Think I did that once, years back when a 240/120 switch was on the wrong setting... but I think it can also happen if someone REALLY messes up the power switch wiring on an AT

BitWrangler wrote on 2024-05-07, 02:29:

Think I did that once, years back when a 240/120 switch was on the wrong setting... but I think it can also happen if someone REALLY messes up the power switch wiring on an AT

Thanks man, as it happens this PSU is 240vAC only, but I'll check out the wiring on the switch tomorrow. No biggie because I was planning on replacing with an ATX PSU + adapters anyway.

The seller on eBay claims it waas old stock that was found in the back room of a PC shop that went out of business years ago.

As with all PSU explosions, sure made me jump though 😁

can be one of those funky Rifa capacitors

rasz_pl wrote on 2024-05-06, 15:09:

shorting RTC is same as giving it half the cycle, so maybe its not oscillating?

I would like to thank you for this most valuable tip. One CUBX-E saved from the trash.

Replaced the little 32768 Hz oscillator and it's back alive!

I am actually playing with a real Apple IIe Platinum. Quite a nice machine. But also a bit troublesome to get going.

PD2JK wrote on 2024-05-06, 14:12:

The board won't power on when I short the 'power switch' header. Funny thing is, when I short the RTC crystal oscillator, it powers up!

I would never have thought to try this. I'll have to remember this trick for future debugging.

Today I received a Pentium 166MHz CPU that I bought for my Digital Venturis 575, Socket 5 PC. It replaces an IDT WinChip C6 200MHz CPU, which has poor FPU performance.

I almost purchased a 133MHz CPU, as I was pretty sure this motherboard was a Version 1.0 that only supported up to 133MHz. Good job I didn't, because upon examining the motherboard, I noticed the extra jumper that enables 166, 180 & 200 support, so in fact I had a Version 2.00 motherboard all along. For only slightly extra money I could have gotten a 200MHz Pentium and maxed it out, but nevermind, the 166 will suffice.

Definitely noticeably improved gaming performance over the WinChip CPU. I can now play Blood without it dropping loads of frames whenever there was any action on screen. 😎

PD2JK wrote on 2024-05-07, 07:39:

rasz_pl wrote on 2024-05-06, 15:09:

shorting RTC is same as giving it half the cycle, so maybe its not oscillating?

I would like to thank you for this most valuable tip. One CUBX-E saved from the trash.

Replaced the little 32768 Hz oscillator and it's back alive!DSC_8928.JPG

Now, that's a big win, seeing the asking prices for CUBX boards these days. Congrats!

Not entirely retro, but related.

Revisiting my Intel Legacy Files Archive section on my website... which eventually led to board transition BIOSes. I mostly stopped because of limited storage in general and logistics, which was solved last summer. This revisit started when I was thinking of changing the mainboard for my Live TV server, running into an Intel board I didn't initially recognize while browsing a certain site.

Since the 1990s Desktop board section is mostly complete there (just missing a few BIOSes for some of the boards), I focused on the 2010s Desktop board section. I was actually missing one board entirely (Intel DH61BF - the board in question), which I just added to the archive. Also found a bunch of BIOSes for some H61 boards, which some are needed as transition points for 3rd Gen CPUs and Windows 8 support... and yet there are more I haven't obtained yet. Thankfully, Intel still had their flowchart that listed 3rd Gen transition BIOS versions for them (yet got rid of the BIOS files themselves...). Edit: Oops, the flowchart was actually on the Internet Archive, as it, too, was long removed.

Also, seems some 2000s Desktop BIOSes (read: a lot) were found that I don't have yet; will have to add those sometime.

May have to create a secondary mirror at some point, but the current size of this is almost 60GB in size...

I have been testing lots of old cards lately, and I was just tinkering with a pile of old Paradise PVGA1A ISA 256K VGA cards. I was running through Checkit! tests on each card to be sure they were working properly, but one gave me a funky screen for one of the text modes. Basically, everything was huge and partially cut off, and text was very dark blue on black, when it was supposed to be a medium gray on black. I realized that the switches on the back of the card were set differently on this card so I set them to match the others and it fixed the issue. Long story short, with the help of this page I realized that... big surprise... the default settings (marked with a *) worked the best.

The one that seemed to cause problems was SW1/2 "VGA MODE SWITCHING CONFIGURATION ". Now, my question is... why is "PS/2 Style" the default and the least problematic mode? Why does the "PC\AT Style" setting cause issues with certain DOS text modes? I know that sometimes this third party documentation (most of it coming from TH99) can be incorrect, so I'm leaning toward it being a case of them just mislabeling the settings. Still, if anyone knows why this might be accurate I'd love to know. 😀

I don't think it matters too much, but I'm using a 440BX system with DOS 7.1 (98SE) and the display is a Dell Ultrasharp 2001FP.

By the way, I am absolutely LOVING the 2001FP for working with old video cards. It seems to handle almost everything I throw at it, and even when a certain mode is misaligned or doesn't look right the auto-alignment works amazingly well and even has a single button right on the face of the monitor to correct alignment instantly. Since it is also 4:3, I would consider this a nearly perfect LCD for testing retro gear. A CRT certainly blows it away in many aspects, but if I'm not going for a great gaming experience and just want reliable testing with lots of flexibility, lower power consumption and a small footprint, this is just great.

I'm sure the difference between a 2.8 512/533 P4 and a 3.06 512/533 CPU isn't much, but I wanted to max out the CPU in my P4 Win98 system. It seems that the 2.8 GHz CPU that I wanted to replace with the 3.06 is the highest speed CPU this very basic motherboard can use. I was hoping that the Ebay seller had sent me a bad CPU, but it POSTed in my P4P 800, so the CPU is fine. Bummer.

I have been playing around with MSI MS-6368 v5 and VIA C3 Ezra 800Mhz CPU as high end compact DOS/Win3 machine. Also using "industry standard" PicoGUS on top of VIA build in SBPro solution which worked without hitch.
Actually I'm impressed with VIA PLE133T compatibility in DOS. Build in Trident Blade iGPU handled DOS programs very well. I also run many DOS scene demos and in general it worked well. There are no native Win31 drivers but standard build in 8bit SVGA modes run well. CPU/Memory seems to be have enough bandwidth that 2d GUI acceleration is not that important.

Unfortunately MB does not have any FSB speed control so with setmul I managed to get it running at 486 level (lowest speedsys CPU mark - 18) so WC2 and test drive 3 are still way too fast.
Also some late 90s dos demos are asking Univbe support. I found Univbe v7 beta version which claims to have Trident Blade support but I could not get it running in pure dos - version I downloaded kept complaining that program is not properly installed.

On truly positive moment with VIA C3 CPU is low power requirement. I was able to run it on open bench completely passive. Even with full 800Mhz speed temps seems to hover low 40-es C.
Then again if you want to have full CPU power this MB supports Tualatin...

I dug out my unopened Xbox 360 E - bought on clearance when the Xbox One was coming out - and decided to set it up, so I could play some old console games. It's very slow and clunky due to Microsoft's glitchy software, but it seems to work alright otherwise.

Dan386DX wrote on 2024-05-07, 02:26:

So today I received a P60 system listed as not working on eBay.

As soon as I plugged the power lead into it, before I even had a chance to switch the system on, the PSU went bang and released a little magic smoke.

Anybody seen this before?

Probably the filter cap. They are hot all the time and will blow if the system is off.

I am currently setting up an all-in-one DOS machine. It is based on a µ-ATX SS7 board with SIS530 Chipset and ESS Solo-1 sound. I downgraded the CPU from a 500MHz AMD-K2 to an Intel Pentium 233 MMX to get full speed control. I also flashed an early mainboard bios to get the option to disable the mainboard cache, which is absolutely essential to reach the lower 386 speed range.

I have yet to make up my mind between the onboard SIS530 AGP graphics or my RIVA 128 PCI card.

The board offers up to 3 pci slots, so I can also install a Voodoo 1/2/SLI option on it. (But that isn't exactly the build purpose)

It also has a single ISA slot, which will cost a PCI slot if used. So using an ISA soundcard with build in or added wavetable is an option.

I am using an IDE2SATA converter to connect a 2.5 inch SSD in a swap bay. This is an easy option to transfer files and back up the system. A CD-Rom and Floppy will likely complete the build.

For me, the attraction is having access to spare parts, since it is based on µ-ATX.

Testing most of today's fleabay finds.

So far:

- FIC P4M-RS350/L works, one of the DDR slots has a few burned pins I need to somehow restore tho. Has a beefy Arctic cooler on it (which tbh looks a bit goofy if you ask me)
- EPoX EP-P55-VX works, needs cleaning as it's horribly filthy.
- HD3870 is pretty much a loss, the more I advanced with dismantling it the worse it looked
- GTX275 might be salvageable, it stops at 2B POST code on my ABIT IP35 so gotta check what's up with it.

UPDATES:
- 3870 is indeed a loss
- GTX275 acts fairly weird. Won't POST but won't beep either, and that's happening on my Gigabyte P35-DS3 as well... wth?
- EP-P55-VX cleaned up well
- K8N freebie got recapped, its Forza FX5500 also got recapped with red polys to match

Left to do:
- test DVD drives (Philips DVDRW + Plextor DVDROM)
- check 80GB HDD the P4M-RS350/L came with
- get a heatsink for the P166 on the EPoX
- test out the Promise controller similar to how PhilsComputerLab did it.

PcBytes wrote on 2024-05-12, 11:24:

Testing most of today's fleabay finds. […]
Show full quote

Testing most of today's fleabay finds.

So far:

- FIC P4M-RS350/L works, one of the DDR slots has a few burned pins I need to somehow restore tho. Has a beefy Arctic cooler on it (which tbh looks a bit goofy if you ask me)
- EPoX EP-P55-VX works, needs cleaning as it's horribly filthy.
- HD3870 is pretty much a loss, the more I advanced with dismantling it the worse it looked
- GTX275 might be salvageable, it stops at 2B POST code on my ABIT IP35 so gotta check what's up with it.

I'll update the post as I progress in testing stuff.

Really enjoy hearing about your flea market finds and what works/doesn’t. Please don’t stop telling us 😀

I found this floppy on Ebay. It's a registered copy of JASC Professional Capture Systems (also known as JasCapture). It's a DOS and Win 3.x program for screencaps.

Haven't found a copy anywhere online so I uploaded it to the archive. I tested it on my 386 machine and it works great. Doesn't ask for a serial. The DOS version is a TSR.

https://archive.org/details/img-2130-1

Managed to fix (somewhat) a burnt slot on my recently bought FIC P4M-RS350L.

What I did was carefully cut away the burnt plastic, just enough to leave the pins slightly exposed to make contact. Harder to explain as I don't have a photo of the slot after the repair on hand...

And yes, it indeed is the Medion-Disney Dream Desk PC version of the FIC mainboard.

Repo Man11 wrote on 2024-05-11, 16:35:

I'm sure the difference between a 2.8 512/533 P4 and a 3.06 512/533 CPU isn't much, but I wanted to max out the CPU in my P4 Win98 system. It seems that the 2.8 GHz CPU that I wanted to replace with the 3.06 is the highest speed CPU this very basic motherboard can use. I was hoping that the Ebay seller had sent me a bad CPU, but it POSTed in my P4P 800, so the CPU is fine. Bummer.

Well, the 3.06 has Hyperthreading, which the 2.8 doesn't. Might very well be why the motherboard doesn't like it?