During the last few days, I only worked on an MB S.7, a PCCHIPS M550 Intel TX chipset, I solved the initial problem of the BIOS, through a recovery floppy, then I started doing various bench with the i166MMX CPU, which was at reduced frequency (133 MHz), then at normal frequency (166), then due to the multi blocked, I changed it with an i233MMX, and at the moment I am finalizing the benches at 233 MHz, later I will do more at 200, reducing the multi to 3X.

DarthSun wrote on 2024-10-08, 19:10:

Quadro FX3450 (6800GS) Maintenance, pasteer. The best 1 slot cooling stem. GPU-n copper, RAM and black bent metal surfaces, dire […]
Show full quote

Quadro FX3450 (6800GS) Maintenance, pasteer.
The best 1 slot cooling stem. GPU-n copper, RAM and black bent metal surfaces, direct contact with the RAM, do not need a pad, just paste. MX4 went to them.
The 6800GT also has such cooling.
While working:

The attachment 1QFX3450Open.jpg is no longer available

The GPU:

The attachment 1QFX3450GPU.jpg is no longer available

Nvidia really milked the 6000 and 7000 series pretty hard, there are just so many variants of the 6800 like I have a Quadro FX1400 which I'm pretty sure is a PCI-E 6800, just a notch down from your one since there's no power input?
There were just lots of them back in the day and now. I mean I also have a working 6800GT AGP but I'm afraid to use it since I've also got a dead 6800GT AGP.

Today's thing is I have so many laptop computers that I'm now stuck in LCD hell - I've just sort of resolved the DSTN polariser issue on my DSTN laptops which had bad polarisers: Re: Vinegar syndrome and choosing new polarising film
So all that messing around with sticky tape had a good outcome.

But I also just got my Toshiba Satellite P20 fixed up the other day making one good one out of 2 bad-ish ones. Then it was stored on its side and fell over. It's not a small laptop at 17" and now there's a big white pressure line in the LCD where that knock has damaged the backlight sheets.
Otherwise it's a very cool laptop, not really made by Toshiba but it's got a 3.2GHz Pentium 4 and a Geforce FX 5700 with 64MB of 128-bit memory. A pretty nice all-in-one retro games machine now that sbemu is a thing.

And I found out my Sony Vaio C1XD Picturebook laptop's polariser is starting to break down:

That's pretty much the same as the one in this video! https://www.youtube.com/watch?v=zbrW7uUYpkw - they must have some particular susceptibility to the polariser breaking down like this.
I'm gonna have to get quite good at polariser replacement now....

Rip them all off and use them wearing polaroid shades 🤣

BitWrangler wrote on 2024-10-11, 02:02:

Rip them all off and use them wearing polaroid shades 🤣

haha, that is definitely *a* solution! 😁
I did kinda try that with this DSTN that had no polariser for a long time, it just made me never want to use it because the laptop was in one place and the glasses were in another.

That doesn't fix the fresnel sheets in the Toshiba P20 either, I am so mad about that... but at least I know now how those pressure spots happen in LCD screens - the back of the LCD takes a sharp knock and the lightguide and fresnel sheets press against either other at the nearest sharp point in the LCD housing, deforming probably the fresnel layers since they're the most delicate and bam, now you have a bright/dark spot on the screen forever. And it's all you'll ever see when you look at the laptop screen.

My Nokia has almost all new capacitors now, I forgot to get few but they're not critical. I accidentally ordered oversized film capacitors for few places where e-cap doesn't really make sense and I had to be slightly creative but overall I don't have to go back into the monitor for a good long time now.

Originals actually were all 105ºC rated japanese capacitors but a number of them had degraded quite a bit, few are in the areas where the PCB has actually browned from heat... Now those bits have some extra long life types in them, with a bit of slack to allow some air to pass underneath.

It is also amazing how much higher performance modern capacitors can have, while being substantially smaller. Compared to originals, some of the replacement e-cap can take 5x as much ripple current while being half the size, and the polymer ones are another class on their own 🤣.

EDIT: it still lives !

Thermalwrong wrote on 2024-10-11, 02:20:

haha, that is definitely *a* solution! :D I did kinda try that with this DSTN that had no polariser for a long time, it just mad […]
Show full quote

BitWrangler wrote on 2024-10-11, 02:02:

Rip them all off and use them wearing polaroid shades 🤣

haha, that is definitely *a* solution! 😁
I did kinda try that with this DSTN that had no polariser for a long time, it just made me never want to use it because the laptop was in one place and the glasses were in another.

That doesn't fix the fresnel sheets in the Toshiba P20 either, I am so mad about that... but at least I know now how those pressure spots happen in LCD screens - the back of the LCD takes a sharp knock and the lightguide and fresnel sheets press against either other at the nearest sharp point in the LCD housing, deforming probably the fresnel layers since they're the most delicate and bam, now you have a bright/dark spot on the screen forever. And it's all you'll ever see when you look at the laptop screen.

Theoretically, one could maybe exploit memory effect in plastics to get them to pull back into shape with careful heating, but temp would have to be super controlled.

I installed a new to me Sound Blaster AWE64 in my Katmai system today . I haven't installed an ISA card in at least 25 years. It had a good, firm snap to it when the edge connector mated. Surprisingly satisfying.

You can easily call me a mad man - resoldered 2x SGRAM chips on a Innovision Banshee by hand. Yes, you read that right, by hand.

It works in Windows (it would present me a whole jailbar picture before)... off to test 3dMark 99 Max after I install a fan on the heatsink.

PcBytes wrote on 2024-10-14, 23:23:

You can easily call me a mad man - resoldered 2x SGRAM chips on a Innovision Banshee by hand. Yes, you read that right, by hand.

It works in Windows (it would present me a whole jailbar picture before)... off to test 3dMark 99 Max after I install a fan on the heatsink.

Patient 😀 That's really nice that resoldering / replacing the RAM resolved it. My problems are all BGA it seems like. So far my PCI banshee came back from the dead with a hot air reflow but I've got an AGP one that still needs re-balling.
Was it bad ram or bad solder joints on yours?
btw I've soldered 4x SGRAM chips on my MS-6168 by hand mostly, it was not fun. Quite a few solder bridges and dry joints

Today I got this green Octek EVGA-16 card back to working. I bought these two cards a few years ago as a pair for parts, at the time the yellow card had broken display from bad RAM while the Octek card had no display because it had no RAMDAC and it seems I didn't know it was a RAMDAC back then. Back then I took all the good RAM off the green card and put it on the yellow card. The Octek card has then sat in a parts bin for maybe 4 years with no RAM and I nicked the VGA connector from it at some point.

So I got some RAM and RAMDACs for cheap from Aliexpress since I don't have DIP RAM or RAMDAC and the card now works 😀

Thermalwrong wrote on 2024-10-14, 23:51:

Patient :) That's really nice that resoldering / replacing the RAM resolved it. My problems are all BGA it seems like. So far my […]
Show full quote

PcBytes wrote on 2024-10-14, 23:23:

You can easily call me a mad man - resoldered 2x SGRAM chips on a Innovision Banshee by hand. Yes, you read that right, by hand.

It works in Windows (it would present me a whole jailbar picture before)... off to test 3dMark 99 Max after I install a fan on the heatsink.

Patient 😀 That's really nice that resoldering / replacing the RAM resolved it. My problems are all BGA it seems like. So far my PCI banshee came back from the dead with a hot air reflow but I've got an AGP one that still needs re-balling.
Was it bad ram or bad solder joints on yours?
btw I've soldered 4x SGRAM chips on my MS-6168 by hand mostly, it was not fun. Quite a few solder bridges and dry joints

Today I got this green Octek EVGA-16 card back to working. I bought these two cards a few years ago as a pair for parts, at the time the yellow card had broken display from bad RAM while the Octek card had no display because it had no RAMDAC and it seems I didn't know it was a RAMDAC back then. Back then I took all the good RAM off the green card and put it on the yellow card. The Octek card has then sat in a parts bin for maybe 4 years with no RAM and I nicked the VGA connector from it at some point.

The attachment double-chips (Custom).JPG is no longer available

So I got some RAM and RAMDACs for cheap from Aliexpress since I don't have DIP RAM or RAMDAC and the card now works 😀

Joints that went bad after decades. In my case I didn't have access to a reflow machine, so I had to make do with my trusty ole' soldering iron (thermally controlled) to do the job. All that's left to do on it is now getting a new VGA header in there as the one it has right now is having a missing screw hook.

Finally got back to the EGA monitor after car trouble, wiring trouble and being diverted by diverse diverting diversions... Got the power socket soldered, got it fixed in place, got basic functionality...

So gave me the oldie heartsy stoppensies when it wouldn't not do nothing at first but then I scrubbed the plug in IPA and off we went. That XT is an Apco Turbo, apparently made by the same ppl that made PB's VX88 for them, Goldstar or Samsung I think, so it's pretty much identical to a VX88. Guess I gotta dig out a keyboard for it to go much further... also a boot disk... Not quite sure what's going on with the HDD and controller, it's telling me a Seagate ST138R isn't working... yeah well it isn't when it is, it's a Kalok.... not sure if that's "badge engineered" too, Seagate with a Kalok label on it. I don't have my necrocomnicon of XT drive controller incantions to hand, so not sure what magick it needs at the moment.

Welp, fixed that Banshee's artefacts for good. It still had some faint dark red lines show up during a few 3dMark99 tests (the spinning table of squares, forgot its exact name, but it's prior to the spinning light test) so I said screw it and redid the legs solder joints on all 4 chips east of the Banshee BGA. Clean picture, everything works as intended.

Found a working VCR left by the previous tenants in an apartment recently purchased for renovation by some friends. There were also three CRT TVs, of which I assume some may have been working, but I have nothing to do with one, nor could I find reasonable means to transport them.

The VCR, though, is nice. Came with a working remote too, which I had to clean and remove the corroded batteries from. I have been years without one, and it is not that I need one right now, but there may still be some old shows on VHS cassettes at my parents' place, surprisingly not all of which are easy to find online. And every once in a while one does find tapes in book scrap/donation bins. I had a working setup of capturing them to PC back in the day, when I was transferring my childhood VHS collection, and still have all the equipment, so maybe I'll get around to doing it again one day.

dr_st wrote on 2024-10-15, 05:31:

Found a working VCR left by the previous tenants in an apartment recently purchased for renovation by some friends. There were also three CRT TVs, of which I assume some may have been working, but I have nothing to do with one, nor could I find reasonable means to transport them.

The VCR, though, is nice. Came with a working remote too, which I had to clean and remove the corroded batteries from. I have been years without one, and it is not that I need one right now, but there may still be some old shows on VHS cassettes at my parents' place, surprisingly not all of which are easy to find online. And every once in a while one does find tapes in book scrap/donation bins. I had a working setup of capturing them to PC back in the day, when I was transferring my childhood VHS collection, and still have all the equipment, so maybe I'll get around to doing it again one day.

a nice find ! Although it may not get used much there are, as you say, the odd occasions when you can. And indeed VHS releases are not always duplicated in DVD or online, some really are slowly fading away into obscurity

Spent half an hour googling trying to answer the eternal question "Are Model Ms really XT compatible or not?" and got, no, yes and maybe in equal amounts. Hmmm, though I didn't see any from the 90s, especially after Lexmark spinoff, that anyone had confirmed to work. So out of what would take less than 10 hours to find, I grabbed the earliest one that I had handy, and that was a 1989 1391401. So this one, plugged into this particular late Turbo XT seems to work so far... I say seems and so far, because it stopped complaining about KB at POST and esc and ctrl-alt-del seem to work, but I have not got a boot yet for other reasons. Status: "Where the hell did I put my 720k boot disks??? I had them in the last 12 months I am sure of it...."

Edit: Got back to it some hours later, still hadn't found the box of 720s with the boot and utils, discovered an LS120 won't rawwrite on 720s, derp, found a USB flop, wrote out a 3.3 image and a 5.0 image, try de boot... nada. Realise I haven't got any lights or grind on the 720... oh.. is it B: drive or an X:drive, as in pining for the fjords. Guess I gotta rip it open again tomorrow and go over all the connections and jiggle switches for oxidation disturbance. I swear this was set up with 3.5" A drive last time I used it. Ugh don't know where in the stash the easiest 360k boot disk is.... probably got spare 720 drives though. Another key on the keyboard seems to work, discovered hitting enter triggers a retry on the HDD boot.

Finally added some active cooling to my ELSA Victory II-A16 (Banshee), which would have originally shipped with a passive heatsink. Donor card from which the HSF came, is a Mac Rage 128P that I couldn't even find a suitable BIOS for, making it useless.

BitWrangler wrote on 2024-10-15, 17:26:

Spent half an hour googling trying to answer the eternal question "Are Model Ms really XT compatible or not?" and got, no, yes and maybe in equal amounts. Hmmm, though I didn't see any from the 90s, especially after Lexmark spinoff, that anyone had confirmed to work. So out of what would take less than 10 hours to find, I grabbed the earliest one that I had handy, and that was a 1989 1391401. So this one, plugged into this particular late Turbo XT seems to work so far... I say seems and so far, because it stopped complaining about KB at POST and esc and ctrl-alt-del seem to work, but I have not got a boot yet for other reasons. Status: "Where the hell did I put my 720k boot disks??? I had them in the last 12 months I am sure of it...."

Edit: Got back to it some hours later, still hadn't found the box of 720s with the boot and utils, discovered an LS120 won't rawwrite on 720s, derp, found a USB flop, wrote out a 3.3 image and a 5.0 image, try de boot... nada. Realise I haven't got any lights or grind on the 720... oh.. is it B: drive or an X:drive, as in pining for the fjords. Guess I gotta rip it open again tomorrow and go over all the connections and jiggle switches for oxidation disturbance. I swear this was set up with 3.5" A drive last time I used it. Ugh don't know where in the stash the easiest 360k boot disk is.... probably got spare 720 drives though. Another key on the keyboard seems to work, discovered hitting enter triggers a retry on the HDD boot.

Well covid must have zapped that neuron that thought it was set up for 3.5" A drive, because the wiring says different, also went over the drive setup jumpers again and yup, correctly configured. I don't think B drive boot fallover was common until 90s. Anywayyyy, all the wrong ends to swap the cable round simply, so went looking for the handiest 5.25 disks... and ended up chasing phantoms again I guess. Had a picture in my head of a lone box of verbatim 5.25s in the corner of a bankers box, and I don't have too many of those, but they seemed devoid of 5.25s... annoying as well as when I came across them not long ago, it was like "heh, better remember these are here, near to hand, in case any of the family sproglets want to know why a floppy is a floppy.." ... meanwhile main 5.25 stash is buried and extremely well bundled up, and at risk of being a 20 business day diversion because of all the stuff I find first.

Ah, cable did have right ends, they were tucked and folded out of casual sight, really tight behind the drive cage and PSU. So the 3.5 is A now... not that it got me toooo much further, not sure whether I've got a crappy drive, crappy disks, crappy image, or crappy drive I wrote them on because getting a lot of A R F I on the dos 5 disk and 3.3 won't even try, anyway, get dumped at A prompt in dos 5 with nothing but command com loaded and not sure if something hinky about that. HDD is sorta working, can dir it and load stuff off it.. can't run fdisk at the moment though, found I already had a dir with some utils in, ran Snooper, and it seems to hang on first screen while analysing HDD, not sure if just super slow 'coz it's clusters are tangled or something. Leaving it for a spell to see if it comes out of the coma, drive might still be accessing.

Anyway, keyboard seems to be working okay, Snooper reports only 84 key support there though, arrow keys seem to be mapped though. That's all I can do at the moment... gotta prep some more disks with diff drive on decent stock I guess to see what is what. Monitor is looking fine, CGA on EGA or officially EGA mode 1 scanrates according to IBM. Blur is either motion or the focus on cam was still hunting. Snooper seems not to be able to tell it's a V40 either, close enough I guess.

edit: btw, hdd is Kalok Octagon K330 with same CHS as Seagate 138R that be why it think it is.

Picked up a Dell GX300 of the local Facebook Market. Has a single Pentium 3 933 cpu, 512mb Rdram and an ATI Rage 128 Pro. It has a dual slot 1 board in it so I'm shopping for upgrades.

BitWrangler wrote on 2024-10-17, 01:53:

Ah, cable did have right ends, they were tucked and folded out of casual sight, really tight behind the drive cage and PSU. So the 3.5 is A now... not that it got me toooo much further, not sure whether I've got a crappy drive, crappy disks, crappy image, or crappy drive I wrote them on because getting a lot of A R F I on the dos 5 disk and 3.3 won't even try, anyway, get dumped at A prompt in dos 5 with nothing but command com loaded and not sure if something hinky about that. HDD is sorta working, can dir it and load stuff off it.. can't run fdisk at the moment though, found I already had a dir with some utils in, ran Snooper, and it seems to hang on first screen while analysing HDD, not sure if just super slow 'coz it's clusters are tangled or something. Leaving it for a spell to see if it comes out of the coma, drive might still be accessing.

Anyway, keyboard seems to be working okay, Snooper reports only 84 key support there though, arrow keys seem to be mapped though. That's all I can do at the moment... gotta prep some more disks with diff drive on decent stock I guess to see what is what. Monitor is looking fine, CGA on EGA or officially EGA mode 1 scanrates according to IBM. Blur is either motion or the focus on cam was still hunting. Snooper seems not to be able to tell it's a V40 either, close enough I guess.

edit: btw, hdd is Kalok Octagon K330 with same CHS as Seagate 138R that be why it think it is.

286 / Snooper :

Extra fast the HDD, 40GB Flat Maxtor :