Bah, my comment did post. To be honest, I sorta expected that, but I did get about 200+ subscribers from that video alone, and since then, I've gotten a notable bump of views. Honestly though that period was not a good time in my life due to some personal family stuff, then followed by moving which why there was a long delay until I got *anything* out, and then a second hitatus. I'm still struggling to make content on a regular basis, but my goal is once a month at a minimium. Anything else is bonus.

Part 2 looks like it will be up tomorrow or Thursday, but I can say this one had over a decade of service before retirement which really says something.

Skip ahead to Part 2 where I do just that. :) I'm working on Part 3, and learning that the SCSI controller on this system actually has quite a bit problems because of the EISA bus. In short. very little works properly with it, and as an added-insult-to-injury, the onboard NIC doesn't even have a Windows NT driver.

Oooph, if you still have the disks or tapes, I'll be happy to archive them. 'd love to cover some 8-bit micros, but the prices are hard to justify since I'm pretty well out of my wheel house with anything that isn't an Apple II (and the last time I used an Apple II was the early 90s).

@@NCommander I had the tapes and disc. I was keeping everything I bought one of each operating system since BASIC. A couple years ago I was going thru a phase and threw out everything up to Windows ME. Sorry I do not have those any more. I do have a Compaq laptop that came with ME on which I loaded WIN 98. I can send that for free.

EISA was Gang of Nine as a response to IBM but it's very very similar to MCA which basically had the same problem with reference disks :/ Apple did have NuBus but that autoconfiguration was very expensive since everything had to fit in a nubus world (PCI worls similarly). EISA is relatively cheap to implement.

@@NCommander You are correct, I'm embarrassed. I had a 486 EISA MB that had dual CPU sockets. It wasn't until a version of NT 4.0 Workstation (I think) that windows supported anything more than a single processor. And when I finally added a 2nd one, it didn't make much difference.

I contacted you by the Facebook :)

That was probably a Profession/GX, the closely related but slightly less cursed cousin of this machine, but great to hear another Intel computer user :)

Yes, it was the GX Professional. Externally looked identical and had a 486DX processor. There weren’t (m)any PCs with such slim “pizza box” cases at the time, so I remember being impressed by its looks and it had decent performance, too, after installing some extra RAM. Eventually, I traded it in for a second-hand NeXTcube.

there's a label on the bottom with the Intel badge and more, but yeah. It's very very generic otherwise.

Intel XPress platform & Intel XTended XPress?

oh wow, talk about people I didn't expect drop by this corner of the Internet. Your videos was one of the main factors to get into YouTubing retrotech, and actually document trying to get this 486 monster going. I actually picked up a C128 due to some of the Commodore content you made. That PSU pinout is immensely useful, and makes the idea of swapping the PSU for a more modern ATX (w/ -12v converter) more practical although I'd still have to trace out RAWST, and the other pins. I'll probably work up the nerve to do a recap as a prelude to getting my 1702 recapped since that rather desperately needs it.

@@NCommander Thanks! Yeah one of your videos popped up on my feed so I started watching -- and your stuff is really good!! You have real perseverance in figuring out issues, and I love that. For this amazing Intel machine, you should put LP486 in the description as (I think?) that's the model number - and when I search google for it, your videos don't show up.

@@adriansdigitalbasement I actually had LP486 and LP486E in the tags, but I just added it to the description directly, thanks!

It's an optical illusion, but I do still want to re-cap this PSU at some point. Dismantling it to its base components is easier said than done though ...

It's not, just a standard ISA one but its not a soundblaster compatible one either :)

That is (kinda) plan B if I smoke the mainboard, but I haven't had great luck with SCSI historically :(. The drive pack itself self-terminated, and hooked on ID 0, and I don't have an easy way to change it, and I find with old SCSI, you spend more time debugging it than I would spend trying to get this system up and running :( The drive itself does look to be standard 66 pin SCSI with Molex 5-pin so finding something it can plug into shouldn't be impossible. I'm hoping to have the drive dumped before the Dallas chips arrive (at the moment, DigiKey still says they're processing my order, so I might have them before 2022 at this rate), but it's pretty dependent on what ports actually work with the Dallas chip dead. I did post a screen shot of how I actually plan to dump the filesystem even w/o working Ethernet.

@@NCommander That Mac SCSI drive is a Quantum ProDrive LT, 700mb dated 1994. Apparently self-terminated, don't recall its ID# but probably 0. Card is an AHA-2930CU, bit of an oddball (2930 have limited device support) but was what I found listed as specific to this HD. The dead Mac had onboard SCSI so wasn't helpful. Anyway, I just plugged it in and it worked. I'm not a SCSI maven but I don't think I've ever seen a SCSI HD that wasn't standard and couldn't be recognized by one or another of the Adaptec cards. Then again, mine are random acquisitions and I haven't gone looking. :)

@@NCommander More thinks... it has an external SCSI port; normally there's an onboard SCSI utility you can access during boot.. you should be able to hook up another drive ... actually, I'd be interested in seeing if that works!

​@@Reziac SCSI HDDs are pretty standardized all around (it's better than IDE in this regard TBH), but it's more an issue that adapters can be very tempermental in general. Mac users have is somewhat easier because Apple basically was standardized from start to finish in this era. As for the external SCSI, I don't know if its the same bus per say. I haven't got anything to hook up to it, and due to the dead Dallas chip, it really isn't happy on EISA stuff so :/

I discuss this in part 2, but it's actually not that trivial; this drive is a SCSI-2 connector, and there's no easy way to connect that to a modern PC. No one makes a SCSI-2 PCI-E card, most PCi->PCI-E adapters only work with half-high cards, and on top that, I need an enclosure, terminator, and the usual luck that comes with SCSI chaining. I ended up with a fairly creative solution to data recovery. Part 2 should drop this Wednesday.

@@NCommander I am tooned for the part 2 :)

@@NCommander Almost universally, SCSI is SCSI -- with one main caveat: Watch out for HVD, or High-Voltage Differential cards. They're relatively rare, thankfully, but they don't mix with SE (Single-Ended) or LVD (Low-Voltage Differential) stuff. You can damage something if you try. Other than that.... You can find an adapter to change just about any SCSI _this_ to any SCSI _that._ I think people think it's a lot more complicated than it actually is. I know you've already come up with workarounds, but this is my advice to anyone having to deal with this: Don't let it scare you away. It's really not that bad. The biggest thing is IDs. Every device needs a unique ID #. The SCSI card itself is usually #7. Normally, you will see hard drives as 0 and 1. Some cards want to boot from SCSI device #0. Some rare others prefer the highest number after the card. Narrow SCSI only uses IDs 0-7. Wide SCSI uses 0-15. Internal SCSI drives usually have a 50-pin ribbon connector. SCSI interfaces with internal connectors will usually also be 50-pin ribbon connectors. It's exactly like IDE, just wider. You need a straight-through ribbon cable - easy to find on eBay if you don't have one. You can also buy ribbon cable and 50-pin IDC connectors and crimp them in a vice (or buy a $20 tool.) It's a piece of cake. Later Wide SCSI devices may have a 68-pin connector with high-density pinout. You can get a simple passive adapter to convert between that and 50-pin (narrow) SCSI. Wide SCSI lets you run a wider data bus, so twice the throughput, and use SCSI IDs 8-15. Wide devices can fall back to narrow, so you can mix them if everything is ID'd less than 8. You might see an 80-pin "SCA" connector on later internal SCSI drives. This was mainly intended for use in servers with hot-swap bays. You can adapt to this, too. The external connector on a SCSI card may be a 25-pin D-Sub (particularly older stuff made for Macs), or a Centronics port (like the printer end of a parallel printer cable), or a 50-pin high-density connector, or a 68-pin high-density connector (if it's Wide SCSI). These are all compatible, and can be adapted with the right cable or passive adapter. There's nothing special about internal and external SCSI. I have a couple of ISA brackets with an HDB-50 connector on the outside, and a 50-pin IDC connector on the inside. Great for using sound cards with built-in SCSI interfaces to connect external drives. In most cases, both the internal and external connectors are the same bus. If it's a dual-bus card, it will usually say so. You do need to terminate both ends of the SCSI bus, and depending on how old the devices are, you might have to even do it on purpose. Almost ALL SCSI devices have a TERM jumper that enables termination. Enable this on the _last_ device in the chain at each end. The card will either auto-terminate (pretty common, and should be defeatable with a jumper or software option, so you'll know if it's doing it), or you might have to enable termination on the card end by installing a resistor pack or closing a jumper. Usually, this will be done for you, and you have undo it if you need to. If you have both internal AND external devices, you need to terminate the last external device and the last internal device -- the ones furthest on the end of the chain in either direction. Do not terminate the card in this case. If you use short cables and slower SCSI, you can often get away with %@$#ing up the termination. But don't push it. There are two ways to terminate an end of the bus: Either a passive terminator (essentially a resistor array), or an active terminator. In the former case, it's usually a jumper or switch on the device. Some external devices require a physical terminator to be installed. It's just a plug with no cable. If you use active terminators, you need to supply something called "TERM POWER" to the bus -- essentially just +5V. You only want ONE device on the bus doing this, and ideally that should be the SCSI card. Everything else should have TERM POWER disabled. Some devices sense the presence of TERM POWER and enable it automatically. Old devices may blow an onboard fuse (or a polyfuse) if power is present from more than one source. In some cases, the fuse might be your SCSI cable or a trace on the connector. :-) But that should be rare. Enable parity anywhere you can. If you can't enable it everywhere, turn it off. OK, that might seem like a lot, but it's not bad. To sum-up: 1) Anything can convert to anything else, usually with a simple connector dongle. Except HVD -- avoid that. 2) Narrow SCSI 1 can link up with Ultra 320 SCSI just fine, and vice versa. The connectors (and low IDs) are your blockers. Solve for that and you're golden. 3) Terminate the furthest two devices. That could be a drive and a card, or two drives with the card in between. 4) If you're using active terminators, enable termination power on the card and disable it everywhere else. 5) Parity -- all or nothing. 6) SCSI ID #7 is your SCSI card. Most other IDs are inconsequential. Stick to 0-7 if you can.

I'll have to try that next time (and get a spare toothbrush to basically do that). Thanks

@@NCommander Get a very soft one. Stiff brushes can damage stuff. -- I keep one in my kit thanks to clients' filthy hardware!

@@Reziac yeah, I saw factory bodge wires and I went "NOPE!" on any idea of dosing this board or washing machine but a very soft one should be OK.

@@NCommander Don't worry about it too much. I regularly bathe whole motherboards and even PSUs (PCBs only, not the whole assembled box) in the sink, use some dish soap to break down grime if it's more than just a layer of surface dust. Iso alcohol works really well for solder flux cleanup. I use any random toothbrush that I got from a dentist visit or wherever else.. You don't need to grind it in there, just a casual brushing will do. Go slow and easy around the bodge wires, but they shouldn't just take any opportunity to fall off. At worst, you might break down the adhesive where they tack parts of it with glue. But the ends should be soldered and won't let go unless you snag it and break the wire. In general, this stuff's tougher than you might think. I have had zero failures that can be attributed to washing, scrubbing, or otherwise cleaning something. I've, for sure, broken stuff that would've otherwise worked. (If you haven't broken _something,_ you haven't tried hard enough. It's just a hazard of the hobby.) But not from cleaning. You gotta do something really dumb. And you will. I have. :-)

How much does the DS1358 have? the DS12887 I replaced the dead one w/ has more memory. Still, doing the filing and clipping a new battery is always an option.

@@NCommander 4K x8 SRAM. Its used to store the EISA (or MCA if you have a PS/2) configuration data. Likely a cost saving part as boards with the DS1287 have that bit of RAM on the board itself.

I noticed this problem when KZfaq updated their captioning tool. It's very frustating to me since I can't fix it on my older videos I just switch editors to DaVinci Resolve which has some easier way to do in video and regular subtitles so I'll see what works, but I really need is the ability for KZfaq to let me move the CCs up about half an inch.

@@NCommander i like kdenlive for video editing personally

@@DouglasWalrath I'm trying DaVinci Resolve because there's a LOT in kdenlive that frustates me, but everything up thus far has been.

@@NCommander well did u use it on linux or on windows? cuz the windows version is still in beta and has plenty of bugs, try the flatpak linux version

@@DouglasWalrath I run Linux, but it's a problem with the workflow. I've been using Resolve to edit my next video and the workflow is considerably easier and I'm not fighting the tool anywhere as much.

It's a 487 slot, but since this is a DX processor, it's empty. There are multiprocessing 486s (486 was the first SMP capable chip), but they're rare. There's a jumper on the board for DX/SX operation which I don't think I've shown (you can see it on the jumper diagram on part 2). There's also an open spot for a cache IC of some sort (the reference manual calls it a "TurboCache")

You can always tell a 487 socket. It's got one extra pin in the inside corner. Kind of odd here, though, since it looks like the 486 is also in a socket. Why you would fit an SX in there, and later upgrade it with a 487 rather than just replace it with a DX (which is what a 487 is....) is beyond me.

@@nickwallette6201 I have heard that there were both dedicated 487s, and 487s which were DXs which failed in some way that didn't effect the FPU. I'm sorta tempted to do some math benchmarking now by swapping the processors with SX/DX models. THis sytem did have SX as an option and there's a SX/DX jumper.

@@NCommander That's not entirely accurate, the original 8086 was designed to be multi CPU compatible and I believe there are some NON PC systems that did this. However on the PC side of things the first MP system was the Compaq Deskpro from 1989, which supported 2 33MHz 386 CPUs.

@@Dargaard I think it was not the DeskPro but the Compaq SystemPro Server System which was capable of dual CPU´s. It could also use 386 and 486 Processors. ( The Multiprocessing was asymmetric )

I actually managed to locate the power pinout some time after I filmed this so it's no longer necessary, but this is roughly what I'd do as a last resort. I generally understand enough about mains voltage to know what is safe and not safe to do, and frankly, given what I've seen some other people do on youtube, I'm honestly amazed no one has electrocuted themselves yet

It's an optical illusion. I actually meant to put a text note saying that its not bulged in real life, and it didn't happen >.

@@NCommander Right, I saw it later in the vid from a different angle and it looked fine. Weird perspective stuff happens on YT particularly on close in work. An artifact of using zoom or macro on the capture device and then whatever processing YT do after upload. Plus I'm just a little jaded, having dealt with the electrolytic capacitor plague as a hardware guy in the field when that was a thing.... I sleep and I see a bulged cap and the smell hits me. Electronics PTSD. Keep up the good work here.

@@chillybrit2334 My camera is pretty crud (I actually used my cell phone for a lot of the shots because it actually does better than the GoPro), but the first footage of the PSU is zoomed in quite badly because I had the camera in the wrong position and didn't realize it >.< Still getting used to filming live action stuff.

@@NCommander I think a lot of YTers use phones for this kind of work rather than GoPro because of the lenses - see bigclivedotcom - another electronics reverse engineer for instance. It's all good. Iterate to perfection. No-one is born with any knowledge, yet you already share yours with us - this is more than enough. These are your early days on YT and leaps will bound over humps. You don't have to have a 100 strong team of video editors behind you right now. You need something that grabs people and keeps them coming back, the rest follows. :) Regards from SN BTW o/

@@chillybrit2334 what's really awkward is the lightening in my apartment is kinda a potato. I was able to fudge around it in this video, but the next one, it's going to be pretty notable that I'm working in an iffy scenario. Thank you for your kind words. Honestly, the reception my writeups I got on SN has discouraged me from posting there so I've moved to my own blog.

SCSI on this system is actually connected to EISA and not ISA and it seems to have a sane default hardcoded so ... that's actualyl entirely plausible. I need to see if the graphics chip is supported.

Yes, this system runs NEXTSTEP nicely, with drivers for every subsystem. I had one around 1995 running NEXTSTEP. It was a very nice PC for its time, although quite expensive.

... there was nothing broken with it ... I was checking to see if the caps were leaking, which is a common problem on these PSUs. THe only problem was the dead Dallas clock chip which I replaced.

Absolutely. I've had to work on some IBM PSUs with rivets holding the AC inlet (and outlet) in place. Drill the heads out, get some #4-40 screws (and nuts, if it's not threaded), and continue on.

I'll go more into details with specifications with the next video part, but here's the manual: www.elhvb.com/mobokive/Archive/Intel/professionalworkstationtps.pdf. I'm still somewhat learning how to do these types of videos and almost everything I've done up to this point has been pre-scripted and not "real life" things, I'm mostly working to expand that so this will def. be included with part 2. It's a Western Digital WD90C30 chip for reference. It's not exactly amazing, but it would get the job done for what this machine was designed for. Capable of 800x600x256 or 1024x768x16. I legitimately have no idea what the soundcard is, there are no identifiable markings on it and the EISA config utility doesn't even show the slot populated so it might just be dead. As the system is right now, it's a 486DX-33 Mhz produced in 1992, although there was other options available, and could be theoretically upgraded to an Overdrive-60 which actually would make running WIndows XP *technically* possible. I might even try it if I can source the OverDrive 60. There's 8 MiB of RAM installed, system max is 32 MiB. I'd be happy to answer any other questions and place those in Part II.

I live in New Jersey. The sun only works if you don't have a cloud of Newark blocking you :)

I'm not an expert, but I do understand the basics of how a PSU operates and how it can discharge. I did bleed this PSU off before opening it each time.