DeathRabbit679 wrote on 2022-11-16, 06:21:

I was fooling around with my new MT32. Then I went to insert a CD and discovered my CD-ROM on my Pentium 1 box is waving the white flag and won't eject any more, so my retro activity became attempting to yank it out, only to discover I'll have to remove the FDD and HDD to be able to remove a screw someone put in an unreachable location. Ran out of time to mess with it. So, ultimately, my retro activity was banging my head on my desk repeatedly.

Ouch, yeah actually using retro stuff can end up that way pretty often. I always end up finding new faults on my laptops when I start using them properly. Hopefully it should be relatively easy to replace the drive eject belt on your CD drive or you've got a spare?

I've had a PC Chips M919 sitting on my 'this needs testing' pile and finally got around to it. The two NPN transistors that make up the motherboards voltage regulator circuitry had holes burnt out in them - this board came to me in a junk lot and I'm guessing the CPU got put in backwards. Testing them somehow one of them still worked? But the other one had a gain of 10 instead of 280 so it's no good anymore.

The heatsinks were gone too so I had to improvise. From the good M919 board I have, I can tell the M919 uses either Toshiba 2SC3420 or NEC 2SD1691 which are BJT NPN transistors. I've used NEC D1691-Y transistors from China and while testing I found one of them was literally backwards? Instead of being ECB with the writing on the front, it was BCE with the writing on the front. I'm wondering if they're fakes or re-marked parts but they were cheap enough and I only needed 2 so it's fine.

It worked and didn't catch fire, but on starting the motherboard just gave code E1 or E2 which for an AMI win bios is not a valid code. Better than "--" codes at least.
Initially I was running the 5x86 133 without a heatsink and it was pretty hot so I jammed this big old Zalman fan on there, which is working great, now it's running stone cold. It's so heavy that it's not going anywhere.

Something notable about these PC Chips boards is that they use 1.2mm pcb thickness instead of 1.6mm and this means they're kinda flexy - there was a visible bend in the top corner of the motherboard with the fake cache chips and the UM8881F chip. The cache chips had some clearly loose pins so I squished the UM8881F chip onto the board with my hand and powered it then it started up.
To fix that I resoldered the UM8881F's pins and now it's working pretty well, even working with the cache stick I've got for the M919:

Hercules really wasn't that generous with the thermal paste!

I had a similar experience inside of a VAIO laptop. Have a loop at this "generous" thermal pad. This is on a Pentium 3 @ 600 Mhz.

I found this after it died suddenly. Coincidence?

Kahenraz wrote on 2022-11-18, 11:58:

I had a similar experience inside of a VAIO laptop. Have a loop at this "generous" thermal pad. This is on a Pentium 3 @ 600 Mhz.

I found this after it died suddenly. Coincidence?

20221112_192235.jpg

With what you are paying for sure you'd pay those 50 cents more for good thermal compound.
I hate it when they cheap out on essential things.

aaaaand the gremlins have raised their ugly heads again. Couldn't write to c drive blue screen every one in a while. GAH... -_- win98 y u no stable?

Spent the evening troubleshooting this 386 build.

First, the 1.44MB drive stopped reading disks. Traced the issue to the I/O controller card. After replacing the card, managed to get disks working again.

Immediately following that, my CF card suddenly stopped working. After further troubleshooting it appears to be a fault with the CF-IDE adapter. Which really surprises me given that it's almost brand new and there are very few electronics on these adapters.

One I/O controller card and CF-to-IDE adapter later and the system is back to working. We'll see how long it holds up. 😒

Bought electronic components including 33MHz and 25MHz plastic package oscillators to modify a compaq motherboard to run 12.5 or 16.5MHz 386 or 486 for experiment with slow down method for some "touchy" games.

Cheers,

Socketed a PLCC32 bios chip.
The same pin kept deosldering and crashing the machine 🙁
Now fixed, I needed a smaller tip I couldn't find.

Shponglefan wrote on 2022-11-18, 23:38:

Spent the evening troubleshooting this 386 build. […]
Show full quote

Spent the evening troubleshooting this 386 build.

First, the 1.44MB drive stopped reading disks. Traced the issue to the I/O controller card. After replacing the card, managed to get disks working again.

Immediately following that, my CF card suddenly stopped working. After further troubleshooting it appears to be a fault with the CF-IDE adapter. Which really surprises me given that it's almost brand new and there are very few electronics on these adapters.

One I/O controller card and CF-to-IDE adapter later and the system is back to working. We'll see how long it holds up. 😒

Nice machine 😉
CF cards have few shottky diodes and they failed a few times on my cards. Resoldered and it's back in business.
At least that has always been the cause for my CFs.

Shponglefan wrote on 2022-11-18, 23:38:

Immediately following that, my CF card suddenly stopped working. After further troubleshooting it appears to be a fault with the CF-IDE adapter. Which really surprises me given that it's almost brand new and there are very few electronics on these adapters.

I had the same problem with a set of three CF to IDE adapters I received last week. One was not dead but worked with some (but not all) cards. And those it did work with showed corruption at POST, hung, or failed to let the system boot.

I inspected it under my microscope and couldn't find any problem with the traces or solder joints. I agree, there isn't anything that should go wrong on these, which is really weird. The seller wanted it back, so I returned it without investigating further.

Thermalwrong wrote on 2022-11-16, 22:19:

DeathRabbit679 wrote on 2022-11-16, 06:21:

I was fooling around with my new MT32. Then I went to insert a CD and discovered my CD-ROM on my Pentium 1 box is waving the white flag and won't eject any more, so my retro activity became attempting to yank it out, only to discover I'll have to remove the FDD and HDD to be able to remove a screw someone put in an unreachable location. Ran out of time to mess with it. So, ultimately, my retro activity was banging my head on my desk repeatedly.

Ouch, yeah actually using retro stuff can end up that way pretty often. I always end up finding new faults on my laptops when I start using them properly. Hopefully it should be relatively easy to replace the drive eject belt on your CD drive or you've got a spare?

I swapped in a spare and then worked on the drive some. I think it was just stuck because reasons, after taking it apart and manually exercising mechanism, it seemed good to go. Put it back in the parts pile because with how expensive even basic 90s PC parts are getting, I don't dare toss anything if I still want to be messing around with these computers 5/10 years from now. I can see it now, <Giant Narrator Voice> "The year is 2032, crime is rampant in the streets, disease spreads unchecked, and IDE cd-rom drives with beige faceplates cost 900 dollars on Ebay. It's a world gone mad!" </GNV>

Dumped one bios.
Tried to bench a board but would randomly crash accessing storage in W98SE.

GX media is really not a good system overall. Didn't get much support.

Just finished PC (P133, 64MB RAM, Super TX3 MB) to test the Diamond Edge 3D 2120 card that I found some days ago.
Luckily the card is OK, it was a bit tricky to set the drivers, but now it is time for some Sega games 😀

I tried to install Windows 98 on a B75 motherboard. I thought it would be easy, but boy, was I wrong!

I tried all the tricks I knew about, trying to limit memory, etc - no luck. The installer crashes in the same spot, when detecting hardware. Could be ACPI related, but there is nothing in the BIOS that could fix it.

There goes my idea to build an all in one Win98 / WinXP / Win10 build. Guess I will have to look for different hardware!

gmaverick2k wrote on 2022-11-18, 22:00:

aaaaand the gremlins have raised their ugly heads again. Couldn't write to c drive blue screen every one in a while. GAH... -_- win98 y u no stable?

Solved the issue. Disconnected second Seagate HDD containing clonezilla backup image and connected main Seagate HDD to sata 2. Plain sailing from now, no issues so relieved

Gave a quick check to the 50 floppies I picked up not too long ago. Only 2 was completely unreadable and one damaged.
Also checked a couple of floppy and CD drives. The floppy drives turned out to be all good, but 2 CD drives were faulty.
Gave a quick check to the ES1869 found in a banged up PC this week, Device Manager don't seem to pick it up.

Nexxen wrote on 2022-11-19, 00:05:

Nice machine 😉
CF cards have few shottky diodes and they failed a few times on my cards. Resoldered and it's back in business.
At least that has always been the cause for my CFs.

Thank you. 😀

Yeah, I'm hoping it's a simple case of a failed component that won't be too tough to diagnose / replace. Though with the few number of components, I can't imagine it will take too long to figure out. 😁

Kahenraz wrote on 2022-11-19, 01:50:

I had the same problem with a set of three CF to IDE adapters I received last week. One was not dead but worked with some (but not all) cards. And those it did work with showed corruption at POST, hung, or failed to let the system boot.

I inspected it under my microscope and couldn't find any problem with the traces or solder joints. I agree, there isn't anything that should go wrong on these, which is really weird. The seller wanted it back, so I returned it without investigating further.

I had similar symptoms. My system would POST, but then hang when attempting to boot. I also noticed the LEDs on it seemed dimmer than normal, so I'm assuming some sort of voltage wonkiness is going on, though it's not with the PSU output since I tested those and its fine.

With few components should hopefully not take too much work to figure out what failed.

Today I added some RGB to my 'Ultimate' Windows 98 build.

Next up I may try adding an Audigy 2 ZS installed alongside the MX300, at which point I think this one will be done.

Shponglefan wrote on 2022-11-19, 22:40:

Thank you. :) […]
Show full quote

Nexxen wrote on 2022-11-19, 00:05:

Nice machine 😉
CF cards have few shottky diodes and they failed a few times on my cards. Resoldered and it's back in business.
At least that has always been the cause for my CFs.

Thank you. 😀

Yeah, I'm hoping it's a simple case of a failed component that won't be too tough to diagnose / replace. Though with the few number of components, I can't imagine it will take too long to figure out. 😁

Kahenraz wrote on 2022-11-19, 01:50:

I had the same problem with a set of three CF to IDE adapters I received last week. One was not dead but worked with some (but not all) cards. And those it did work with showed corruption at POST, hung, or failed to let the system boot.

I inspected it under my microscope and couldn't find any problem with the traces or solder joints. I agree, there isn't anything that should go wrong on these, which is really weird. The seller wanted it back, so I returned it without investigating further.

I had similar symptoms. My system would POST, but then hang when attempting to boot. I also noticed the LEDs on it seemed dimmer than normal, so I'm assuming some sort of voltage wonkiness is going on, though it's not with the PSU output since I tested those and its fine.

With few components should hopefully not take too much work to figure out what failed.

Dimmer LED light happens when I forget to plug the FD connector.
VRM busted? You'll tell in the near future. 😀

Shponglefan wrote on 2022-11-19, 22:46:

Today I added some RGB to my 'Ultimate' Windows 98 build.

Next up I may try adding an Audigy 2 ZS installed alongside the MX300, at which point I think this one will be done.

This is cute 😉
Bravo!