ediflorianUS wrote on 2023-03-21, 09:17:

here is a photo of the lot (not sure how real it is):

Niiiiice!
Are you going to test them all?

bearking wrote on 2023-03-21, 09:45:

ediflorianUS wrote on 2023-03-21, 09:17:

here is a photo of the lot (not sure how real it is):

In my opinion, that is all garbage! Maybe it is worth 50 euros...

I wouldn't touch it myself.

TrashPanda wrote on 2023-03-21, 16:09:

bearking wrote on 2023-03-21, 09:45:

ediflorianUS wrote on 2023-03-21, 09:17:

here is a photo of the lot (not sure how real it is):

In my opinion, that is all garbage! Maybe it is worth 50 euros...

I wouldn't touch it myself.

Gold scrappers are probably thinking around $2 a chip though, so for a few hundred the number ramps quick. Really you gotta be seeing several pieces that are worth ~$50 or better as usable/collectible CPUs to make it worth the bother. How many times they've been picked is a thing too, if they've been through several sets of hands, you've probably got left with the lowest third of useful value, the lowest third clock speed and the lowest third gold content.

ok so 0.5e/cpu is out of worth-value then.

Sit-rap. (should I start new Marchintosh topic?)
tested evrything ,
- Macbook white works , no backlight
- MacBookPro 2.53 GHz Core 2 Duo (P8700) 5,5 - A1278 - 2326* made it to work (with transfer mb from the 2351*)
- PowerBook G4 booted-up , broken screen
- Macbook A1150 , C2D 2.16 Ghz 2gb ram , works fine , may have a bad key
- MacBook Pro "Core 2 Duo" 2.4 13" Mid-2010 Specs 2351* - cracked glass and bad mb, had some oxidation and may be missing some smd/half a trace on one chip(work-in-progess) , pulled the screen + magsave-powerconnector. -pwr adpr light code
-MacBook Pro "Core i7" 2.66 15" Mid-2010 A1286 - broken, in so many way's , had oxidation all over board. I am unclear if I will ever be able to fix-it. -pwr adpr light code
-MBP Late 2008 Aluminum - MB466LL/A - MacBook5,1 - A1278 - 2254 -2ghz C2D , booted up , has a bad case
-Same as above , no post. (probably will transfer board and other things I need to make a working MBP Late 08). -pwr adpr light code (possible oxidation mb).

Picked up this neat AIR 486SE board. Lots of discrete PALs/GALs/logic on this one. Seems to have a Dolch-branded Phoenix BIOS. It needs a new Dallas RTC and I’d bet money the NVRAM’s battery is shot. It runs a fixed 33MHz FSB, probably derived via the 66.67 MHz oscillator.

I’ve been curious about the Intel EISA chipset. Apparently, this initial revision was somewhat buggy. Performance is acceptable, but the board lacks L2. If only there were cache interposers.

Also, picked up this Weitek 4167. Figured this was the perfect board with which to give it a spin. It was slightly faster than the DX2-66’s FPU in the Weitek demos. Neat.

EDIT: It also has EISA CFG-controlled onboard parallel/serial. That’s pretty interesting.

two dallas chips ? and two fpu's ?

WJG6260 wrote on 2023-03-23, 00:51:

Picked up this neat AIR 486SE board. Lots of discrete PALs/GALs/logic on this one. Seems to have a Dolch-branded Phoenix BIOS. I […]
Show full quote

Picked up this neat AIR 486SE board. Lots of discrete PALs/GALs/logic on this one. Seems to have a Dolch-branded Phoenix BIOS. It needs a new Dallas RTC and I’d bet money the NVRAM’s battery is shot. It runs a fixed 33MHz FSB, probably derived via the 66.67 MHz oscillator.

I’ve been curious about the Intel EISA chipset. Apparently, this initial revision was somewhat buggy. Performance is acceptable, but the board lacks L2. If only there were cache interposers.

Also, picked up this Weitek 4167. Figured this was the perfect board with which to give it a spin. It was slightly faster than the DX2-66’s FPU in the Weitek demos. Neat.

EDIT: It also has EISA CFG-controlled onboard parallel/serial. That’s pretty interesting.

F5553759-E653-424E-8B0C-318BF0530DDC.jpeg

I think that is one of the most densely populated boards I have ever seen (that wasn't just a memory board).

What exactly does this board do that would require so many additional components?

I don't know anything about how PCB design was done back then, but I sincerely hope all that stuff didn't need to be traced manually. 🤣

nuno14272 wrote on 2023-03-23, 14:15:

two dallas chips ? and two fpu's ?

dallas is not only RTC but also battery backed SRAM, EISA requires more memory to store config
weitek is not a normal FPU, its memory mapped https://www.geekdot.com/weitek-abacus-fpu/

Ozzuneoj wrote on 2023-03-23, 14:37:

I think that is one of the most densely populated boards I have ever seen (that wasn't just a memory board).
What exactly does this board do that would require so many additional components?

Intel really sucked at custom chips and integration before Pentium era. I count ?23? PALs 😮

nuno14272 wrote on 2023-03-23, 14:15:

two dallas chips ? and two fpu's ?

Exactly what rasz_pl said. A lot of EISA boards have “less-conventional” Dallas setups, since they need some sort of battery-backed NVRAM to store EISA configuration information. Most use a 1387/1488/something of that sort, but this uses a separate NVRAM module and more conventional DS1287 for RTC features only. P-EISA boards like those with the HiNT Caesar chipset do not use Dallas NVRAM modules, however. They have starkly different NVRAM setups.

The Weitek is indeed memory-mapped, and it’s sort of a strange one. They were sold for a short time as specialized floating point units, and their performance was supposedly eclipsed by the DX2-66’s time. I say supposedly, because the quick test I ran showed otherwise; that being said, there’s probably still merit in the statement that most end users were better off not paying large sums for a 4167 when the internal co-processor in their 486DX2-66 was more than adequate.

Ozzuneoj wrote on 2023-03-23, 14:37:

I think that is one of the most densely populated boards I have ever seen (that wasn't just a memory board). […]
Show full quote

WJG6260 wrote on 2023-03-23, 00:51:

Picked up this neat AIR 486SE board. Lots of discrete PALs/GALs/logic on this one. Seems to have a Dolch-branded Phoenix BIOS. I […]
Show full quote

Picked up this neat AIR 486SE board. Lots of discrete PALs/GALs/logic on this one. Seems to have a Dolch-branded Phoenix BIOS. It needs a new Dallas RTC and I’d bet money the NVRAM’s battery is shot. It runs a fixed 33MHz FSB, probably derived via the 66.67 MHz oscillator.

I’ve been curious about the Intel EISA chipset. Apparently, this initial revision was somewhat buggy. Performance is acceptable, but the board lacks L2. If only there were cache interposers.

Also, picked up this Weitek 4167. Figured this was the perfect board with which to give it a spin. It was slightly faster than the DX2-66’s FPU in the Weitek demos. Neat.

EDIT: It also has EISA CFG-controlled onboard parallel/serial. That’s pretty interesting.

F5553759-E653-424E-8B0C-318BF0530DDC.jpeg

I think that is one of the most densely populated boards I have ever seen (that wasn't just a memory board).

What exactly does this board do that would require so many additional components?

I don't know anything about how PCB design was done back then, but I sincerely hope all that stuff didn't need to be traced manually. 🤣

Yeah, it’s a mess 🤣

Intel’s EISA chipset is always on boards like this. rasz_pl’s point—that Intel sucked at custom chips—is seemingly accurate, as every single Intel-based EISA board looks like this. I believe there’s only two that are in Baby-AT form factor: this one, and the A.I.R. 486ES. While this board is marked as such, it is 100% a 486SE. The 486ES has 8 EISA slots. I think integration may have improved with the 350DT chipset, which was supposedly the bug-fixed revision. The build quality of this board is tank-like, but the design and layout are absolutely atrocious.

rasz_pl wrote on 2023-03-23, 14:51:

dallas is not only RTC but also battery backed SRAM, EISA requires more memory to store config weitek is not a normal FPU, its m […]
Show full quote

nuno14272 wrote on 2023-03-23, 14:15:

two dallas chips ? and two fpu's ?

dallas is not only RTC but also battery backed SRAM, EISA requires more memory to store config
weitek is not a normal FPU, its memory mapped https://www.geekdot.com/weitek-abacus-fpu/

Ozzuneoj wrote on 2023-03-23, 14:37:

I think that is one of the most densely populated boards I have ever seen (that wasn't just a memory board).
What exactly does this board do that would require so many additional components?

Intel really sucked at custom chips and integration before Pentium era. I count ?23? PALs 😮

The PALs are everywhere on this thing! No wonder it lacks L2—there’s just no room! 🤣

I wonder if Intel EISA chipsets did not sell well because of the sheer cost of implementation. Swaths of discrete logic are everywhere on their early boards, like you noted, and that could not have been beneficial for production costs.

Got for $5 and was listed as not working. It works though and in great condition!

Love the Teac 5.25 drives. Works of art.

Yeah, they're built like a tank.

Just like this NOS Zalman 400 Watt PSU, because +5V lives lines matter. 40 amps, yay.

WJG6260 wrote on 2023-03-23, 00:51:

Picked up this neat AIR 486SE board. Lots of discrete PALs/GALs/logic on this one. Seems to have a Dolch-branded Phoenix BIOS. I […]
Show full quote

Picked up this neat AIR 486SE board. Lots of discrete PALs/GALs/logic on this one. Seems to have a Dolch-branded Phoenix BIOS. It needs a new Dallas RTC and I’d bet money the NVRAM’s battery is shot. It runs a fixed 33MHz FSB, probably derived via the 66.67 MHz oscillator.

I’ve been curious about the Intel EISA chipset. Apparently, this initial revision was somewhat buggy. Performance is acceptable, but the board lacks L2. If only there were cache interposers.

Also, picked up this Weitek 4167. Figured this was the perfect board with which to give it a spin. It was slightly faster than the DX2-66’s FPU in the Weitek demos. Neat.

EDIT: It also has EISA CFG-controlled onboard parallel/serial. That’s pretty interesting.

F5553759-E653-424E-8B0C-318BF0530DDC.jpeg

Beatiful find. Congrats!

nuno14272 wrote on 2023-03-23, 14:15:

two dallas chips ? and two fpu's ?

Weitek is not a "Normal" FPU - only specialized apps can use it.

One Dallas chip is for RTC and storing bios settings, other one is for storing EISA configuration.

Maybe I can still use it.

Found this nice Microsoft PS/2 keyboard at the thrift for $4. Just needs a little cleaning.

Found this CRT, a Samsung Syncmaster 750 S - display - CRT - 17":

(photo from the Web, too lazy to get up from the couch :p )
Playing DUNE or DUNE 2 or any Westwood game is very nice !

Plus it was at a very reasonable price (€35), and like new.

Also, it's not too small, and not too huge. Just perfect. 😀

Picked up this case today for £5. Suspicion that it was kept outside 😀

LewisRaz wrote on 2023-03-25, 17:37:

Picked up this case today for £5. Suspicion that it was kept outside 😀

Could definitely have been in a garage. Mine constantly gets leaves in it in the fall. 😁

Bought an IBM P150+ with the roundish heatspreader and some extra bent pins.
It didn't look too bad but man, was I covered in sweat when I finally straightened them all out!
We both survived.

Bought this nice IBM Aptiva Pentium 150MHz tower. I already had the exact same machine in desktop form factor. For some reason, I felt like it needed a tower as a friend. After cleaning, it looks pretty good and works flawlessly. That dampened sliding door mechanism needed some plastic glue to get it working.

WJG6260 wrote on 2023-03-23, 15:10:

P-EISA boards like those with the HiNT Caesar chipset do not use Dallas NVRAM modules, however. They have starkly different NVRAM setups.

Because later boards store the ESCD data into BIOS main eeprom. This means if you change your hardware too much, the main eeprom would get tired in certain areas pretty fast.