Did some more work testing that 486 motherboard.

I soldered a temporary bodge wire to bypass the broken trace, so I could see if the rest of the board was working. Didn't want to spend time doing a detailed trace repair unless I knew the rest of the board could be fixed.

Fortunately, it powered right up. Looks like that one broken trace was the main issue.

I did run into a problem testing the ISA slots. The 4th slot had a decent amount of corrosion in it. With the POST card, it failed to display the POST readout or the CLK/ FRAME LEDs.

While I had originally cleaned it with some Deoxit, it clearly wasn't enough.

I gave it a few more rounds of Deoxit and some scrubbing with a plastic brush, after which it started working.

Next up, I'll need to fix that trace properly. I'll have to first desolder the keyboard socket, since it's too close to the trace. The keyboard socket also has some corrosion in it, so it might be worth replacing with a new one anyway.

Went through the Trident cards on the ISA486GXI, had only minor problems with them, spoiler the best ones were the ones you'd expect, so were the worst ones. The real shitshow started when I tried to run my Orchid Kelvin 64 for comparison, either the board doesn't like it, or it doesn't like the i/o card, or it's all down to the glitch in the worn VGA socket which I may have to replace. Yeah, couldn't get a stable boot with it, and kept messing up the i/o so hdd and floppy were dropping off. That card seemed a little quirky when I put it in and wasn't eager to go on the the 486VC but seeemed to settle down... now IDK, either took some damage with aborted boots and much on and off, or it's back to behaving "in character" and I was a sucker to trust it... prologic thing with UMC chip. .... Anyhooooo, I could not get numbers for the Kelvin 64 on that board is the long and short of it... I managed after that to get my only ET4000 to run, which is in a Cardinal SnapPlus, and that came in 3rd. I was not sure that was a great comparison card, so tried to get the Ati Graphic Ultra going and i/o hell broke out again, ran fine for some tests but froze on disk access loading the last and then I couldn't get a good boot again... so maybe gotta retire that i/o card or use it for slower machines. So I plan to sort all those numbers out and do a post.... my plan for chip testing in the Tridents beyond go/no-go might be a non starter though, had one with badly seated chips and that gave error codes under POST BIOS string and I think wasn't going to load OS from there. I was thinking they'd still pass and go into a limp mode on 256KB good, but maybe not. However, still might be easier doing 2 chips at a time on a TVGA9000 pass/fail.

In other news, got some grabbers come in so might actually be able to use my logic probinator with Pulsing or something.

Edit: Oh it was a couple of pages back now, just in case it wasn't clear what tridents I meant... Re: What retro activity did you get up to today?

vutt wrote on 2024-10-22, 20:09:

So I have been playing around with my recent acquisition after full recap - MSI MS 6163 Master Pro v2
First I did put officially supported Slot1 Coppermine 933 PIII in and run it trough my regular stress tests - no problems whatsoever. Rock solid at 133FSB speed.

Tested yesterday properly 2nd part of my purchase above - SL-02A++ Slotket. Worked just fine with full auto jumpers with 133FSB speed on MSI MS6163v2. Well I had to choose FSB speed in BIOS manually.
Seems to be solid Coppermine adapter.

FWIW my solid Coppers slotkets are the AA370TS and AOpen's FCPGA adapter. The former worked great on a BE6-II v1.2, and the latter was bundled w/ an Acer S61 board I bought.

I didn't have much time retroing in recent weeks, today I got around to do something with the notebooks I've been sitting on for a while. Saved some data, then I thought I do some benchmarking.

The integrated SiS GPU seems to perform around the level of a TNT2 M64. Which is not all that awful for what it is. It would be a competent W98 laptop.

I intercepted some mid '00s hardware that was bound for the dump. This included two towers with LCD monitors, keyboards, and mice, and a big old Pentium M laptop. All HDDs were removed but the rest is there. One system is an LGA-775 with a Pentium D in it, and the other is AM2 with a Sempron. Haven't tested them yet, since I've been playing with the laptop. The thing is kind of hideous but it has 2GB of RAM and a Radeon X700, so I wanted to run some benches. I had a spare 20GB HDD that I used to use with an Itox G5m100 with Win 2000 already installed. So I plugged that in and have been going through the ritual of trying to find drivers and whatnot. I tried a few ATI video drivers and none wanted to work, so finally I took BlackWingCat's Catalyst 10 Legacy version and monkeyed with the .INF to make it install. Had to add this line:

"Radeon X700 Series" = ati2mtag_RV410, PCI\VEN_1002&DEV_5653

I ran 3D mark '01 and the system shutdown for some reason. So I tried again with the CPU underclocked to 1.33GHz and then it completed with 13669 3Dmarks.

BitWrangler wrote on 2024-10-26, 03:39:

Went through the Trident cards on the ISA486GXI, had only minor problems with them, spoiler the best ones were the ones you'd expect, so were the worst ones. The real shitshow started when I tried to run my Orchid Kelvin 64 for comparison, either the board doesn't like it, or it doesn't like the i/o card, or it's all down to the glitch in the worn VGA socket which I may have to replace. Yeah, couldn't get a stable boot with it, and kept messing up the i/o so hdd and floppy were dropping off. That card seemed a little quirky when I put it in and wasn't eager to go on the the 486VC but seeemed to settle down... now IDK, either took some damage with aborted boots and much on and off, or it's back to behaving "in character" and I was a sucker to trust it... prologic thing with UMC chip. .... Anyhooooo, I could not get numbers for the Kelvin 64 on that board is the long and short of it... I managed after that to get my only ET4000 to run, which is in a Cardinal SnapPlus, and that came in 3rd. I was not sure that was a great comparison card, so tried to get the Ati Graphic Ultra going and i/o hell broke out again, ran fine for some tests but froze on disk access loading the last and then I couldn't get a good boot again... so maybe gotta retire that i/o card or use it for slower machines. So I plan to sort all those numbers out and do a post.... my plan for chip testing in the Tridents beyond go/no-go might be a non starter though, had one with badly seated chips and that gave error codes under POST BIOS string and I think wasn't going to load OS from there. I was thinking they'd still pass and go into a limp mode on 256KB good, but maybe not. However, still might be easier doing 2 chips at a time on a TVGA9000 pass/fail.

In other news, got some grabbers come in so might actually be able to use my logic probinator with Pulsing or something.

Edit: Oh it was a couple of pages back now, just in case it wasn't clear what tridents I meant... Re: What retro activity did you get up to today?

Was able to finangle a doom score out of the ATI this morning, and unleashed shocking results on an unsuspecting public Lowend 486 ISA VGA Graphics Quick Test, Mostly Tridents plus Oak, ATI, Tseng ....

Continued my recent quest of repairing Sound Blaster cards. This time I replaced a broken RAM socket on a Sound Blaster 32 (CT3670).

The replacement socket with metal clips is just so much nicer than the plastic clips. Really wish Creative Labs hadn't cheaped out on these.

Before:

Socket Removal:

After:

Shponglefan wrote on 2024-10-27, 22:09:

Continued my recent quest of repairing Sound Blaster cards. This time I replaced a broken RAM socket on a Sound Blaster 32 (CT3670).

Really nice work. I love it when faulty parts can be easily identified and replaced. With modern hardware everything is often so miniaturized that even if you know where the problem is, you can't do anything about it without some very fine-grade equipment.

dr_st wrote on 2024-10-28, 05:35:

Really nice work. I love it when faulty parts can be easily identified and replaced. With modern hardware everything is often so miniaturized that even if you know where the problem is, you can't do anything about it without some very fine-grade equipment.

Thank you! I agree, older thru-hole stuff is definitely easier to work on. Some modern components are so tiny they're barely visible without a microscope.

Managed to install Win98SE on a G41 motherboard with a core2 CPU: https://theretroweb.com/motherboards/s/asus-p5g41-m-le
It was an adventure.
Long story short, I had to make a DOS bootable USB drive with the win98 install files on it, coax the BIOS to boot from the USB (needed to disable USB2 in BIOS and use USB1.1 for that), format the SATA drive while booted in DOS from the USB, copy the setup files to it and start the installer. That is because this motherboard refuses to boot a HDD with DOS on it. The installer copied the needed files, but again after restarting it refused to boot from the HDD to complete the install.
The way I solved it was: on the leftover unpartitioned space on the drive I installed XP SP3 (which went very smoothly) and afterwards used NTLDR to chain boot win98. 😀 This way the win98 setup was able to complete.

That's really bizarre. Why would a G41 board refuse to boot a DOS HDD? Why would USB 2.0 have to be disabled to boot from a USB flash drive? And isn't the ICH7 southbridge 100% ATA compliant?

dr_st wrote on 2024-10-28, 16:03:

That's really bizarre. Why would a G41 board refuse to boot a DOS HDD? Why would USB 2.0 have to be disabled to boot from a USB flash drive? And isn't the ICH7 southbridge 100% ATA compliant?

I wish I had answers to all of those, but I don't. It's just how in practice I found things to be.
There's actually more weirdness going on, like for instance: if in BIOS I set SATA mode to "compatible" instead of "enhanced", and ask it to enumerate both SATA and PATA devices - the BIOS will see a SATA drive ONLY if no PATA devices such a CDROM are connected. If an IDE CDROM is connected, it refuses to detect SATA drives.
And that is true both for BIOS revision 0308 (which the motherboard came with) and 0506 (latest, which I flashed in desperation)

I adopted celeron heatsink to PIII SECCII. To make this possible without removing or damaging the back plastic cover (without cutting off standoffs), a distance must be used. Without it, the heatsink would have no contact with the die. Besides, you have to drill holes, Celeron has a different spacing.
Why? Because I like the look of the low heatsink (it doesn't cover the mobo)

Added a heatsink to the MVP3 chipset on my Epox EP-51MVP3E-M.

In preparation for the 40th anniversary of the AMIGA 1000 in July next year I took on a small endeavor (though not small for me) to learn some new skills and refresh some older skills. Basic outline: Design an Amiga-1000-lookalike-bezel, print and paint it, solder the components on an AMIGA 500++ PCB, tweak, adjust, etc… and put it into a Checkmate 1500 case… and of course, play some games

Nice work. I was having Amiga stuff come to mind from a different angle, but that is turning the story inside out, so I'll start from the beginning...

I am somehow in a "speaker drought" meaning I thought I had dozens of little 8 ohm speakers around, the type to hook up to speaker headers, or for general low-fi noise. They all seem to have disappeared, or I have at some point, put them "somewhere safe" effectively disappearing them. So upon needing a speaker to connect to my motherboard bench tester, I had to scavenge. First thought was a little speaker module intended to be bolted under a dashboard of vehicle, for external speaker for radio or comms equipment. Stick it on the meter to check it's somewhere near 8 ohms, get 1.5 kilohm, what? Okay so sneak a peek inside, it's actually got a tiny amplifier in there with it, guess it takes input modulated on 12V supply... quick think about whether I want to try blowing up my speaker header trying to use it in amplified mode, nah, or steal just the cone out of it, hmmm nope, I might require this for intended purpose and was "borrowing" it... so start looking around again for a donor and see...

An external modem, kicking around since I got it in a box of junk, 14.4 K practical peripherals model, with a tell tale speaker aperture in the case... Okay, let's steal that. cracked it open, find the 2" speaker unhooks readily from the header. But what else do we have in here... oooh, two 1Mbit EPROM, tasty, oh what's that huge chip on the board? 68302 ... a 68k based microcontroller. I didn't clock the clockspeed, but had a chuckle to myself wondering what early 90s customers might have bought one of these and plugged it into a computer less powerful than the dang modem. Seemed a bit like overkill anyway. Looked it up, definitely a 68000 class in there, some sources say more similar to 68020, umm isn't that a 68010 then maybe? Speeds up to 33Mhz, ooh nice.. so am sticking a mental pin in this for further investigation, in case I could use it to make a bodged up classic 68k accelerator for an Amiga 500... a couple to four times the speed is alright for smoothing out some period games, don't really need the overkill ones.

Yay speaker, small though, 2" ? size down from typical PC Speaker, on a mini 2 pin header, in case I think of a reason to put modem back together, I stuff tinned ends of a 4 pin speaker header (strangely I had one with no speaker on) in it and tape it up so they stay there. Plug this bodge into the test board, fire it up, RAM county clicks, ready to go boop, seems to be working... Oh.. let's try Modplayer... gah, copy on drive was corrupt, find it at dos.retropc.se stick it on a disk, xfer, good to go, only the DrWho theme mod on disk, hey didn't I have some archived on disks I saw.... so unpacked 3 disks of mods... about 30 and sat there all last night going through them. Oddly, this speaker was a bit overkill for a modem too, quite nice quality for such a small sucker, I have definitely heard many worse case speakers. So I was quite enjoying the output knowing the limitations of PC squeaker sound reproduction.

BitWrangler wrote on 2024-10-29, 14:37:

I am somehow in a "speaker drought" meaning I thought I had dozens of little 8 ohm speakers around, the type to hook up to speaker headers, or for general low-fi noise. They all seem to have disappeared, or I have at some point, put them "somewhere safe" effectively disappearing them.

There's a reliable seller on ebay from which I've purchased hardware many times that has a number of small speakers for sale, let me know if you're interested and I can drop you a link. I have no affiliation with them other than having bought from them before.

Swapped a couple more damaged RAM sockets on a pair of Sound Blaster AWE32 cards (CT3900 and CT3980).

Had a bit of a panic when the CT3980 didn't seem to be recognizing the installed RAM. Spent about 30 minutes checking continuity, reflowing solder joints, etc. Until I finally realized I had forgot to set the jumper to enable external RAM. 😅

Before:

After:

Upgraded my KG7-RAID's main drive to a 74GB Raptor on a J-Micron SATA-IDE adapter that I did a resistor mod on.

This thing officialy FLIES. 🤣