Got a few things for a "good" price.
Not all items are pictured - among them an ASUS PCI/I-P5MP3 board and a small lot of Pentium60 cpus(luckily some with fdiv bug).
I'm most delighted about the MB-8433UUD... PS/2 mouse support out of the box seems like a blessing.
I also need to buy a faster crystal for the P5MP3 and since I'm naughty 'll order 75- and 80- besides the "casual" 66mhz.

zuldan wrote on 2025-01-09, 09:03:

zuldan wrote on 2025-01-07, 10:36:

Got a mystery pack today. Here is the contents. Not sure how rare the 3700 is. Need to do some research on the 7600.

Well turns out the mystery pack was worth it. This 3700+ (ADA3700AEP5AR) is for socket 754. It’s the fastest desktop processor for the 754. They seem to be pretty rare. Only 1 sold of eBay. What GPU would be a good match for this processor?

Nice find !

AGP Radeon x800 or GeForce 6800 ? (mid/late 2004)
With an nforce3 or K8T800 motherboard

It's the fastest desktop s754.... but there is a faster Mobile 4000+ (released one year later) (https://en.wikipedia.org/wiki/List_of_A ... ,_62W_TDP))

Sony Vaio PCG-F809K arrived.

Good points; it's in decent condition - case isn't overly scratched (it's mostly dirt), and the screen looks unmarked.

Bad points:

I failed to see that the battery compartment cover was missing, so the battery flops about crazily without any means of keeping it installed 🙁 ... and of course it doesn't power on (as expected). It *does* have corrosion inside, near the RTC battery module/lid sensor/battery connector, though not nearly as bad as documented here: If you own a Sony VAIO PCG-Fxxx series laptop, inspect it for corrosion as soon as possible!

zuldan wrote on 2025-01-09, 09:03:

zuldan wrote on 2025-01-07, 10:36:

Got a mystery pack today. Here is the contents. Not sure how rare the 3700 is. Need to do some research on the 7600.

Well turns out the mystery pack was worth it. This 3700+ (ADA3700AEP5AR) is for socket 754. It’s the fastest desktop processor for the 754. They seem to be pretty rare. Only 1 sold of eBay. What GPU would be a good match for this processor?

There was an undocumented desktop s754 3800 Athlon as well. I only ever saw one (I had it in my collection a few years ago) and never found anything online or in AMD documents, but it worked and was detected right. Can't remember details, and as I've never seen one in another collection, I would freely describe it as very rare 😀

debs3759 wrote on 2025-01-09, 15:54:

zuldan wrote on 2025-01-09, 09:03:

zuldan wrote on 2025-01-07, 10:36:

Got a mystery pack today. Here is the contents. Not sure how rare the 3700 is. Need to do some research on the 7600.

Well turns out the mystery pack was worth it. This 3700+ (ADA3700AEP5AR) is for socket 754. It’s the fastest desktop processor for the 754. They seem to be pretty rare. Only 1 sold of eBay. What GPU would be a good match for this processor?

There was an undocumented desktop s754 3800 Athlon as well. I only ever saw one (I had it in my collection a few years ago) and never found anything online or in AMD documents, but it worked and was detected right. Can't remember details, and as I've never seen one in another collection, I would freely describe it as very rare 😀

Was it any faster though? I remember at one point AMD switched up the ratings/numbers, because initially the "PR" was the "TR" Thunderbird rating from the XP, how high clock Thunderbird core it represented in performance. Then some way into the A64s they switched up the basis, I forget what to, but that's also when the Sempron numbers seemed to get inflated, so they were on a different scale I think. Anyway, thinking it might have been marked in the last week of production before changing to new scale or something.

megatron-uk wrote on 2025-01-09, 14:10:

Sony Vaio PCG-F809K arrived. […]
Show full quote

Sony Vaio PCG-F809K arrived.

The attachment IMG20250109130444.jpg is no longer available

The attachment IMG20250109130609.jpg is no longer available

Good points; it's in decent condition - case isn't overly scratched (it's mostly dirt), and the screen looks unmarked.

Bad points:

The attachment IMG20250109130525.jpg is no longer available

The attachment IMG20250109132358.jpg is no longer available

I failed to see that the battery compartment cover was missing, so the battery flops about crazily without any means of keeping it installed 🙁 ... and of course it doesn't power on (as expected). It *does* have corrosion inside, near the RTC battery module/lid sensor/battery connector, though not nearly as bad as documented here: If you own a Sony VAIO PCG-Fxxx series laptop, inspect it for corrosion as soon as possible!

Huh, what does it do when you try to power it? I assume you're using the stock charger?
I ask because one of my F807K laptops that I made up from spare parts has something weird with its embedded controller / power microcontroller, in that it will only run when the DC input is 18 to 18.4 volts rather than the standard 19 to 19.5. The laptop works well enough off my bench PSU that I'm not going to troubleshoot it further just yet but I wonder if that weirdness could affect other Vaio laptops, I have no idea how that could have gone wrong on mine.

If you want to troubleshoot it, it's best to run the laptop from a bench power supply ideally with good power consumption metering since you can see if it's drawing DC power at all. It could be something as simple as a broken DC jack or the fuse. Or if there's no reaction to the power button at all, perhaps the flex cable for the button board is damaged.

Here's a picture of a neomagic version of the mainboard for a quick example of its innards:

All the main caps are polymer tantalums which don't go wrong too often, but perhaps there's a short on the mainboard stopping it from coming on.
Regarding your mention of the "blink of death" in the other thread, it shouldn't be that because the speedstep controller is on the processor MMC module and the ones I've seen don't use the ADP3421. Reseating some of those sub-boards with contact cleaner might help though, since both the CPU and BIOS chip are both removable.

Thermalwrong wrote on 2025-01-09, 16:55:

Huh, what does it do when you try to power it? I assume you're using the stock charger? I ask because one of my F807K laptops th […]
Show full quote

megatron-uk wrote on 2025-01-09, 14:10:

Sony Vaio PCG-F809K arrived. […]
Show full quote

Sony Vaio PCG-F809K arrived.

The attachment IMG20250109130444.jpg is no longer available

The attachment IMG20250109130609.jpg is no longer available

Good points; it's in decent condition - case isn't overly scratched (it's mostly dirt), and the screen looks unmarked.

Bad points:

The attachment IMG20250109130525.jpg is no longer available

The attachment IMG20250109132358.jpg is no longer available

I failed to see that the battery compartment cover was missing, so the battery flops about crazily without any means of keeping it installed 🙁 ... and of course it doesn't power on (as expected). It *does* have corrosion inside, near the RTC battery module/lid sensor/battery connector, though not nearly as bad as documented here: If you own a Sony VAIO PCG-Fxxx series laptop, inspect it for corrosion as soon as possible!

Huh, what does it do when you try to power it? I assume you're using the stock charger?
I ask because one of my F807K laptops that I made up from spare parts has something weird with its embedded controller / power microcontroller, in that it will only run when the DC input is 18 to 18.4 volts rather than the standard 19 to 19.5. The laptop works well enough off my bench PSU that I'm not going to troubleshoot it further just yet but I wonder if that weirdness could affect other Vaio laptops, I have no idea how that could have gone wrong on mine.

If you want to troubleshoot it, it's best to run the laptop from a bench power supply ideally with good power consumption metering since you can see if it's drawing DC power at all. It could be something as simple as a broken DC jack or the fuse. Or if there's no reaction to the power button at all, perhaps the flex cable for the button board is damaged.

Here's a picture of a neomagic version of the mainboard for a quick example of its innards:

The attachment f707-mainboard.jpg is no longer available

All the main caps are polymer tantalums which don't go wrong too often, but perhaps there's a short on the mainboard stopping it from coming on.
Regarding your mention of the "blink of death" in the other thread, it shouldn't be that because the speedstep controller is on the processor MMC module and the ones I've seen don't use the ADP3421. Reseating some of those sub-boards with contact cleaner might help though, since both the CPU and BIOS chip are both removable.

It does precisely nothing. No blink, no flash from the led's along the bottom of the monitor. No pops, speaker noises or release of magic smoke.

I have acquired a standard, 19v, PCGA-AC19 adapter, which I believe was the original part for these machines - it's a genuine Sony part; 19.5V DC at 3.3A.

The Neomagic board looks only slightly different to my ATI version around the ATI chip itself, but largely appears to be the same:

I can't believe that the power control board/sub-board was damaged anywhere near enough to cause it to not power on... unless there's something on these sub-boards which just dies:

Very small amount of corrosion on mine (and a tiny amount of discolouring of the chassis around it), it cleaned up easily with white vinegar and 99% isopropyl.

The MMC-2 module used in this appears to use an ADP3420.

I'll go hunting for 19v DC on the motherboard next.

megatron-uk wrote on 2025-01-09, 17:12:

It does precisely nothing. No blink, no flash from the led's along the bottom of the monitor. No pops, speaker noises or release […]
Show full quote

Thermalwrong wrote on 2025-01-09, 16:55:

Huh, what does it do when you try to power it? I assume you're using the stock charger? I ask because one of my F807K laptops th […]
Show full quote

megatron-uk wrote on 2025-01-09, 14:10:

Sony Vaio PCG-F809K arrived. […]
Show full quote

Sony Vaio PCG-F809K arrived.

The attachment IMG20250109130444.jpg is no longer available

The attachment IMG20250109130609.jpg is no longer available

Good points; it's in decent condition - case isn't overly scratched (it's mostly dirt), and the screen looks unmarked.

Bad points:

The attachment IMG20250109130525.jpg is no longer available

The attachment IMG20250109132358.jpg is no longer available

I failed to see that the battery compartment cover was missing, so the battery flops about crazily without any means of keeping it installed 🙁 ... and of course it doesn't power on (as expected). It *does* have corrosion inside, near the RTC battery module/lid sensor/battery connector, though not nearly as bad as documented here: If you own a Sony VAIO PCG-Fxxx series laptop, inspect it for corrosion as soon as possible!

Huh, what does it do when you try to power it? I assume you're using the stock charger?
I ask because one of my F807K laptops that I made up from spare parts has something weird with its embedded controller / power microcontroller, in that it will only run when the DC input is 18 to 18.4 volts rather than the standard 19 to 19.5. The laptop works well enough off my bench PSU that I'm not going to troubleshoot it further just yet but I wonder if that weirdness could affect other Vaio laptops, I have no idea how that could have gone wrong on mine.

If you want to troubleshoot it, it's best to run the laptop from a bench power supply ideally with good power consumption metering since you can see if it's drawing DC power at all. It could be something as simple as a broken DC jack or the fuse. Or if there's no reaction to the power button at all, perhaps the flex cable for the button board is damaged.

Here's a picture of a neomagic version of the mainboard for a quick example of its innards:

The attachment f707-mainboard.jpg is no longer available

All the main caps are polymer tantalums which don't go wrong too often, but perhaps there's a short on the mainboard stopping it from coming on.
Regarding your mention of the "blink of death" in the other thread, it shouldn't be that because the speedstep controller is on the processor MMC module and the ones I've seen don't use the ADP3421. Reseating some of those sub-boards with contact cleaner might help though, since both the CPU and BIOS chip are both removable.

It does precisely nothing. No blink, no flash from the led's along the bottom of the monitor. No pops, speaker noises or release of magic smoke.

I have acquired a standard, 19v, PCGA-AC19 adapter, which I believe was the original part for these machines - it's a genuine Sony part; 19.5V DC at 3.3A.

The Neomagic board looks only slightly different to my ATI version around the ATI chip itself, but largely appears to be the same:

The attachment IMG20250109170406.jpg is no longer available

I can't believe that the power control board/sub-board was damaged anywhere near enough to cause it to not power on... unless there's something on these sub-boards which just dies:

The attachment IMG20250109135910.jpg is no longer available

Very small amount of corrosion on mine (and a tiny amount of discolouring of the chassis around it), it cleaned up easily with white vinegar and 99% isopropyl.

The MMC-2 module used in this appears to use an ADP3420.

That might be good, no flash at all means that it's not detecting a short and shutting off. Or isn't getting to that point yet.
Do you have a bench power supply? It might be worth doing what I did with my dodgy F807K where I started at 17v and ramped up the voltage until I could see the laptop start reacting when I pressed the power button.
Also with a bench power supply you could possibly see what current it's drawing when attempting to power.
Hopefully you get the F807K fairly soon then you could swap around parts to compare 😀

I'll go hunting for 19v DC on the motherboard next.

Good idea 😀 Since you've got the screen off you can easily check the DC jack and fuse area at least.

Not very Retro...but I found this card interesting and it was very cheap. "GeForce 9300GS PCI 512M".

I do not find very much about it. Maybe its not a common card. All hits I get are about the PCI-e variant.

At least it works and its PCI with DVI and 512M RAM 😀

Thermalwrong wrote on 2025-01-09, 17:19:

]
Good idea 😀 Since you've got the screen off you can easily check the DC jack and fuse area at least.

No, I don't have a bench PSU. I'll try probing for voltage first, then try swapping the sub-board from the 807k when it arrives (or the spare, stand alone, sub-board (unknown condition) I have been able to source from a parts supplier).

If that still doesn't work, I'll try finding a power supply to try your suggestion.

I have high hopes for swapping the sub-board from the 807k...

Predator99 wrote on 2025-01-09, 17:20:

Not very Retro...but I found this card interesting and it was very cheap. "GeForce 9300GS PCI 512M". […]
Show full quote

Not very Retro...but I found this card interesting and it was very cheap. "GeForce 9300GS PCI 512M".

I do not find very much about it. Maybe its not a common card. All hits I get are about the PCI-e variant.

At least it works and its PCI with DVI and 512M RAM 😀

The attachment 20250109_180159.jpg is no longer available

The attachment 20250109_180206.jpg is no longer available

The attachment 20250109_180113.jpg is no longer available

Tested in some old non AGP board?
This is a strange card.

debs3759 wrote on 2025-01-09, 15:54:

zuldan wrote on 2025-01-09, 09:03:

zuldan wrote on 2025-01-07, 10:36:

Got a mystery pack today. Here is the contents. Not sure how rare the 3700 is. Need to do some research on the 7600.

Well turns out the mystery pack was worth it. This 3700+ (ADA3700AEP5AR) is for socket 754. It’s the fastest desktop processor for the 754. They seem to be pretty rare. Only 1 sold of eBay. What GPU would be a good match for this processor?

There was an undocumented desktop s754 3800 Athlon as well. I only ever saw one (I had it in my collection a few years ago) and never found anything online or in AMD documents, but it worked and was detected right. Can't remember details, and as I've never seen one in another collection, I would freely describe it as very rare 😀

Interesting. I think finding undocumented processors from after the year 2000 (that aren't engineering samples) is basically unheard of these days, since there are so many people and websites dedicated to CPU collecting. Do you have any pictures of it, or any records from when you got it or sold it?

Predator99 wrote on 2025-01-09, 17:20:

Not very Retro...but I found this card interesting and it was very cheap. "GeForce 9300GS PCI 512M". […]
Show full quote

Not very Retro...but I found this card interesting and it was very cheap. "GeForce 9300GS PCI 512M".

I do not find very much about it. Maybe its not a common card. All hits I get are about the PCI-e variant.

At least it works and its PCI with DVI and 512M RAM 😀

The attachment 20250109_180159.jpg is no longer available

The attachment 20250109_180206.jpg is no longer available

The attachment 20250109_180113.jpg is no longer available

I guess the 9400GT PCI was too spicy for some ppl. 🤣 It's gotta be almost the same with slightly slower clocks right?

BitWrangler wrote on 2025-01-09, 19:13:

Predator99 wrote on 2025-01-09, 17:20:

Not very Retro...but I found this card interesting and it was very cheap. "GeForce 9300GS PCI 512M". […]
Show full quote

Not very Retro...but I found this card interesting and it was very cheap. "GeForce 9300GS PCI 512M".

I do not find very much about it. Maybe its not a common card. All hits I get are about the PCI-e variant.

At least it works and its PCI with DVI and 512M RAM 😀

The attachment 20250109_180159.jpg is no longer available

The attachment 20250109_180206.jpg is no longer available

The attachment 20250109_180113.jpg is no longer available

I guess the 9400GT PCI was too spicy for some ppl. 🤣 It's gotta be almost the same with slightly slower clocks right?

It likely has half the memory bus width and half the unified shaders, going from the 9300GS vs 9400GT specs online.

I actually ran some tests with a 9400GT PCI a month or so ago. It was in my usual Athlon XP test system, and while it was definitely the overall highest-scoring PCI card I have tested in 3dmark 2001SE, the experience was NOT smooth. Having such a huge bottleneck on such a relatively powerful card makes for some crazy performance swings. One second it's 30fps, the next its 130fps... it's just all over the place.

I also stumbled upon a PCI Radeon HD5450 last year, and I intend to test that as well. I'm curious which card manages the PCI bandwidth better.

... I'm guessing neither would provide a good experience though. Maybe if you could lock game frame rates to 30 or 45 or something, it might be able to solidly provide that in games from the very early 2000s. But even then there will probably be hitches.

PCI was just so horribly limiting... it was completely dead for gaming by maybe 2005, with the last card that made sense for light gaming being the Geforce 6200, and even that was borderline.

I think if someone had continued to make AGP 8x cards though, they probably would have been viable all the way up to Nvidia's Kepler or first gen Maxwell. Imagine a GTX 680 or 750 Ti in AGP... 😮

Ozzuneoj wrote on 2025-01-09, 19:08:

debs3759 wrote on 2025-01-09, 15:54:

zuldan wrote on 2025-01-09, 09:03:

Well turns out the mystery pack was worth it. This 3700+ (ADA3700AEP5AR) is for socket 754. It’s the fastest desktop processor for the 754. They seem to be pretty rare. Only 1 sold of eBay. What GPU would be a good match for this processor?

There was an undocumented desktop s754 3800 Athlon as well. I only ever saw one (I had it in my collection a few years ago) and never found anything online or in AMD documents, but it worked and was detected right. Can't remember details, and as I've never seen one in another collection, I would freely describe it as very rare 😀

Interesting. I think finding undocumented processors from after the year 2000 (that aren't engineering samples) is basically unheard of these days, since there are so many people and websites dedicated to CPU collecting. Do you have any pictures of it, or any records from when you got it or sold it?

I have been terrible at keeping proper records. I can tell you I sold it in the CPU-World forums some years ago, but searching without the full part number would take me longer than I could handle (very poor health, extreme fatigue and a kidney stone are conspiring to keep me inactive).

debs3759 wrote on 2025-01-09, 19:44:

Ozzuneoj wrote on 2025-01-09, 19:08:

debs3759 wrote on 2025-01-09, 15:54:

There was an undocumented desktop s754 3800 Athlon as well. I only ever saw one (I had it in my collection a few years ago) and never found anything online or in AMD documents, but it worked and was detected right. Can't remember details, and as I've never seen one in another collection, I would freely describe it as very rare 😀

Interesting. I think finding undocumented processors from after the year 2000 (that aren't engineering samples) is basically unheard of these days, since there are so many people and websites dedicated to CPU collecting. Do you have any pictures of it, or any records from when you got it or sold it?

I have been terrible at keeping proper records. I can tell you I sold it in the CPU-World forums some years ago, but searching without the full part number would take me longer than I could handle (very poor health, extreme fatigue and a kidney stone are conspiring to keep me inactive).

Very sorry to hear about your health problems.

I was just curious and I have searched CPU World forums and found nothing about it. No one has listed it on the CPU World list for Socket 754 processors either, even as an engineering sample.

https://web.archive.org/web/20240330221509/ht … thlon%2064.html

(archived page from a couple months ago because CPU-World is currently offline to prevent bots from scraping the site)

Anyway, no big deal. It's just not every day that an undocumented 20 year old processor comes up, especially from a socket as short lived as Socket 754 (with so few other processors for it to get lost among). There aren't a whole lot of stones left unturned to find stuff like this, you know?

Ozzuneoj wrote on 2025-01-09, 19:23:

It likely has half the memory bus width and half the unified shaders, going from the 9300GS vs 9400GT specs online. […]
Show full quote

BitWrangler wrote on 2025-01-09, 19:13:

Predator99 wrote on 2025-01-09, 17:20:

Not very Retro...but I found this card interesting and it was very cheap. "GeForce 9300GS PCI 512M". […]
Show full quote

Not very Retro...but I found this card interesting and it was very cheap. "GeForce 9300GS PCI 512M".

I do not find very much about it. Maybe its not a common card. All hits I get are about the PCI-e variant.

At least it works and its PCI with DVI and 512M RAM 😀

The attachment 20250109_180159.jpg is no longer available

The attachment 20250109_180206.jpg is no longer available

The attachment 20250109_180113.jpg is no longer available

I guess the 9400GT PCI was too spicy for some ppl. 🤣 It's gotta be almost the same with slightly slower clocks right?

It likely has half the memory bus width and half the unified shaders, going from the 9300GS vs 9400GT specs online.

I actually ran some tests with a 9400GT PCI a month or so ago. It was in my usual Athlon XP test system, and while it was definitely the overall highest-scoring PCI card I have tested in 3dmark 2001SE, the experience was NOT smooth. Having such a huge bottleneck on such a relatively powerful card makes for some crazy performance swings. One second it's 30fps, the next its 130fps... it's just all over the place.

I also stumbled upon a PCI Radeon HD5450 last year, and I intend to test that as well. I'm curious which card manages the PCI bandwidth better.

... I'm guessing neither would provide a good experience though. Maybe if you could lock game frame rates to 30 or 45 or something, it might be able to solidly provide that in games from the very early 2000s. But even then there will probably be hitches.

PCI was just so horribly limiting... it was completely dead for gaming by maybe 2005, with the last card that made sense for light gaming being the Geforce 6200, and even that was borderline.

I think if someone had continued to make AGP 8x cards though, they probably would have been viable all the way up to Nvidia's Kepler or first gen Maxwell. Imagine a GTX 680 or 750 Ti in AGP... 😮

I didn't test mine super well, but it seemed to be about identical to the 8400 PCIe I tried in the same system. I didn't see any spiky performance, but probably didn't go back much, just "sober" ploddy 30fps things mostly. I guess if it's really lumpy in real use though, maybe the HD2400Pro PCI isn't so bad, with only about 60% theoretical perf of 9400. Somehow lost track of what I was intending these for, had a couple of tualatin boards with no AGP and they'll be fine with GF6200. Thought I had something s775 that took early C2D but forgot what. The Johnstown Atom board isn't likely to profit much from anything beyond the GF6200 either.

I got these mem sticks this week. The Corsair ones are for a 2005 high end PC with a P4 670, an Asus P5ND2 SLI and a X1800XT 512MB. And they run incredible, 675MHz CL2-3-2-5 seems to be stable.

The OCZ Sticks I haven't tested yet but they are the first sticks I've seen which are rated for 800MHz CL3, so they are a welcome addition to my little collection.

I've seen those memory heatsinks with the little copper heatpipe before. I don't know if it has any practical use, but it's super cool.