With a large area pan-head like that, it's quite likely to be friction/stiction combined with mild corrosion that's keeping them from turning, so I would recommend using a penetrant, Kroil, WD-40 or something.

I managed to get one of the rusted screws out with a few drops of synthetic oil and a high quality steel hex bit. The cheap allen key set I used initially was just stripping it. I was lucky to get the last screw out at all.

Working on a Sony PSX DESR-5700 that has no audio and cannot read CDs that I throw in there. It uses 100VAC, which in my neck of the woods, it's 120VAC that we use and I'll need a converter/stepdown to make the PSX work on a 120VAC line or find a way to replace the power supply that has 120VAC out of the box.

appiah4 wrote on 2022-03-24, 12:01:

I am fearing that they may indeed be rivets but the hollows look too hexagonal to be just that, so I am still hoping someone will come up and say "Oh they are this and this kind of screw, I took one apart before.." But knowing Intel they may well be unremovable rivets.

Which doesn't explain why the PII 400 SECC1 I have in my hand has no HSF on it.. Maybe it's an OEM part that had an aftermarket HSF that was later removed? No idea.

I found this post in a 5-year-old thread started by some shady-sounding 😁 "appiah4" fellow...

Re: Are all Slot 1 P3 coolers the same?
Gatewayuser200:
"Some heat sinks on SECC1 slot 1 processors are just not intended to be removed.

I've seen some OEM SECC1 slot 1 processors with screw/rivets holding the heatsink in place. They use a torx bit, but because the splines are so small and fragile it's very easy to round it out. Once it rounds out the only way to remove them is will a drill or drill press which is a pain."

Sounds like pretty much the same situation you're facing here.

Working on an 'ultimate' Windows 98 build based on an Athlon XP 2500k.

Modern parts include a Fractal Design Define R5 case, Seasonic 650W ATX power supply, and Samsung EVO 870 SSD.

Probably going to try to include at least two sound cards, a Diamond MX300 (pictured) and an Audigy 2 ZS (to be added).

Not sure on the GPU I'm going to use. Currently it has a GeForce 6800. Might keep it or might swap it for something a bit earlier gen.

The seasonic has beefy 12v?

Shreddoc wrote on 2022-03-24, 20:45:

I found this post in a 5-year-old thread started by some shady-sounding :D "appiah4" fellow... […]
Show full quote

appiah4 wrote on 2022-03-24, 12:01:

I am fearing that they may indeed be rivets but the hollows look too hexagonal to be just that, so I am still hoping someone will come up and say "Oh they are this and this kind of screw, I took one apart before.." But knowing Intel they may well be unremovable rivets.

Which doesn't explain why the PII 400 SECC1 I have in my hand has no HSF on it.. Maybe it's an OEM part that had an aftermarket HSF that was later removed? No idea.

I found this post in a 5-year-old thread started by some shady-sounding 😁 "appiah4" fellow...

Re: Are all Slot 1 P3 coolers the same?
Gatewayuser200:
"Some heat sinks on SECC1 slot 1 processors are just not intended to be removed.

I've seen some OEM SECC1 slot 1 processors with screw/rivets holding the heatsink in place. They use a torx bit, but because the splines are so small and fragile it's very easy to round it out. Once it rounds out the only way to remove them is will a drill or drill press which is a pain."

Sounds like pretty much the same situation you're facing here.

Well, I tried to remove them with a torx driver and I managed to unscrew two before I managed to 'round out' the other two. I could take a drill to it but that will defeat the purpose of mounting it on another CPU so I just put the removed screwsback on. That cooler will just stay with that CPU for life now.

appiah4 wrote on 2022-03-24, 21:25:

Shreddoc wrote on 2022-03-24, 20:45:

I found this post in a 5-year-old thread started by some shady-sounding :D "appiah4" fellow... […]
Show full quote

appiah4 wrote on 2022-03-24, 12:01:

I am fearing that they may indeed be rivets but the hollows look too hexagonal to be just that, so I am still hoping someone will come up and say "Oh they are this and this kind of screw, I took one apart before.." But knowing Intel they may well be unremovable rivets.

Which doesn't explain why the PII 400 SECC1 I have in my hand has no HSF on it.. Maybe it's an OEM part that had an aftermarket HSF that was later removed? No idea.

I found this post in a 5-year-old thread started by some shady-sounding 😁 "appiah4" fellow...

Re: Are all Slot 1 P3 coolers the same?
Gatewayuser200:
"Some heat sinks on SECC1 slot 1 processors are just not intended to be removed.

I've seen some OEM SECC1 slot 1 processors with screw/rivets holding the heatsink in place. They use a torx bit, but because the splines are so small and fragile it's very easy to round it out. Once it rounds out the only way to remove them is will a drill or drill press which is a pain."

Sounds like pretty much the same situation you're facing here.

Well, I tried to remove them with a torx driver and I managed to unscrew two before I managed to 'round out' the other two. I could take a drill to it but that will defeat the purpose of mounting it on another CPU so I just put the removed screwsback on. That cooler will just stay with that CPU for life now.

Sounds like the way to go. Those things appear very easily rounded, just by looking at them. There are various hacks to remove stuck fixtures. Jamming or sticking something to the head, so it can be unscrewed. But honestly sometimes these things are more trouble than it's worth.

Built P3/Windows 98 machine recently. Sort of a higher-end setup with Tualatin 1266, Voodoo 3 3000 and Vortex 2:

Tried previously with Geode NX1750, which was too fast at 1200-1400Mhz ... even for UT99 (game speedups).

And one ongoing project with P2 450:

chris2021 wrote on 2022-03-24, 21:09:

The seasonic has beefy 12v?

That's a good question. Should I be worried if it doesn't? 😳

You should mainly be concerned with 5V rail for socket A.

stef80 wrote on 2022-03-24, 22:56:

You should mainly be concerned with 5V rail for socket A.

Looking at the old PSU I took the system from, it's got +5V @ 22A. Whereas this new PSU is rated +5V @ 20A.

Looks like 1000W Seasonic PSUs are rated at +5V @ 25A. Perhaps I should have gone for one of those instead?

At any rate, everything seems to be working so far. Got Windows 98 installed without issue, other than having to set the SSD partitions to LBA.

I may keep the system underclocked anyway, since I'm not planning on running anything too new on it (relative to the time the OS came out).

appiah4 wrote on 2022-03-24, 21:25:

Shreddoc wrote on 2022-03-24, 20:45:

I found this post in a 5-year-old thread started by some shady-sounding :D "appiah4" fellow... […]
Show full quote

appiah4 wrote on 2022-03-24, 12:01:

I am fearing that they may indeed be rivets but the hollows look too hexagonal to be just that, so I am still hoping someone will come up and say "Oh they are this and this kind of screw, I took one apart before.." But knowing Intel they may well be unremovable rivets.

Which doesn't explain why the PII 400 SECC1 I have in my hand has no HSF on it.. Maybe it's an OEM part that had an aftermarket HSF that was later removed? No idea.

I found this post in a 5-year-old thread started by some shady-sounding 😁 "appiah4" fellow...

Re: Are all Slot 1 P3 coolers the same?
Gatewayuser200:
"Some heat sinks on SECC1 slot 1 processors are just not intended to be removed.

I've seen some OEM SECC1 slot 1 processors with screw/rivets holding the heatsink in place. They use a torx bit, but because the splines are so small and fragile it's very easy to round it out. Once it rounds out the only way to remove them is will a drill or drill press which is a pain."

Sounds like pretty much the same situation you're facing here.

Well, I tried to remove them with a torx driver and I managed to unscrew two before I managed to 'round out' the other two. I could take a drill to it but that will defeat the purpose of mounting it on another CPU so I just put the removed screwsback on. That cooler will just stay with that CPU for life now.

Did you apply oil to lubricate before trying to back them out? I warned you. 😀

Shponglefan wrote on 2022-03-25, 00:12:

Looking at the old PSU I took the system from, it's got +5V @ 22A. Whereas this new PSU is rated +5V @ 20A. […]
Show full quote

stef80 wrote on 2022-03-24, 22:56:

You should mainly be concerned with 5V rail for socket A.

Looking at the old PSU I took the system from, it's got +5V @ 22A. Whereas this new PSU is rated +5V @ 20A.

Looks like 1000W Seasonic PSUs are rated at +5V @ 25A. Perhaps I should have gone for one of those instead?

At any rate, everything seems to be working so far. Got Windows 98 installed without issue, other than having to set the SSD partitions to LBA.

I may keep the system underclocked anyway, since I'm not planning on running anything too new on it (relative to the time the OS came out).

I recently went digging through my old PSU collection looking for a PSU to power my 3200+.. got a ton of P4 PSUs with stupidly good +12v rails, a lot of AT PSUs and P3 capable ones .. and buried under them all was a ...

Shaw 680watt PSU with 35amps on the +5v rail, now this PSU looks shmick ..clean inside and out with little wear and tear on it .. but fuck its a SHAW

Now it tests out perfectly, all rails are at their correct voltages under load and there are no fishy burning electronic smells coming from it . even the damn fan is silent, which leaves me in a position I dont want to be in .. do I trust a SHAW PSU with my rare XP 3200+ unlocked Barton.

Its the only PSU I have that will possibly run the 3200+ without issue, its on a Nforce2 Ultra 400 board, so the damn thing is a power hungry beast and I have tried a few other PSU but they can barely supply enough voltage and with just the bare essentials and the CPU underclocked it gets close to maxing out the +5v rail so its no Bueno for them PSUs.

So I turn this over to fellow Vogons users .. do I dare trust this system to a SHAW PSU ?

I midstyped. I meant to inquire about the 5vdc amperage. I'm a dork.

Kahenraz wrote on 2022-03-25, 00:58:

appiah4 wrote on 2022-03-24, 21:25:

Shreddoc wrote on 2022-03-24, 20:45:

I found this post in a 5-year-old thread started by some shady-sounding :D "appiah4" fellow... […]
Show full quote

I found this post in a 5-year-old thread started by some shady-sounding 😁 "appiah4" fellow...

Re: Are all Slot 1 P3 coolers the same?
Gatewayuser200:
"Some heat sinks on SECC1 slot 1 processors are just not intended to be removed.

I've seen some OEM SECC1 slot 1 processors with screw/rivets holding the heatsink in place. They use a torx bit, but because the splines are so small and fragile it's very easy to round it out. Once it rounds out the only way to remove them is will a drill or drill press which is a pain."

Sounds like pretty much the same situation you're facing here.

Well, I tried to remove them with a torx driver and I managed to unscrew two before I managed to 'round out' the other two. I could take a drill to it but that will defeat the purpose of mounting it on another CPU so I just put the removed screwsback on. That cooler will just stay with that CPU for life now.

Did you apply oil to lubricate before trying to back them out? I warned you. 😀

Now you get the impact wrench with the 1/2" to 3/8" drive adapter and the 3/8" to 6mm hex and chuck in a fresh bit.... and set it to max, and TIGHTEN, snap the damn things off 🤣

Took a look at the C code I wrote for an extremely large number calculator while I was in college around 20 years ago.

Uses no large number library and precision can go to infinite decimal places.

I figured most of it out by working out the calculations on paper.

Some of my comments in the code are quite funny.

The numbered revisions of the code go to 195.

Looks like I was working on converting it to do calculations faster by using arrays of 64-bit ints instead of character arrays.

I plan on working on it again and possibly merging it with another program in order to be able to figure out insanely large prime numbers very very quickly.

Shponglefan wrote on 2022-03-25, 00:12:

stef80 wrote on 2022-03-24, 22:56:

You should mainly be concerned with 5V rail for socket A.

Looking at the old PSU I took the system from, it's got +5V @ 22A. Whereas this new PSU is rated +5V @ 20A.

Looks like 1000W Seasonic PSUs are rated at +5V @ 25A. Perhaps I should have gone for one of those instead?

Enermax 350W (few posts above) have 32A on 3.3 and 5V 😀. You can find them NOS for about 20-25 $/€.
It would make excellent retro PSU, if recapped with quality caps. Original ones are so-so (not the best period in time, caps-wise).

Kahenraz wrote on 2022-03-25, 00:58:

Did you apply oil to lubricate before trying to back them out? I warned you. 😀

Yeah, I tried oil as well as WD40 to break the rust and that inhuman grip, but it only worked on two of the screws. That thing is bolted onto there by Thor himself.

Not only the capacitors but I'd say the amount of soldering on some of those (good) old PSU is a lot.. I've opened different ones of the 350W Pentium4/Prescott oriented line and while they seems to have quite many components for those times and they are indeed heavy, once seen the back side of the PCB it felt like nowdays they are using like half the amount of metal to solder the traces.