So, I've set this LuckStar up three times now and have full function every time. No clue what was going on.

The POST card effects nothing. This last time, I installed the wrong video card, but still functioned perfectly (just way faster). I even added a network card and transferred about 16GB of games. Definitely going to call this GOOD and add it to the collection.

Over the weekend I cleaned and benched a Shuttle FB61 with Pentium 4 SL6WK (3.0/800/512), 2x1GB DDR400, and a very similar Radeon 9600.

I had to use my normal bench PSU (300W Enermax), since the 200W PSU in the Shuttle had plague™ and was broken, most likely from not being powerful enough. It's going to get a nice 350W retrofitted (needs ATX and P4 connectors).

ANYWAY, to the point: it was a similar setup to what I was messing with earlier and now that I got the 9600 in the board, I can compare the "same" video card in two similar systems with wildly different CPU clocks and memory, especially since the LuckyStar has 133Mhz SDRAM. The results are expected, but now quantified.

I had my brothers' Shuttle race my Shuttle, for funsies; mine got blown out of the water. Of course, it DID finish first, but not a fair race, as the GF4 was missing many feature tests. My Shuttle runs Athlon 64X2-3800+(2.0Ghz) 2x512MB DDR400, and GeForce 4 MX4000. I didn't get pictures of his while it was here, but it was way slicker-looking than mine, with a 2mm aluminum plate on the front.

Going to do a recap now. Not sure what, but it's going to get done, 🤣

I decided to do PCChips M861g(1.6a) because it looked like I could get it done quickly, not many caps.

I initially was just going to to the power stage, it had some puffers, but I ended up doing the whole thing. 34 on this one. Most of them tested fine, just the puffers were actually bad.

I'm about to test it soon. If it works, I'll prepare a drive for it. I recently got a pile of 40 and 80 GB Seagate 7200.7, so I've been pairing boards to drives.

Alright! It POSTed first try and then sailed right into DOS and completed the first test. It failed the memory test, but I'm not surprised by that.

What I AM surprised about, is how many boards this stick has gone through and NONE of them have detected the failure. (this is a 1GB stick, always detected as 512MB) After the first restart, the BIOS also told me my memory was crap. I've never seen it straight up say "Extended memory test FAILED, F1 to continue" or something like that. I switched the sticks out for two random ones, both worked and passed both BIOS and Speedsys memory tests. I ended up changing RAM yet again, so I could tune it properly (was DDR333, now I got 2x512MB DDR400).

All my testing was with the onboard VGA so far. WindowsXP doesn't know anything about it and the Hyperion drivers did not contain video drivers. I didn't want to bother with the onboard anyway, so I tried installing the 9600LE. It wouldn't output video and after a few reinserts, wouldn't POST anymore. Removed, it POSTed, but the 9600LE was NOT working. I tried the universal GF2MX200: it POSTed AND gave me video.

Next problem:
3Dmark2001 crashes as soon as the first test starts to load. Ugh. I felt I already knew the problem, so I went right for that: the VIA driver for AGP. I just reinstalled the microsoft basic driver in the device manager and restarted. It was now able to run 3Dmark2001... almost. It crashes randomly now. HARD crash, dead-reboot.

I'm working as I'm typing, so I got updates already. My gut was telling me what's wrong. The PSU. I ran out of PSU! I swapped the PSU (300W Enermax) with a Antec TruePower 650 and it was able to complete the test! It is also a much smoother experience at the desktop.

The 9600LE works in the board now, too! I've tried doing another run with the 9600LE and got the boost expected.

Seems like a good board so far. I don't like the driver situation, but it works pretty good. I'll continue testing.

While I was doing that, I also disassembled the empty XT -style frame and will wash it before getting some fitment details worked out. I'm going to need to make a few parts for it to be completely usable.

I don't know the status of the PSU that came with it. It was opened before and will need to be again.

Just did some quick mock-up to get to know what I'm working with. I started with a "modern" board.

I then tried a board without audio, to add more clearance for the expansion slot bracket.

I moved to try and take care of the other clearance issue: the drives

The easiest route, of course, is a baby AT board, but...

This Gemlight, I knew was going to fit already, so I also threw a bunch of stuff on top to eye-ball clearances.

Not sure which direction I want to go yet with that.

The M861 is still trooping along well, no issues. I'll call it good by tomorrow.

I have some time-off this week, so I'll be trying to re-arrange and get more surface area.

Today, I made a quick repair to the Boston BA-7800s I got a while back. I tested them today and had flaky volume and basically no "notch" on the two that were supposed to have it. It comes apart easy, the face has to be unbolted from the pots and one screw on the back.

I'm glad nothing serious was going on once I had one open. I just cleaned it with alcohol and added some grease.

This is not computer related, but an important issue:
The timing belt on my Passat broke and completely took the valves out with it. I got lucky on a 04 Honda Accord, but I will have to spend the whole first day of my vacation making it drivable.

I will feel better once the Honda is safe to drive. I was thinking about sharing the pic of the car, but you can't see the mechanical failure. What you CAN see is this:

I test drove the car after receiving it (about 3-4 miles) and noticed this AFTER... Looking closer around, that sh*t is EVERYWHERE and there were little mushroom-looking guys growing around the floormats. But, hey, the price was indeed right. I did detail work years ago and still have my equipment, it's just going to take some time.

I started working on the Biostar M7VIG 400.

I was feeling adventurous and removed ALL the caps at once! I think this method will cut time, but not for the feint of heart. There are only 3 smaller sizes and 2 larger ones on this board, so I figured this would be an easy one to play with. 25V22µF, 16V100µF, and 16V1000 are the small ones, 16V2200µF and 6.3V3300µF are the large ones. ALL of the 6.3V3300 tested BAD.

Cleaning it without the caps was great. Real easy to scrub the normally hard-to-reach area(s). All around good time. This is going to be fun to re-populate.

I had to take some creative liberties when replacing the caps this time. Most of the 3300s were replaced with 2200s and two of the three 16v2200µF were substituted with 1800µF. This was due to not having the correct parts. Should work fine still.

I'll use this Sempron for testing:

Is there a specific name for that plate/spacer thing? I call it "Athlon Armor" It seems like a cool thing to get for my other Athlon systems.

Huzzah! Came right on!

Passes initial test, so I'm installing XP. I have a board with a similar chipset, so if I do it next (ECS L7VMM2), I can streamline the loading-into-windows part of the experience.

I ditched the onboard VGA and went with 2x512MB DDR333.

Major Jackyl wrote on Today, 02:20:

Is there a specific name for that plate/spacer thing? I call it "Athlon Armor" It seems like a cool thing to get for my other Athlon systems.

Copper shim.
You can still find reviews around. It was a thing back then.