that chips are often salvaged like that in China and other Asian countries that we dump all our ewaste to, so chips from sources like this are a total gamble and I treat it as such.
It's also why I prefer it when Chinese IC sellers just give you the chip still attached to a cut off piece of PCB, at least no one's given it a baptism by fire then 😁

Hmm, I hadn't considered that reason when I see cut-off PCB pieces with the chip still on there, like lots of the AMD 5x86 133 PQFP chips are still attached to their PCB segments. I considered it barbaric but what you saw with the chip not getting even hot once installed, maybe it was kept under sustained heat long enough that most of the gold bond wires melted and so the BGA pads connect to nothing. Like you say it looks legit and it's not something worth faking particularly.

Gotta do the reball on my Voodoo Banshee AGP at some point, since I got my CT6760 Voodoo Banshee PCI working I've had little incentive, but I got the BGA balls and the stencil / tray in preparation for it.

Susanin79 wrote on 2024-09-19, 10:31:

Started to disassemble my new Cordata SX20 laptop.
Found that there were previous attempts to fix it and two missed screw inside the laptop 😀 CMOS battery is absent, so it would be tricky exercise to find the spare.
External battery leaked too and didn't looks original inside. FDD was stuck, but I hope that after the cleaning and new lube apply it will works fine.

Wow that one's quite a rare survivor in that condition, the screen looks great for its age. Was the floppy drive belt-driven or like a desktop type direct-drive FDD?

Today and the previous day and a couple months ago I've been working on my Toshiba LI-3300, which is a Toshiba XM-3301BC SCSI 1x speed CD-ROM drive in an external scsi housing. It is glorious:

It uses a caddy which took me 4 months to find, but before that I replaced every single capacitor in it, they are 30 years old and it seems this drive was run til it stopped running. They had all leaked:

When I found the caddy it was super disappointing because it would not read a CD, it would spin up briefly and spit out the disc. I spent a while testing and figured out you can use these drives without the caddy! Just tape the cover-pusher-arm out of the way, put a coin on the CD and some blu-tack around that. Rest it on the spindle then press the microswitch that gets hit when the caddy is inserted and it'll move the carriage up and start reading:

Putting a disc in, sometimes it would spin up but make strange noises when trying to spin the disc. Eventually I found that heating the PCB briefly with a hairdryer before powering up the drive would allow it to work, that lead me to think there's another bad cap or a hairline broken trace, but the fault was more pernicious than that. I went back over it and found that I'd left an original radial cap which smoothed the 5v rail and on removal that stank of fish oil:

I'd cleaned as well as I could but the board still didn't look great. Cleaning more by hand and tinning the traces made no difference. Thankfully I have access to an ultrasonic cleaner and put the board in that, maybe close to an hour total in 5 minute bursts cleaning new things each time, sometimes it really did stink of fish oil even though all the caps were good, there was cap juice under the ICs where I couldn't clean. After all that, which had the unfortunate side effect of destroying *all* the soldermask labels... it works again! After drying it with the hairdryer, blowing out water from under the ICs with a blower then letting it dry out for a while, I put it back together and the drive spun up and read quietly on the first try 😁

Not through the woods just yet since the external SCSI enclosure power supply is not starting up, but this caddy drive works again! So great, and no faffing about with the laser this time - see here for my last experience, which in hindsight I doubt was even the laser tracking anyway, the motor speed regulation is equally important and I ignored it back then - Re: What retro activity did you get up to today?

Gotta make some website or something for some of these drives, the Toshiba LI-3300 doesn't seem to exist according to google but it's right in front of me.

Once upon a time, I had an external Mitsumi similar to that. Any older optical can prolly read the Domesday book.

Caddies are a bugger to find, I had 3 in the house and it took me 2 years 🤣

Bought new brand of a vintage keyborad.

Cleaned and upgraded the CPU to A6-3500 for this FM1 build, also added an 32GiB SDD for the OS.

TheChexWarrior wrote on Yesterday, 10:00:

Bought new brand of a vintage keyborad.

It's a very nice!

Thermalwrong wrote on Yesterday, 00:45:

Hmm, I hadn't considered that reason when I see cut-off PCB pieces with the chip still on there, like lots of the AMD 5x86 133 P […]
Show full quote

Daniël Oosterhuis wrote on 2024-09-19, 07:54:

I did install it on the board, and it didn't even get slightly warm, whereas these chips usually get real hot, real fast. It mig […]
Show full quote

Postman5 wrote on 2024-09-18, 06:22:

What made you think the chip was dead? Have you tried installing the chip on the board? Or using a tester to examine the contacts? Do you know that the chip Banshee has unused contacts?

I did install it on the board, and it didn't even get slightly warm, whereas these chips usually get real hot, real fast.
It might look like I didn't use it, but that's because I reballed that chip to practice before reballing the original IC.

I used the exact same methods to solder the old one back on after reballing, and it did get hot right away, though I initially didn't get video output.
But that was just a case of cleaning the PCI pins and reinserting the card, after that all was fine.

rasz_pl wrote on 2024-09-18, 14:18:

If you ever seen how those chip are salvaged you would have less doubts 😀 Open fire pit with PCBs roasting and some child banging them on the ground so the chips fall off.

Oh, I'm fully aware that chips are often salvaged like that in China and other Asian countries that we dump all our ewaste to, so chips from sources like this are a total gamble and I treat it as such.
It's also why I prefer it when Chinese IC sellers just give you the chip still attached to a cut off piece of PCB, at least no one's given it a baptism by fire then 😁

Hmm, I hadn't considered that reason when I see cut-off PCB pieces with the chip still on there, like lots of the AMD 5x86 133 PQFP chips are still attached to their PCB segments. I considered it barbaric but what you saw with the chip not getting even hot once installed, maybe it was kept under sustained heat long enough that most of the gold bond wires melted and so the BGA pads connect to nothing. Like you say it looks legit and it's not something worth faking particularly.

Gotta do the reball on my Voodoo Banshee AGP at some point, since I got my CT6760 Voodoo Banshee PCI working I've had little incentive, but I got the BGA balls and the stencil / tray in preparation for it.

Susanin79 wrote on 2024-09-19, 10:31:

Started to disassemble my new Cordata SX20 laptop.
Found that there were previous attempts to fix it and two missed screw inside the laptop 😀 CMOS battery is absent, so it would be tricky exercise to find the spare.
External battery leaked too and didn't looks original inside. FDD was stuck, but I hope that after the cleaning and new lube apply it will works fine.

Wow that one's quite a rare survivor in that condition, the screen looks great for its age. Was the floppy drive belt-driven or like a desktop type direct-drive FDD?

Today and the previous day and a couple months ago I've been working on my Toshiba LI-3300, which is a Toshiba XM-3301BC SCSI 1x speed CD-ROM drive in an external scsi housing. It is glorious:

The attachment ebay-picture.jpg is no longer available

It uses a caddy which took me 4 months to find, but before that I replaced every single capacitor in it, they are 30 years old and it seems this drive was run til it stopped running. They had all leaked:

The attachment XM-3301-capacitors.JPG is no longer available

When I found the caddy it was super disappointing because it would not read a CD, it would spin up briefly and spit out the disc. I spent a while testing and figured out you can use these drives without the caddy! Just tape the cover-pusher-arm out of the way, put a coin on the CD and some blu-tack around that. Rest it on the spindle then press the microswitch that gets hit when the caddy is inserted and it'll move the carriage up and start reading:

The attachment IMG_44641 (Custom).JPG is no longer available

Putting a disc in, sometimes it would spin up but make strange noises when trying to spin the disc. Eventually I found that heating the PCB briefly with a hairdryer before powering up the drive would allow it to work, that lead me to think there's another bad cap or a hairline broken trace, but the fault was more pernicious than that. I went back over it and found that I'd left an original radial cap which smoothed the 5v rail and on removal that stank of fish oil:

The attachment IMG_4508 (Custom).JPG is no longer available

I'd cleaned as well as I could but the board still didn't look great. Cleaning more by hand and tinning the traces made no difference. Thankfully I have access to an ultrasonic cleaner and put the board in that, maybe close to an hour total in 5 minute bursts cleaning new things each time, sometimes it really did stink of fish oil even though all the caps were good, there was cap juice under the ICs where I couldn't clean. After all that, which had the unfortunate side effect of destroying *all* the soldermask labels... it works again! After drying it with the hairdryer, blowing out water from under the ICs with a blower then letting it dry out for a while, I put it back together and the drive spun up and read quietly on the first try 😁

The attachment IMG_4467 (Custom).JPG is no longer available

Not through the woods just yet since the external SCSI enclosure power supply is not starting up, but this caddy drive works again! So great, and no faffing about with the laser this time - see here for my last experience, which in hindsight I doubt was even the laser tracking anyway, the motor speed regulation is equally important and I ignored it back then - Re: What retro activity did you get up to today?

Gotta make some website or something for some of these drives, the Toshiba LI-3300 doesn't seem to exist according to google but it's right in front of me.

Ugh, just spent hours trying to get the switch mode power supply working. The output capacitors on the secondary side had mostly leaked and the power supply was resetting constantly and not giving a voltage. I'm not an EE and high voltage electronics are not something I'm comfortable with, my current method to discharge the main SMPS capacitor causes a welding spark and the complexity here is beyond me.
But then I was looking at the specs of the drive vs the power supply and the PSU is specced for 40w output (5v 3a & 12v 2a) which is wasted on this Toshiba drive which uses a measly 700ma on both the 5v & 12v rails.
So I've removed the switch mode power supply from the external caddy and hooked up a 12V DC jack where the mains plug went, from that there's a 5v step-down regulator so that's 5v & 12v sorted from a 12v input which I've got lots of. Ran a connector with resistor off of that for the external enclosure's power LED.
Now the drive is working and tomorrow I'll make up a 3d print to hold the DC jack in place and it's done 😀

Caddy drive is wonderful, my last experience with them was in the school library where they spent out on PCs with CD-rom drives and encyclopedia software that was outdated in probably less than a year.

BitWrangler wrote on Yesterday, 03:19:

Once upon a time, I had an external Mitsumi similar to that. Any older optical can prolly read the Domesday book.

Caddies are a bugger to find, I had 3 in the house and it took me 2 years 🤣

It's a shame the information on these older CD drives is so scarce now, for me there's really something special about these 1st gen 1x speed drives. This drive and my LMSI drive are quite special but I hope to one day find one of the Mitsumi LU-005 drives, that's what we had back in the 90s and I really miss it.

The caddies are horrid to find, I'm tempted to make up a 3d print to make these drives usable without the original caddy but there probably aren't many of these drives still in a running state now.
The really great thing about caddies is that they so strongly resemble regular CD cases so people don't tend to think they're anything special. If mixed in with regular cases and especially big box games then they'd be impossible to find.

Actually yeah, mine were in with a bunch of CD Jewel cases so I might have missed them at least once. I know I've got one SCSI internal and one Mitsu or Panasonic interface internal that needs one. I am hoping to dig out a 1992 correct drive for the Pondering the recreation of an early "Multimedia" PC... if it exists. which might end up being a caddy load.

It was a bugger to dig out info on the older CD drives in the mid 90s even, a lot of tech info was on dial in BBS and faxback services of the manufacturers and never made it to the web. It was easier to find out about some of them from the information in some of the older Linux HOWTOs, CDROM Howto is maybe the one I'm thinking of.