I installed a Chinese card cooler on my GeForce 3 Ti 200. I had to hacksaw and file portions of the base to make it clear the RAM sinks, and when I tried to install it I discovered that the short fins were preventing me from fully inserting it into the AGP slot so I had to bend some of them in.

PD2JK wrote on 2024-05-16, 13:56:

Blew up a Caviar 80MB hard drive, WDAC280.

I can't make anything out of it. Diode or capacitor?

DSC_9039.JPG

Could have been a small inductor. Peel of some of the black case material and there should be some wound wires visible ("wire wound chip inductor"). You'll likely need some good magnification.
Looking at the datasheet of the spindle driver there's no other component beween Pin 12 and Cp, but according the packaging / footprint I'd strongly suspect it to be an inductor .
Try measuring the resistance above it. If it's open/high resistive remove and bridge it. That should make it work again if there's no other problem.

That circuit is a charge pump for generating a higher-than-Vcc voltage to drive the high side switches of the motor drivers.

Anyone know of any water blocks for the socket 370 era chips? I've got a crazy idea to water cool this dual p3 system I'm building..

Veeb0rg wrote on 2024-05-17, 08:38:

Anyone know of any water blocks for the socket 370 era chips? I've got a crazy idea to water cool this dual p3 system I'm building..

Socket 370 is compatible with Socket A, so look for those that fit Socket A instead.

ChrisK wrote on 2024-05-17, 06:06:

Could have been a small inductor. Peel of some of the black case material and there should be some wound wires visible ("wire wo […]
Show full quote

PD2JK wrote on 2024-05-16, 13:56:

Blew up a Caviar 80MB hard drive, WDAC280.

I can't make anything out of it. Diode or capacitor?

DSC_9039.JPG

Could have been a small inductor. Peel of some of the black case material and there should be some wound wires visible ("wire wound chip inductor"). You'll likely need some good magnification.
Looking at the datasheet of the spindle driver there's no other component beween Pin 12 and Cp, but according the packaging / footprint I'd strongly suspect it to be an inductor .
Try measuring the resistance above it. If it's open/high resistive remove and bridge it. That should make it work again if there's no other problem.

That circuit is a charge pump for generating a higher-than-Vcc voltage to drive the high side switches of the motor drivers.

I will look into it if I have the time. Thanks!

Repaired a couple of traces on a mobo and soldered a tiny qfn.

Nexxen wrote on 2024-05-17, 23:49:

Repaired a couple of traces on a mobo and soldered a tiny qfn.

Wow, nicely done that looks really really tiny! What does it do?

Repo Man11 wrote on 2024-05-16, 22:48:

I installed a Chinese card cooler on my GeForce 3 Ti 200. I had to hacksaw and file portions of the base to make it clear the RAM sinks, and when I tried to install it I discovered that the short fins were preventing me from fully inserting it into the AGP slot so I had to bend some of them in.

I had to do the same on my GF3 Ti200 which I think was a dell OEM one that either didn't have a heatsink because it came from a scrapper or had a passive heatsink. I used dremel for mine - somehow I was cutting off the metal bracket on the back and cut a notch into the PCB and somehow missed any crucial components so it still works great. Got to make sure to remove the heatsink before dremel next time 😀

---------
The last couple of days I got my Toshiba T2450CT in extremely ruined condition. About the only thing working without repair so far is the hard drive, which had the T2450CT recovery floppy images on it! Those are going on Archive.org soon 😀

I think I might make a thread on it in system specs to track its progress because it had so much wrong that I should log the faults for other people to come across in future.
The LTM09C012 LCD is unobtainium and the front and rear polariser appear to be ruined. I took off the front polariser and tried fixing the LCD controller PCB then used a polarising filter to view the LCD but still nothing visible on the LCD 🙁

The mainboard was broken to start with since all the Elna brand caps (with the exception of one Elna LongLife cap???) had let go their juices and the ultrasonic helped get them clean:

I got this post code reader for a parallel port to upgrade my 8-led post reader since it lets me go forward and back while seeing previous entries on my led post reader requires capturing a video. The post code reader was really annoying though and required an external 5v PSU, using a stupid USB-A to USB-A cable! I have had a bad experience with those before so I grafted a power-bank PCB and battery from a vape onto it so now it's portable 😁

Also I put some foam pads on the underside and over the power bank's PCB so it's not blinding and the reader can't short on things so easily now.
It came in handy too, the T2450CT stopped posting the day after I fixed the DC section's capacitors and the post code reader for the T2450CT was stuck on 01/02 which according to the T2450CT's maintenance manual is the programmable interrupt timer initialisation. That turned out to be a bad connection on the joint PCB since the BIOS is on a separate PCB from the CPU, connected by a flex PCB with 2x board-to-board connectors, faults are inevitable with that many connectors. Re-seating the connector allowed the T2450CT to work again

Transferred a Socket 7 Baby AT board into an ATX chassis. It really does fit beautifully with four screws aligning, for the top two I've used nylon standoffs + hot glue. With a 3D printed backplate, it really does look the part. Even the PCI slots align - I never tried this before and didn't realise it was this easy.

Obviously not a solution for every baby/AT board, but mine has both ATX and AT power headers, so is blessed with pins to be bridged by the momentary touch button on the case instead of a switch.

Thermalwrong wrote on 2024-05-18, 01:44:

Nexxen wrote on 2024-05-17, 23:49:

Repaired a couple of traces on a mobo and soldered a tiny qfn.

Wow, nicely done that looks really really tiny! What does it do?

Traces were connected to IDE pins, IIRC "something" enable (I'll post later with exact definition). DA0 and DA2 (pins 35 and 36)
It was slow and boring, but there was enough space as it was between PCI2 and PCI3.

Next time I'll use longer wires to anchor at distant points to avoid touchups removing the whole thing. But here I wanted to keep the silkscreen and avoid the angle, guess I should have used the angle to anchor.
Also, I have to buy yellow soldermask.

Thermalwrong wrote on 2024-05-18, 01:44:

and required an external 5v PSU, using a stupid USB-A to USB-A cable! I have had a bad experience with those before so I grafted […]
Show full quote

and required an external 5v PSU, using a stupid USB-A to USB-A cable! I have had a bad experience with those before so I grafted a power-bank PCB and battery from a vape onto it so now it's portable 😁
post code reader (1).JPG
Also I put some foam pads on the underside and over the power bank's PCB so it's not blinding and the reader can't short on things so easily now.
It came in handy too, the T2450CT stopped posting the day after I fixed the DC section's capacitors and the post code reader for the T2450CT was stuck on 01/02 which according to the T2450CT's maintenance manual is the programmable interrupt timer initialisation. That turned out to be a bad connection on the joint PCB since the BIOS is on a separate PCB from the CPU, connected by a flex PCB with 2x board-to-board connectors, faults are inevitable with that many connectors. Re-seating the connector allowed the T2450CT to work again

Nice job 😀 kudos!

Dan386DX wrote on 2024-05-18, 04:46:

Transferred a Socket 7 Baby AT board into an ATX chassis. It really does fit beautifully with four screws aligning, for the top two I've used nylon standoffs + hot glue. With a 3D printed backplate, it really does look the part. Even the PCI slots align - I never tried this before and didn't realise it was this easy.

Obviously not a solution for every baby/AT board, but mine has both ATX and AT power headers, so is blessed with pins to be bridged by the momentary touch button on the case instead of a switch.

I fit all my baby AT boards into ATX test stands. It is easy to drill and tap a new M3 hole for a standoff. The hardest part is getting the location correct.

PD2JK wrote on 2024-05-17, 14:26:

ChrisK wrote on 2024-05-17, 06:06:

Could have been a small inductor. Peel of some of the black case material and there should be some wound wires visible ("wire wo […]
Show full quote

PD2JK wrote on 2024-05-16, 13:56:

Blew up a Caviar 80MB hard drive, WDAC280.

I can't make anything out of it. Diode or capacitor?

DSC_9039.JPG

Could have been a small inductor. Peel of some of the black case material and there should be some wound wires visible ("wire wound chip inductor"). You'll likely need some good magnification.
Looking at the datasheet of the spindle driver there's no other component beween Pin 12 and Cp, but according the packaging / footprint I'd strongly suspect it to be an inductor .
Try measuring the resistance above it. If it's open/high resistive remove and bridge it. That should make it work again if there's no other problem.

That circuit is a charge pump for generating a higher-than-Vcc voltage to drive the high side switches of the motor drivers.

I will look into it if I have the time. Thanks!

You're welcome. Let us know how it goes.

There are a couple of systems I have that I mostly have because I enjoy having them. One of these is my "Not so Super" Socket 7 system based on a TXP4 - 500 MHz K6-3+, 256 megs of PC133, a PCI FX 5500. It scores just over 2,800 point in 3D 2000, and since I had it out yesterday I decided to install Half-Life on it and see what it's like to play it on such a low end system. At 800x600 it is playable, but there are some places where the frame rate dips pretty low. I think back to July of 2001 when I built my P55T2P4 system with a K6-2+ and a Pine TNT2 32 meg PCI card and played Jane's WW-2 Fighters; having never had anything better, I found it very enjoyable.

On a Toshiba Sat Pro laptop, a core2 duo, I am replacing Windows 2000 Pro with Windows 2000 Advanced Server.

It dual boots with Slackware-64 15, the current release of Slackware, and Windows 2000. 2k was peak Windows, and back in the day I found 2k server was the best desktop version: disable a few things and change some settings, and you have 2k pro with proper terminal services available.

I've got an iso of 2000AS from Microsoft's page on archive.org, and rather than burn a CD I've got a Windows 98 virtual machine sharing the iso by SMB 1 (or whatever 98 uses). Then the laptop will be PXE booted into The Universal TCP/IP Network Bootdisk, and run the win2kAS setup.exe across the LAN. Got all this working, except actually running the setup.exe.

Just copying the drivers and games to the Linux partition, then will do the reinstall.

Win2kAS's install will wipe out LILO, but that should be fixable by PXE booting into the Slackware installer, with a rootfs option pointing at the Slackware partition on the HDD. Double-check /etc/lilo.conf , then rerun lilo.

\i386\winnt.exe worked under DOS, the setup.exe in the root of the CD can't be run from DOS.

But there's no partition available for temporary space! The Windows installer only allows me to exit.

And I don't want Windows to end up on D:, so will make 2 partitions, use the latter for temp, install Win2k onto the first, then tidy up afterwards with GParted.

This was not a bright idea 😀

Today I recapped my TRS-80 Model 100...and killed it. It was working mostly, the screen would flicker in and out and those two usual capacitors had started leaking but nothing bad yet. I put a new barrel battery in it and bought a cap kit, spent all evening replacing them all. Now it's completely dead. I double checked all the caps, they are all correct and properly oriented. So now I'm depressed, kind of makes me want to get rid of every bit of my retro stuff.

Migrated my Tualatin/440BX rig to my 11y old SandyBridge case. I need to make some space and could not throw nice CM Silencio case away. It still has all relevant legacy bays. Everything fit perfectly except MB power arrangement since it uses modern bottom PSU layout and P3B-F MB ATX power connector is designed for top PSU cases.
Initial testing with summer temps outside my 1.4Ghz Tualatin is hovering in 45 - 55C zone. I might need to add additional exhaust fan. While I turned my Seasonic Gold PSU upside down with intension to draw out air I don't think its effective since P3 system is not putting enough load to make PSU fan work.
Air flow i important since I will be probably add passively cooled Voodoo 2 later.

Edit: After switching to 120mm exhaust only - clearly it's more effective even with lowered ~1100 RPM. Kind of obvious, because it's right next to hottest source - CPU. My GF4 4200 ti is relatively cool card.

vutt wrote on 2024-05-19, 09:38:

Migrated my Tualatin/440BX rig to my 11y old SandyBridge case. I need to make some space and could not throw nice CM Silencio ca […]
Show full quote

Migrated my Tualatin/440BX rig to my 11y old SandyBridge case. I need to make some space and could not throw nice CM Silencio case away. It still has all relevant legacy bays. Everything fit perfectly except MB power arrangement since it uses modern bottom PSU layout and P3B-F MB ATX power connector is designed for top PSU cases.
Initial testing with summer temps outside my 1.4Ghz Tualatin is hovering in 45 - 55C zone. I might need to add additional exhaust fan. While I turned my Seasonic Gold PSU upside down with intension to draw out air I don't think its effective since P3 system is not putting enough load to make PSU fan work.
Air flow i important since I will be probably add passively cooled Voodoo 2 later.

Edit: After switching to 120mm exhaust only - clearly it's more effective even with lowered ~1100 RPM. Kind of obvious, because it's right next to hottest source - CPU. My GF4 4200 ti is relatively cool card.

Another problem is that the warm air rises inside the case, while the fan in the bottom mounted PSU tries to counteract this (if it was actually running, off course) so wouldn't help much either way. The exhaust fan seems like a really good idea 😀

I tidy up some HW back, one was Processors, they range from the 486 up to the Pentium 775, between the 486 there are two 486 DX2-80 is a DX4-120, the latter is the first one I have, while I already had a couple of DX2-80, obviously being low voltage CPUs (about 3V), they can't be installed in the MB Socket 2 (or earlier), except that they have the voltage reduction circuit on board, the alternative would be an intermediate socket with the circuit that lowers the voltage .

Among the other CPUs some S.7, all from Intel, a P100 a P133 and a P200MMX, of these I already have others that I keep in my collection, but the P133 is the one I usually use to try the MB S.7, after if everything goes well, I switch to the MMX or K6, there speed depends on the type of regulator, if it is linear at most I install a 200 MHz, if it is switching I go around 400 MHz with K6 CXT CPU.

Speaking of S.7 motherboards, I think of repairing the Soyo SY-5BT with cache chip without a PIN, I saw a video on Necroware where it repaired a chip, milling and soldering two PINs, in my case it might be simpler, since the missing pin is only one .

Cleaned up a free PS3 I got from buying a LG M2432 monitor+TV combo. Phat, CECHJ04 model.

gerry wrote on 2024-05-14, 20:02:

they are the best plugs it seems https://www.youtube.com/watch?v=UEfP1OKKz_Q :) […]
Show full quote

BigDave wrote on 2024-05-14, 17:15:

In the UK at least, it doesn't help that most plugs sold now, are usually only found with 13amp fuses fitted. I sold consumer electronics in the 80s/90s, initially retail, and back then, most equipment didn't come with a fitted plug, so as part of the sales process we'd also fit the correct fused plug for the customer, if nothing else to stop them blowing themselves up when they got home! I'm always amazed when I've bought vintage equipment from the pre-fitted plug days, at how badly some people wired their plugs (A recent Amstrad computer purchase as an example), so always the first thing I check before getting it anywhere near a socket. Then again, I guess most people were never taught how, and certainly a great many then, and probably now, didn't realise that you could even get fuses other than 13amp.
Ignorance is bliss, except where electricity is concerned, it needs respect.

they are the best plugs it seems
https://www.youtube.com/watch?v=UEfP1OKKz_Q
😀

Sorry to continue this, but yes, they are indeed, and as that great video demonstrates, UK plugs are very safe, even more in recent decades with the introduction of insulated pins.

Unfortunately, they were, still are at the mercy of the person fitting them, the cables should all be cut and stripped as seen so as to fit the plug neatly. More often than not back then, people used to just connect the wires as is, usually far too long so either squashed inside or protruding outside of the plug, and then terminals not tightened properly using the knife out of the kitchen draw. Forget to fit the clamp to hold the cable in place should they pull the cable, not the plug to remove from socket. Finally, due to the mentioned squashed cable inside, wouldn't be able to screw the plug top cover down properly, exposing the insides and putting the owner at risk every time they removed it, or, they would try to screw it together, which caused the moulding to cut into the wires. Scary, but that isn't the worst thing I've seen relating to a plug. I bought an old amber screen DOS computer in the 90s, and the 'tight' or lazy owner, had wired the 3 bits of equipment into just 1 plug! they had twisted all the wires together, so barely inserted into the terminals, and lots of exposed wires. I guess the only reason he was still alive, is because it hadn't been moved. Absolutely shocking! OK, I'm done with plugs, promise.... unless anyone remembers those 80's mini 3 pin plug systems made by Masterplug?