0003 Why did manufacturers make slightly non-standard PC parts?

[@kpanic23]

Hi Adrian! When the battery in the Dallas timekeeper module runs flat, it doesn't retain _any_ settings, not even over just a soft reboot. When you "save" your settings, they go right to the dump. Those DS1287 modules don't have any external power pins to power them off the PSU while the machine is running (that's also why you have to hack into them to connect an external battery). I do actually have a couple of those AT-style motherboards with double PS/2 connectors, I believe there was some de-facto standard for them. There were AT cases where you can break out the hole for the second PS/2 port, but unfortunately I don't have one. There also were ATX I/O shields with holes for the two PS/2 ports, so you could mount those boards into an ATX case. Sometimes they also have a pin header next to the PS/2 ports to connect a standard DIN socket for an AT keyboard.

[@senilyDeluxe]

I have some computers with Dallas RTCs where the settings survive a reboot, even a reset, but if the power goes out for just a fraction of a second - boom, data's gone. And I have some computers with Dallas RTCs that behave the same - settings won't even survive a warm restart.

[@steeviebops]

+1 When the DS1287 ran flat in my SLT/286, it would give a POST error on every boot, even if you loaded defaults and reset.

[@fnjesusfreak]

A PS/2 30/286 will hold data on a reboot even with a flat Dallas chip.

[@abooogeek]

Seems you are lucky, every mainboard I got hands on with a Dallas RTC clock will not allow you to do any change in the settings and will render the mainboard as a brick if you remove it. The only way I had was to feed them a replacement. I understood then why at Computer Reset there was this pile of ASUS P/I-P55TVP4 left untouched. Soldered Dallas/Odin RTC battery clock and no way you can change the settings.

[@xicorpt]

I also think both ideas would solve Adrian´s problems, and maybe the "Glitch Works GW-12887-1" could be the ideal solution.

[@jotarokujo3395]

I think these should be called "Micro Mail Calls". Kinda fits in with Micro being smaller than Mini, and Micros also being kinda relevant.

[@DeLorean58]

I came to suggest the same thing.

[@jabbawok944]

Me 3

[@marvintpandroid2213]

Me 4

[@korgied]

And my axe

[@DanPellegrino486]

Big yes

[@mulad]

I may have missed it, but I'd recommend hitting F9 when in the BIOS on that suspect motherboard to load default settings. It may not help, considering that the battery in the CMOS+RTC chip has gone bad and may have damaged things, but it's probably better than just going through the menus and flipping settings back and forth. There tend to be some hidden values that you can't modify from the screens.

[@thorsteinj]

Definitely worth trying a reset, just imagine all the values potentially corrupt behind the scenes… also, flash the bios if possible could solve things.

[@detalite]

15:00 It will work in ATX case. There were even special cover plate for those motherboards. I a few years ago i converted one motherboard to two PS/2 conectors and used that plate.

[@jandjrandr]

Pretty sure those motherboards in that era were built for ATX with the back cover plate or PS/2 like clone cases if my memory serves me right. That was the era of uncertainty when PS/2 keyboards and mice were getting popular and had adaptors for AT to PS/2 and serial/USB to PS/2 to accommodate both until PS/2 died. Those connectors are just notoriously difficult to insert.

[@Zeem4]

I think that those AT boards with PS/2 ports were actually a standard, but one that didn't really take off. I've seen quite a few with exactly that layout from several manufacturers, and generic AT cases with a removable piece on the back to allow the PS/2 ports to be accessed.

[@erichkohl9317]

Hey Adrian! I'm glad the parts got to you in one piece. Sorry about those motherboard issues; I really had no idea what to expect. Perhaps in the future if you can iron out those problems it can still be salvaged. And that floppy drive -- it actually came out of a Gateway 2000 that I recently saved from my employer (they were going to have it recycled). After the bezel broke off I just went ahead and replaced it with a Gotek, which is why I included it in the box. And the AWE32 came with the Micron system -- that's why it fit into that one slot you put it in. But yeah, the fact that it's such a long card probably makes it difficult with certain motherboards and/or cases.

[@Inject0r]

It definitely makes this an interesting video! No need to apologize for that! 😉 As for the Dallas: it just needs a new battery on there. It’s a 30 minute job with a dremel and soldering iron. After that, Adrian should load the default BIOS settings + time and save the settings. After that he should be starting to set the BIOS settings. The AWE32 could still fit in the first slot of the Intel Advanced/EV board. When I was younger, my brother gave me his “old” AWE32, and I plugged it into that exact board. I just needed to put some tape on the card and give the PCB a slight bend, but it worked flawlessly for about 5 years! Both boards would be great contestants for a testing board. The Micron board maybe a bit more, due to the fact that I believe the Intel Advanced/EV deserves a full P1 build! 🙄😁

[@erichkohl9317]

@@Inject0r Yeah, if I had known that we'd all be lusting after our old Pentium systems that we had in the '90s I would have saved the case from the Micron ... instead I gutted it and just saved the parts. Regardless, I got lucky recently, because I found an eBay listing for a Micron Millennia that was very similar to what I had, so I bought it. Windows 95 FTW! ;-)

[@Inject0r]

@@erichkohl9317 love it!

[@johnnystauersbl1620]

it is a completely normal error on scop probes. and can be repaired. remove the molded part with an eg knife so you have the raw metal clip, put a piece of shrink flex on the wire, solder wire on again, shrink over wire, don. we always have a roll of silicone cable to replace with when they get too short 😅

[@VintageTechFan]

I think in this case its just that the molded part is way too hard and the bend protection is therefore purely cosmetical.

[@bobblum5973]

I had similar problems with the clip on Tektronix scope probes back in the late '70s/early '80s, and did the same fix with heat shrink tubing after trimming and resoldering. Those weren't originally molded with rigid plastic, they were more like pliant vinyl; if yours are more rigid that may be contributing to their short lifespan.

[@cyxceven]

Yeah, why doesn't he just fix it?

[@TrimeshSZ]

@@bobblum5973 I think everyone had that problem - when I read your post something deep in my brain went "196-3120-01", presumably because it was a part I ordered so many times.

[@SeanBZA]

You need to get a bus extender for those ISA cards, simply a card slot and then a small card edge connector, so you can plug it into the motherboard, and get the card high enough to clear the CPU.

[@amirpourghoureiyan1637]

A right angle adapter would help as well

[@shawnmulligan3471]

@@amirpourghoureiyan1637 or even a 2 or 3 port riser card for from PC would probably work, if it's just for testing cards; although in the case of the Sound Blaster it kind of looks like if you put it in the absolute farthest slot and just bent it a little it would hit the heatsink but not really interfere with it, and it would probably work ok

[@spokehedz]

I would extract the bent-wire-clip portion from the rubber molded connection, and then solder on my own ground wire. If I was feeling particularly fancy, I might consider putting some heat shrink over it, but I might just leave it as a bare wire so I can just use a IC leg-clip or maybe a regular old croc clip on whatever was handy.

[@AndyHullMcPenguin]

I also agree with this. These clips 'aint rocket science. The silicone cable may be a little strange, but shortening it by a few mm isn't going to affect your accuracy in any measurable way. Snip off the connector just beyond the break. Tin the end of the wire, remove the sleeve from the clip. Solder the wire back on, add heatshrink to taste. Rinse, repeat with the other broken ones.

[@jabbawok944]

Sugru is good for replacing failed strain relief.

[@SidneyCritic]

It could be where the wire attaches to the clip, but I wonder if it's the wire hardening and cracking/breaking. Either way, I would just fix it.

[@danielcogburn4355]

Yeah I'd just cut off that rubber molding with a razor blade (x-acto or similar) and resolder it back together.

[@josephkartychak6789]

In regards to the Micron “AT” motherboard, it was more common than you would think to have ps2 ports side by side. I’ve come across a lot of AT cases that had the holes set aside on the back in order to utilize both ports. What you’ll usually see is a larger AT keyboard hole with a smaller ps2 hole next to it, actually overlapping. You could also throw it in an ATX case that has an I/O panel specifically set up for it.

[@ironhead2008]

Yeah, it's not that weird. I'll wager the higher ups at Micron wanted their board to be "proprietary" but didn't want to spend any money so the engineers did the bare minimum and the suits said "eh, good enough". Putting a PS/2 KB port on the board reinforces this theory: It basically means they wouldn't have to ship an AT to PS/2 adaptor with their systems.

[@nickwallette6201]

Gateway used those on some Micronics-OEM’d boards. I have the 486DX2 full tower with a PS/2 KB port. There’s a footprint for a PS/2 mouse port, but simply adding the port did not enable mouse functionality. More work to do apparently. :-)

[@szr8]

@@nickwallette6201 Maybe there was an option in the BIOS to enable it?

[@markshade8398]

Yep. Those cases were very common once the overall industry adopted the PS2 port standard.

[@TheRealColBosch]

I don't think I've ever seen Adrian so angry before.

[@8o86]

even his hands are angry

[@wearwolf2500]

I think he's been getting more and more negative in videos lately and honestly I find it very off putting. I'd much rather watch someone get excited about stuff than complain about stuff.

[@redace001]

Perhaps he needs some gummies, might be low sugar / hangry. ;P

[@StevenJPiper]

@@wearwolf2500 Strange I don't notice this. I think you're hypersensitive

[@senilyDeluxe]

"What were they thinking?" isn't that AVGN's tagline?

[@TheHempage]

That Epson drive is useful for Amiga or Atari ST computers, you can change it from DS1 to DS0 by moving a jumper on the back. That is the drive I am currently using with my Atari ST

[@86smoke]

9:30 - Adrian, there are some AT cases that have this kind of ps/2 connectors on the back (I'm pretty sure that Fujitsu did those as well as Biostar). I have 3 of those in my collection, so they are not uncommon. And this ATX Linkworld power supply is rubbish, throw it away as it presents danger to parts connected to it. Also, for the CF-IDE adapter to work properly, try disabling 32bit mode; those early pentium mobos used to drive me nuts with its sloppy implepantation of all standards possible. Also I know some of those boards does not like IDE adapters at all.

[@shadowflash705]

Yeah... Linkworld killed a motherboard out of the blue once.

[@SuperVstech]

OMG! Adrian! You made the classic pc blunder… NEVER plug in a ps2, or AT keyboard connector with the power on! I have replaced SO MANY motherboard fuses from customers doing this…

[@Treveliian]

IRQ/DMA for the CDRom ports on the soundcard are conflicting with motherboards onboard IDE would be my guess.. either that, or that soundblaster card might required -5v

[@CATech1138]

i vaguely remember an early proprietary sound blaster cdrom back in the 2x speed days

[@bryanpratt3933]

He didn't disable the onboard IDE controller, they were conflicting with the XT-IDE. Watching these videos is frustrating.

[@kellyherald1390]

I was yelling at the screen saying to remove the sound blaster card! That card and motherboard was from the days of Plug-n-Pray.

[@Treveliian]

@@kellyherald1390 Yeah i like adrian but sometimes im just like... newb lol but you put a mac classic infront of me and i'll make similar mistakes.. no one is an expert for all systems.. we pick a handfull to be really good at, and the rest just enough to get by.

[@FastLoad]

Maybe a conflict on IDE with the SoundBlaster card, troubleshooting 101: Test one card at a time.

[@shawnmulligan3471]

That Sound Blaster doesn't appear to have IDE - it has Sony / Panasonic / Mitsumi interfaces, but no IDE port. However, for the XT-IDE I would suspect maybe a conflict with the onboard IDE, which the BIOS did allow you to disable. For the XT-CF (or the XT-IDE if it's using the onboard IDE BIOS), there might be a problem with the CF card being 4GB; the BIOS may have a 500MB or 2.1GB limit. Other troubleshooting options would be trying an ISA IDE controller without a BIOS (I'm sure he has plenty of those), a PCI IDE controller, or even a SCSI controller (which wouldn't be subject to the drive size limitations in the BIOS).

[@emolatur]

Around 22:33 ... it's probably worth digging deeper into the bios on that machine. Around that era I remember lots of weird (in the sense that they made me ask why they couldn't decide automatically) settings related to bus speeds and timing, it may simply be that the bus itself is running at a speed that the xt-ide card isn't happy with.

[@wojiaobill]

it's nice to see we're still using these computers in the year 7199

[@uomoartificiale]

I think we just witnessed the peak of a canadian rage mode... :D

[@HighTreason610]

I think for a time the double PS/2 ports were almost a standard to themself, usually on workstation boards. That said, they seem to be be offset one way or another on some boards. Occasionally, the board maker left vias for a regular DIN port there, so you could desolder the PS/2 ports and just stuff a DIN port on, but not all of them have it and I can't quite see if this one does. Wouldn't be in too much of a hurry to modify it if it did, given it's a Micronics board and while they were usually well built, they sure do seem to be quite picky about what configurations they'll work in and also seem to always end up being a little slow for some reason. Plus the Intel/Phoenix BIOS they liked using was pretty bad - the way yours behaved is actually par for the course in my experience. It'd most likely work if you replaced the RTC and went our of your way to put it in the one configuration it'd like, even more so if you disabled PNP. Your time would be better spent with a more generic board, though.

[@eshwayri]

Adrian, that motherboard likely came out right when we were switching over from AT to ATX. The early ATX cases (Antec was quite popular) were still able to take AT motherboards, and that motherboard may have included an ATX back-plate; the two ports would certainly be in the visible area of the ATX back-plate. On current cases you would need one that supports EATX due to the length. I have that pleasure with my Tyan Tomcat III and IV.

[@stonent]

I want to say I saw some cases from that era that had a small punch out that partially intersected with the AT keyboard connector to allow for the double PS/2 port setup like that.

[@geezerdiamond]

I had the SyQuest EZFlyer230 as a secondary hard drive on my Amiga 1200 back in the day. It was a superb drive. I’ve re-bought one recently for my current 1200 for a bit of nostalgia 😃

[@UpLateGeek]

I just went back and checked, and I didn't see you disable the onboard IDE controller when you tested the XT-IDE card. That would probably cause a conflict, which would explain why the CF card wasn't recognised. As for the NVRAM being corrupted and not saving the settings between soft resets, it's possible that it's not wired to power the NVRAM from the board at all, and it's relying on the built-in battery to power it. EDIT: I've also got a board that has two PS/2 connectors in that same location. It originally came with an AT connector and the two footprints for the PS/2 connectors, so I desoldered the AT connector and soldered in two PS/2 connectors. I didn't have an AT case, so my idea was that they'd both be accessible when mounted in an ATX case, since it's got the larger opening. I never saw the underneath of the board in the video, so it's possible that board has an AT connector footprint in the correct location under the PS/2 connector.

[@claymccauley]

I'm not familiar with that specific model of the AWE32, but if it has an IDE port, you may need to disable it with a jumper. If I recall correctly, I had to do this on mine for some systems.

[@Bubu567]

Strain relief is so important. I would like to see it dissected so we can figure out what exactly they did wrong.

[@williamsquires3070]

I think Adrian needs a new shirt that says, “It Freaking doesn’t work!” 😏

[@jk180]

I don't know, the ranting is kind of fun. It's refreshing to know that we all get frustrated over the same stuff

[@lauram5905]

That 20 year old probe you switched to will probably last a lot longer (knock on wood) just from still using leaded solder (hopefully) If you take the other commenters’ suggestions to resolder it, try looking into a soft solder with a high lead content or bismuth added into the alloy, you may find it handles a lot more strain cycles and doesn’t break apart as quickly. Maybe even use a thicker gauge of stranded wire to connect the ground clip to prevent the wire from shearing as well. The combination of costcutting and ROHS has caused the death of many headphones, USB cables, and other high strain cycled cables.

[@JohnWohlers]

Check the sound blaster, the on board cdrom controller has an ide mode. If it's set to the same io / irq space you could be having conflicts. Remove the io addreds jumpers and the irq jumpers fully to completely disable the sb board controller.

[@AceStrife]

Definitely should get an ISA riser board or extender cables for situations where you can't fit certain cards into slots on a motherboard. Makes things so much easier. Just hope they're not manufactured like the recent NZXT riser cable (12v short to ground).. shoddy companies pushing out cheaply manufactured parts is unfortunately common, especially from China where the rest of the world's standards and regulations mean nothing.

[@kellyherald1390]

I was working at Micron Computer at that time that motherboard was used. I also remember that there was a version of that board with the AT keyboard port. And when you flashed the BIOS you had to be sure you had the correct version of the flash. If you had the PS2 board and you flashed it with the AT flash you bricked the flash chip. Ah, those were the days. I definitely remember the M54Hi boards. Brings back memories. I also remember when they would first power up the systems and occasionally a vendor had installed a tantalum capacitor backwards and you'd hear a loud bang and if you were the unlucky one to witness it you'd also see a small fireball shoot through the air.

[@iceowl]

if it came from 1995, i seem to remember that cases of the time had a space for those PS/2 ports. but mostly, stuff like that is made to make it very inconvenient to put it in an OEM case. i learned that a bunch of times, only a couple of years later in 1997, when i was tasked with cleaning out the former projector room in my high school's lecture theatre. lots of old proprietary stuff that makes you ask "why", and the answer is "they don't want you to make frankenputers out of their product. they don't even really want you to be able to use their product. they want you to throw it in the trash."

[@toxicpsion]

i still have a syquest EZ-135 IDE, and still works (most of the time). HD platter cartridges are more robust than i would've expected.

[@ropersonline]

Adrian's definition of "super quick": Best part of an hour.

[@capitanschetttino8745]

Hello Adrian! I know its been quite a long time, but those old soundblasters took up a lot of resources (irq interrupts and dma channels) and it would not be surprising if you have one in conflict, especially with the video card or the IDE controller. Greetings from Argentina!!

[@MikeStavola]

Pretty sure the Rigol ground lead is like, a standard ground lead. It looks just like the Pomona ones I use on my systems

[@DxDeksor]

As others have said, some boards don't remember settings even after a warm reboot with a dallas chip. That'd be the first thing I'd try to fix on my list before trying to understand the other problems this board has, because it's possible that the other issues are related to the corrupted nvram.

[@chrishartley1210]

I remember working on an HP pc, the motherboard had failed. At first glance it appeared to be a standard board, made by MSI if I remember correctly. The IO shield was fixed but a little drilling and cutting would sort that, I thought. Except that all the components were about 1mm closer than standard meaning that the entire backplate was shorter by about 10mm. There was absolutely no way a standard motherboard would fit into that gap. Clearly MSI had made a special board for HP. Crazy.

[@DLim454]

you can clip or plug directly your ground connection onto the scope (usually it has one) which is ground for all the probes so you dont need to use those ground attachement for each probe. 😊

[@soniclab-cnc]

Oh the dreaded faulty scope probe. I’ve had some experience with this. lol. Always charming when you finally you realize the reason you’ve been getting strange signal and scratching you’re head for 30 minutes...

[@Rorschach1024]

When I worked with Walt Russell, who was employee #7 at Compaqit was explained to me that Compaq made ISA bus cards just a bit taller than the actual IBM standard in order to prevent the use of compaq boards in standard clones.

[@bluhnd]

For the non working cards on that Micron motherboard, could this be due to IRQ conflicts?

[@AshtonCoolman]

Why do manufactures still do it today?! I'm looking at you Dell and HP!

[@awilliams1701]

gamer's nexus fan? lol

[@awilliams1701]

@@tspawn35 so does HP. HP uses all 3. However ironically Gamer's nexus found that Dell had an exceptionally good, but useless due to it's proprietary nature, power supply. It was like 80+ gold based on their testing.

[@AshtonCoolman]

@@awilliams1701 I'm a Gamers Nexus fan and a 22 year veteran of IT. Their PCs are destined to be ewaste just like Steve said 😂

[@awilliams1701]

@@AshtonCoolman yeah I used to do IT too. I was hoping it changed after I left IT for programming 10 years ago. I was sad to see that it hadn't.

[@altareos]

A mini mini computer was a microcomputer, so this show should be a Micro Mail Call!

[@jsdutky]

You can buy replacement (no-name) ground leads that will work with most scope probes, $12 (or less) for a pack of four from Amazon.

[@Hiraghm]

I recently saw a Gateway P5-75 that had twin PS/2 ports like that motherboard; the keyboard port had been modified by Gateway so you could access both PS/2 ports.

[@rtechlab6254]

I've had older boards reject CF cards before. I have a really weird vme one going up on my channel in a few weeks that led me a merry dance. Those probe leads are a standard part, RS certainly sell then. Same type on my Tek, Hantek and ye olde Scopex

[@thechillhacker]

Its an insulated alligator clip, a wire, an open spade connector, and some heat shrink... go a little higher on the gauge, nice multiwire stranded, 12 or 14 gauge gel coated speaker wire is one of my favorites for stuff like this.

[@fragglet]

It's when the time gets to 99 o'clock that you know you've stayed up waaaaay too late playing with old computer hardware...

[@silmarian]

I worked tech support for these things and, having had to help people install replacement motherboards, I was glad there weren't more connectors or headers than there already were.

[@andrewb9830]

For the motherboard with the odd port location. You could probably drop it in an ATX case and order a blank I/O shield and drill the proper holes for the ports.

[@simontay4851]

The mounting holes are in different locations on ATX boards. The screw holes on an AT board wont line up with the mounting locations in an ATX case.

[@EvilTurkeySlices]

@@simontay4851 I have my Baby AT board in an ATX case.

[@pjm780]

o/~ In the year 7199 Adrian's old computer is only kinda-sorta still alive o/~

[@argoneum]

Cut the insulation and check if the ground lead didn't act as a fuse at some point. Happens when some live circuit is connected by someone forgetting that oscilloscopes aren't isolated from the mains ground. Yes, I did that too, also got some used probes from Swiss meteo station with burnt ground. Common thing, easy to fix 😸

[@alextrusty2585]

On this IDE problem - perhaps, it's ISA bus divider/waitstates related... check it out in BIOS

[@xaenon]

1994 Bios? You're likely running into the BIOS limit of the era, which for most machines was 504 or 512 or 528 megabytes (depending). Nobody ever thought we'd need hard disks larger than that. When drives started showing up with larger capacities (630 mb, 700ish mb, 850 mb, 1.0, 1.2 and 1.6 gigabyte drives all started showing up at about the same time, and the prices started dropping rapidly), you had three options for the machines the older limited BIOSes: You could update or replace the original BIOS, you could install an add-in card with its own BIOS and IDE controller, or you could run drive-overlay software. Most machines would just fail to detect the 'oversize' drive, but ones that would freeze at that detect phase were not unheard of.

[@andrasszabo7386]

I had the exact same problem with my Rigol scope. I have dismantled the crocodile clips and soldered a new thicker wire into it and on the other end I have crimped an eyelet type connector and opened it with a needle file.

My original AWE32 had one (or two) SIMM slot(s).

[@Roxor128]

Put a couple of 8-meg SIMMs in there and you could play any mod file with hardware mixing at full quality. Most are under a meg, but going by my own collection, they can go up to nearly the ten-meg mark. Cubic Player has to sacrifice some of the sample quality to play larger files on a GUS by reducing the bit-depth and sample-rate to fit the data into GUS RAM, and given the AWE32 only came with 512KB by default, it'd have to do the same there, too.

[@dintyshideaway9505]

That is a strange MIcron Motherboard, but since I worked for Gateway 2000 at that time, there are a couple of explainations. At that time PC makers would contract with corporate and government entities and would be required to continue to use identical commonly replaced for a set period, usually three years. There were no standards, so the components would vary from contract to contract. Remember that custom configuration would require FCC testing, which was expensive. Since at Gateway we were using the 440BX (? I recall could be wrong) Intel chipset we also used PS/2 ports as they substantially reduced tech support calls for non-functioning mice. We used a standard clock with a CRxxx battery, and a lower watt 145watt ATX power supply. The 133 was a late second generation Pentium, so I have never seen a large OEM make one using an AT Power supply, Dallas RTC, or or 32 pin memory. Companies who had no issues with Tech Support, would not pay the royalty to use PS/2.

[@BadManiac]

The dallas chip contains timer / clock generator as well, some boards can't work without the clock. It will very likely work if you replace or mod that Dallas chip.

[@spudd86]

Actually IIRC AWE32 stuff is totally different from WaveBlaster Daughterboards, in fact I'm pretty sure some AWE cards had the pins for a waveblaster board so you could have *two* wavetable synths on your one card. The AWE wavetable synth doesn't show up on the MPU-401 interface, it's its own thing entirely, only works with games/drivers that support it directly. This is largely because of the RAM soundbanks and the fact that the AWE32 supported up to 28 megs of wavetable RAM. Loading that over MPU-401 would be a bit slow. But yeah, basically an AWE card is an SB16 with extra stuff. Though it's really an entire extra device, it's only really one card because it integrates into the SB mixer.

[@_irdc]

One thing I’d suggest regarding the motherboard is to test the ISA bus frequency. These Pentium motherboards often had programmable clock generator chips, allowing for setting various clock frequencies via the BIOS. Corrupting the settings might have set it to something the VGA card could handle but the IDE-buses (noting here that IDE is pretty much ISA-on-a-cable with no active controller in between) couldn’t. The Dallas problems might be solvable by connecting an external battery to the chip (some variants allowed this).

[@mortismaker1520]

Hi Adrian, I may h ave missed it, but did you pull the Dallas RTC? I've seen them do some odd stuff when they die. Also if you pull it and reset BIOS to factory; that might help too. Best...

[@agevenisse3252]

The Pentium board wasn't just Year 2000 compliant. It was Year 7000 compliant as well. Many late AT boards were available with either a AT keyboard connector, or two PS/2 connectors btw. You can switch them out.

[@KernArc]

I've had an Epson floppy drive like this one in my very first PC, c.a. 1994. Boy, was that thing LOUD while seeking! One day I couldn't take in anymore, took it out of my PC and with me to my school. I've asked my CS teacher whether I could swap out that thing for a drive from one of the school's lab PCs and, to my surprise, she was fine with that. So I went home with well used but quiet Mitsumi drive and left my Epson one in the school PC. It was working fine otherwise and it's not like a loud floppy would make much difference in a school lab full of adolescent boys so no harm done. I'm not sure I remember correctly anymore but I seem to recall spending the following evening defragging a bunch of floppies, listening to the soothing sounds of well-behaved head seek stepper motor... and if not, that would seem like the right thing to do ;-)

[@thechillhacker]

~27:20 That's a little bit of a fluke Me: points to the 8000A multimeter in the stack of test equipment on the bench. No, THAT'S a little bit of a Fluke!

[@jammi__]

You should find the reset CMOS jumper on the thing. The Dallas chip may have some battery juice corruption going on internally as well.

[@retro-futuristicengineer]

I have a similar weird board from Fujitsu, but not from a Fujitsu/Fujitsu Siemens PC, but cramped obviously as an upgrade into a Mini Tower that looks more appropriate for 386/486 instead of Pentium. I think it even has a header for PS/2 Mouse, so you can use this to connect the mouse if your case does not have the dual PS/2 exposed. I have already seen dual PS/2 AT cases in some videos, but not yet in real-life. The keyboard port lines up pretty well with the AT DIN connector. And yes, the behaviour is, to a certain extent, normal. I have multiple boards with Dallas that do not save any settings (configure BIOS, save and exit, all settings lost). What I have never seen was 99 o'clock. And by the way, Maxim, the new owner of Dallas, still sells DS12887A+, but these have the "Pin 21 Mutation". They have Pin 21 exposed which the non-plus versions did not had. Pin 21 is CMOS Reset (Negative Active). The clock will be stored properly und run fine, but the settings will be resetted if this pin is grounded (which it is on some boards). But with an empty clock - all gone even after a soft-reboot, which also applies to save and reboot in the BIOS.

[@gigigigiontis8]

About the motherboard that does not save settings: it still might be the Dallas chip. I've seen a video on This Does Not Compute where a similar thing happened (board not keeping settings even with mains power applied)

[@VoVilliaCorp]

I feel the same way about Proprietary stuff, I couldn't tell you how many dells I have worked on over the years and I shout "why, why do this" all the time lol

[@mal2ksc]

Suggestion: If you get some replacement grounding leads, and they still suck, try putting some heat shrink tubing over the end and strain relief that keeps breaking. This won't stop it, but it should slow it down by removing some of the mechanical stress.

[@brianhginc.2140]

For the probe GND clip, exacto-blade cut the heat-shrink tubing on the 'Ohm' shaped spring and re-solder the broken wire with a new heat-shrink tubing in place. The heat-shrink tubing doesn't need to be perfect or even there at all. There is nothing in the location where you attach it to the probe other than a solid outer GND cylinder core. You wont short or damage the probe if your craftsmanship isn't perfect.

[@NovaSilisko]

Same as it ever was... I got a Lenovo Thinkcentre from 2011 or so for $17 from a friend, the insides of it are so disgustingly proprietary. The USB connections have some weirdo 12 pin connector, I had to do a lot of digging to get the front panel switch pinout, and a bunch of other stuff I'm forgetting. About the only standard things were the ATX power connector, SATA, and PCIe. I had to gut a bunch of it to really be able to use the case with anything else, although I did beat it into submission eventually. But I also have a sack of parts that are basically useless outside of that case.

[@TheYuppiejr]

Did you pull the heatsink and validate that the processor is actually a Pentium 133 mhz model? Strange behavior with an unintentional overclock could cause those sorts of issues if the previous owner set the jumpers incorrectly for the CPU parameters or voltage values. I was dealing with an old 486 that was acting strangely and it turned out to be a DX4 model installed in a 5v only board. Also, I'd probably remove all but one stick of memory and test it out with Gold Memory tester or something similar just to make sure you starting from a known good point.

[@Troppa17]

Yeah, there should not be an Pentium 133 on an Socket 5 especially on some wired Micronics Computers Board. If I remember correctly, you can use a Socket 5 CPU in an Socket 7 Mainboard but not the other way around. So I wonder if the PCB under the actually CPU - that provides the CPU fan with power - will also provide the CPU with the needed additional supply current and clock signals... In this case I would think there might be a compabilty problem between CPU with that thing and mainboard. Might be clocking issues, might be not enough power or ripple due drawing more power than the mainboard can handle or something along these lines.

[@molten_software]

I think there is a header on those boards for a breakout PS/2 port, so you can use the board in a regular AT case, and then a PS/2 port elsewhere since one gets blocked

[@MD4564]

9:50 I use to modify my back planes to fit the motherboards Also manufactures always skimp on the cables and such where they can cheap out. And finally in the day of PC repairs, I have always noticed there is no standard designs back in those days.

[@RetroDepot]

Concerning the HDD. The XT-IDE has a BIOS ROM. But you may need to set the cylinder/sectors in the bios for CF card. I often have to play around with these on later AT boards.

[@VintageProjectDE]

Adrian, the ground clips for the Rigol scopes should be available as cheap spare parts from Rigol or their distributor. As an alternative, you could try to purchase the ones from Testec (product no. 20011 - 11 cm; 20010 - 22 cm). Here in Germany, they're sold for about 1.50 EUR per piece, and repairing them really isn't worth the effort. You could also try to get in touch with them. Maybe they'll send you a couple sets of probes. ;)

[@eformance]

Is there a footprint for a 5 pin DIN on that motherboard? It looks like it was designed to populate an AT keyboard connector or 2 PS/2 connectors.

[@thepirategamerboy12]

That EZ Flyer drive looks really similar to a Sparq drive I have. Both are by SyQuest.

[@50shadesofbeige88]

9:48 I have a similar ATX-ish motherboard. I put mine on a piece of acrylic and drilled holes for standoffs. So the "case" is more like an open air bench. This way it's at least usable, and if I want to put it into some kind of wrapped acrylic enclosure later, I can.

[@steelhorseman6883]

I don't know if you know it or not, but once DOS is up and running you can out the disk and the computer will still run as long as the computer is left on. You have to have a boot disk for everything.

[@winstonsmith478]

I'd carefully cut into the probe ground clip plastic with a hobby knife/razor, solder the exposed conductor to the clip, and cover the area with heat shrink tubing.

[@markae0]

27:48 I think you should put the date on the ADB sticker , because with the passage of time things may go bad. As another example, on a cars oil filter , one should also put the date of installation.

[@NJRoadfan]

The motherboard is out of a Gateway 2000 machine. They did the whole dual PS/2 port thing and that is likely a Micronics made board. Some generic baby-AT cases actually had the punchout for the mouse port! The "ATX" power supply is out of a Dell machine and it likely does NOT use a standard ATX pinout. Some early ATX Dell motherboards (made by Intel) used a weird power supply pin out for some really stupid reason and moved all the +3.3V pins to the extra plug you noticed.

[@Dukefazon]

I just ran into a slightly non-standard PC part this summer, Tweeted at LGR Clint and he didn't know what the hell is going on. You know the end plate of a card, right? Let's call it face plate. One end gets held to the chassis with a screw and the other end is slightly narrower that goes into a little notch in the PC case. Well, I have a PC case that looks like a normal AT case, I think it could house a full-sized AT motherboard, it has the proper power switch, turbo, reset and the speed display, but the motherboard mounting holes are a bit weird. But that's not a problem, I have plastic standoffs, the problem is that I just can't put in any cards, because they hit the bottom of the case (that part is raised but no holes) so I can't screw the cards in place. And I don't know what to do with this one, I don't know if it accepts a different type of motherboard with different cards or what is going on and it was not a cheap PC case, cleaned everything and right when I went to put in the stuff and try the different parts if they work together, I got stuck on the very first step. I don't even know if my motherboard, CPU, RAM and cache memory is good because I got so bummed out by this I didn't even test things further...

[@EPtechser]

Concerning the mobo, I had a similar issue with an Aopen pentium mobo with an AMI bios. I did a bios update and that made the SD to IDE adapter work. I see your mobo has an Intel bios, but hey, it is worth a shot.

[@AshenTechDotCom]

replace the dallas when they fail...yeah...it can cause alot of weird issues on those micron boards from my exp, they tend to be rock solid though, also, swap the ram, sometimes those boards dont like mixxed ram you can also just find the power contacts on the socket and try applying correct voltage to test, i will say, when the battery fails those things can act like this and other weird symptoms. some are just.. perplexing... one had a failing battery back then...but only showed issues when specific apps ran... eventually we tried swapping the dallas with an extra we had...and...it started working flawlessly... the shop i worked for sent the failed one off because they wanted it back, it had failed far too early... the sent 2 dozen replacements in exchange for the failed one... not sure how one of the owners had contacts at the maker, but they wanted it back and paid to get it back... im guessing they wanted to break it open and see why it failed so early... not sure if they are still around but, there were replacements a company made that had a battery socked on an adapter, they sold 2 versions on for a 2032 the other for one of those AA battery connectors that also worked with block lithium batteries.

[@markae0]

28:37 "anyone has thoughts" my thoughts are 1) Turn down every clock speed to its minimum ( to see if it fixes freeze) 2) Remove old RAM with minimum (known working)RAM test again. 3) I agree, during a warm reboot of the BIOS settings, it should remember new setting if you correctly saved it. 4) Is it running the correct BIOS version? Does it match the BIOS chip. Can you download the BIOS and examine it in HEX?

[@PaCav-et3kt]

for the grond lead have a look on digikey using the term "oscilloscope probe ground lead"

[@Drucklufttroete]

Could it be possible that there is a DIN5 footprint left of (and half under) the PS/2 mouse connector on the Micron board? Maybe they made two variants of it - one for nonstandard OEM cases and one for generic ones. If there is one, you might be able to replace the connectors and mount the board in a normal case.

[@okfj]

Mini + Mini = Micro, so a Micro Mail Call (MMC). Where Micro can also relate to Microcomputer. For those that don't catch the reference, from Wikipedia "A microcomputer is a small, relatively inexpensive computer with a microprocessor as its central processing unit." So a Micro Mail call dealing with computer parts all goes together.

[@michaelmichalski4588]

In the 7199 if Adrian still has time, and if he's still online, Old computers will all be repaired and Adrian will have no hair

[@AshenTechDotCom]

that pci-isa board worked in alot of cases they started selling cases with oblong holes for the kb/mouse ports like that when ps/2 started to pickup wide use.

[@AshenTechDotCom]

all those old micron bords tend to be solid as a rock even seen some left running for decades and wake up when the mouse was touched...(family moved the grandparents to a retirement place and the power was never cut off... the family ended up having a feud and the parents got to move back home since they had come to hate living in AZ wanted back to WA... their grandson actually was my friend, we helped them get moved back home grandad had left the system running in the basement, its running windows NT 4, hes STILL GOT IT, it even had a UPS that had kept it up whenever the power may have gone out all those years... we know this due to the uptime on the system... blew us away... granda had us build him a top of the line computer and... hes in is 80's now, and playing MMO's like a madman... he couldnt even have a computer at the retirement place from what he said,they banned them and laptops because a small group of the community found them infuriating...even seeing others use them or knowing others had them caused them upset... dude is also stupid into heredity and geneology, got his wife a nice laptop and tablet, she needed the laptop to make proper use of her fancy sewing machine that she couldnt make full use of due to not being able to have a computer hooked up... they didnt like the heat or lack of seasons, the old guy even said the arthritis wasnt as bad as when big storms were coming in in FL.. he would KNOW it was going to be bad, by how bad he hurt due to the pressure changing...

[@ianhanschen]

Some AT cases have a tiny ps2 knockout right next to the keyboard connector hole (really, as part of the edge of the 5 pin din hole) - I remember seeing that in the 90's. But it was never standardized. I don't think the computer store I worked at ever sold one of these boards.

[@hailmynova]

I know makes no sense but "welcome to the basements black list" just sounds corny but awesome at the same time lol enjoying this channel more and more.

[@Backwoods-Bob]

Company I worked at had a laptop that used a DIN connector for the power supply, same as the keyboard mini-DIN keyboard connection. Guess what happened when the ps was plugged into the wrong port?

[@vladimus9749]

I'm surprised no one's called out the weird amount of ram. Looks like possible ram corruption to me.

[@petertorda5487]

SB AWE 32, normally all of them have sim slots for memory upgrade (28MB max), this looks like some more budget version. Anyway wavetable part is based on EMU8000 synthesizer chip from EMU system (I guess similar was used in Proteus series of EMU rack modules). EMU system, was really something in 80's and on beginning of 90's, especially with Emulator IIx and EMAX I samplers (used by everybody, Depeche Mode, Vangelis, etc...). But AWE 32 is not that case, as they used worst possible General Midi samples, and compressed in 1MB rom. Theoretically chip with professional capabilities, but I'm not sure if EMU released any interesting sound banks in soundfont format, most likely they keep it for EMAX. As you mentioned Waveblaster, I guess that first original was based on similar synth chip but with MC68k, and have got much better ROM with samples from Protheus series, but without any RAM, and also without any effects, so it sounded extremely dry.

[@blackIce504]

I still have an LS120 drive and its IDE, takes normal floppy disks as well and even supported the 2.88mb disks, I think the autogame card was auto as there was no need for jumpers for ISA address settings.