First post, by aspiringnobody
Well,
I'm having a bad vintage PC week. I recently discovered that my working-totally-fine gateway PC that I upgraded with a modded Tualatin 1.4GHz drew way too much power from the Gateway's rather anemic VRM. So, I ordered an ASUS CUSL2 hoping that all my cares would fade away. I can't even understate how wrong I was (lol). Sadly, I passed up a TUSL2 that was confirmed working, but with leaking caps, because I wasn't sure if my modded Tualatin would work in a real Tualatin motherboard. It sold for like $30.
So, fast forward to yesterday, and my CUSL2 arrived. I excitedly opened up the packaging and got to work taking out the Gateway OEM, Intel made D815EEA and replacing it with the much lauded ASUS CUSL2. Had some fitment issues (but got those resolved) -- and BAM! -- no post. Put in the post code analyzer. No activity at all. CPU was getting warm, so definitely getting power. Pulled out my Tualatin 1266 and popped it in -- instantly came to life!
It was at this point I went down a bit of a rabbit hole. You see, the Tualatin 1266 came pre-modified from the Korean eBayer who sells them. I eventually decided I wanted the fastest possible, so I tried to source a 1.4 the same way -- but he doesn't sell them. So I ordered his PCB to do the mod myself (I figured, hey, I can do that!). Success -- it worked for months in my D815EEA. I didn't notice this until yesterday, but the thickness of the PCB I received for the mod is easily three times the thickness of the ones he uses for his ready-made CPUs. Either that, or between when I ordered the first one and when I ordered the PCB he's switched to the cheaper, thicker, PCB. So a word of caution there -- they might not be as good as they used to be.
Anyway, having figured this out I decided I must be having some kind of CPU contact issue -- and thus I spent a couple of hours trying different ways of seating the processor. Trying to get every last micrometer of pin out of it I could. Finally, I gave up. I gently placed it in the socket and closed the latch with no pressure on the CPU. I was going to store it away, and I wanted to protect the socket and the pins. For some reason I tried one more time -- and BOOM! activity on the post code analyzer!! So less pressure seemed to be good... weird, right?
I eventually, after trying some other things, reduced the tension on the heatsink and got it more or less posting reliably. However, booting to windows was a crapshoot. Sometimes, it boots fine. Sometimes it hangs. Sometimes scandisk tries to start but crashes. Sometimes scandisk starts and completes normally. But if/when I got to windows, should I try to run 3dmark -- 100% locks up or reboots before it finishes the first test. Checked the forums. Someone else reported (as a response to someone else's thread) that they tried the pin mod on their Tualatin but couldn't get it stable. VGA drivers would always get corrupted, something along those lines. He/She decided that they messed up their CPU trying to mod it and never came to a resolution.
I took a step back and thought that maybe the Windows ME install I was on (that had originated with the D815EEA) might be corrupted in some way, so I reinstalled Windows. Made it through the entire installation without a hitch. Installed the chipset drivers (which said they weren't needed on my version of Windows) and the Intel IDE driver. Installed the drivers for my ATi Radeon 9600XT. Installed DirectX 9.0c. Rebooted countless times. Solid as a rock.
Went in for the moment of truth -- launched 3dmark03. It made it all the way to the end of the first benchmark. Farther than I've seen it go since this all started. And then, without warning, it rebooted right at the end of Wings of Fury. Windows ME tried to load, got to scandisk, and the computer locked up hard. I loaded up memtest v2 and went to bed for the night -- hoping desperately that I'd find bad RAM or something that I could fix.
Which brings us to today -- and 8 successful passes of memtest awaiting me when I woke up. So much for easy. Today at work I took the hot air rework station to the bottom of my poor Tualatin 1.4 and cooked it good trying to mash the PCB down as hard against the pins as possible to eek out any last depth I could on the pins. I'll try that when I get home but I don't have much faith. After that I'm not sure where to go next, but whatever I do will be chronicled in the thread below.
Such a disappointing start to my journey with one of the most beloved vintage motherboards this side of the P2B.
- Evan