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.
When I worked for a NeXT VAR/OEM in 1993, we had a few of those workstations at our booth at the NeXTworld Expo provided by NeXT. The (alleged) 'story' is that Jobs was working on NeXTstep for Intel by 91 and pressured Intel to produce a pizzabox design similar to the NeXTstation. Our booth was a handful that had NeXTstep 486 (later just NeXTstep and eventually Openstep when NeXT's hardware was phased out) machines running on those workstations. They did the job great and ran very fast compared to some of the other 486 computers demonstrating product around the show.
Speakers in the power supply are common on post-WWII electronics from the 60-70s. Probably had a former military contractor working on it. Only reason I know for putting it there is to make use of dead air and move a little air around.
Cool find. We used to have one of these at home bought 2nd hand when I was a child. Up until now I wasn't able to find anything about it anywhere and assumed it was made by IBM as ours came with an IBM monitor and an IBM Model M keyboard.
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.
So, I have one of these (and EISA adapters like Token Ring) - We should talk. Are you still looking for a 'Type 1' (DB-9 to hermaphrodite connector) Token Ring cable?
This system is a clone of the NCR corporation workstation used typically for point of sale only a bit more powerful over your typical POS system since it's also where you would have stored things like inventory databases and such. It was a perfect solution for modernizing small business that didn't have things like a back office or not much of one. A company I worked for out of high school deployed these around town so I was quite familiar with them since I also serviced them. I am glad you found the EISA configuration diskette too as that's probably just as useful as the bios firmware. Often times if the clock battery dies, so will the critical EISA configuration data as well throwing up a flag as it would with IBM's Microchannel. Interestingly you needed a utility diskette for that bus too. I used to keep around 30 or 40 diskettes around just for these types of systems.
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.
Although EISA cards promised easier integration and didn't have to worry about IRQ conflicts, the big problems with EISA is that the boards couldn't self-configure; you had to have the magic floppy that came with each card. At the time, Apples and even Amigas had self-configuring cards for their slots. A classic overthink by IBM.
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.
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 ...
What’s the big prominent socket on the opposite side of the CPU and the FPU socket good for? I can’t remember seeing anything like that on a motherboard before.
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.
Wow I remember these systems and yes there isnt a lot of information on them same with there servers from the same era no intel badging there usually in strange un marked cases
those little captions in the video get blocked out by the youtube captions (tried moving them but they always go back to the bottom of the screen), could u consider moving those in-video captions to the top of the screen in future videos?
I noticed this problem when TH-cam 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 TH-cam to let me move the CCs up about half an inch.
@@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.
I had this machine in 1995 for running NEXTSTEP 486, it was a great little workstation. I have been trying to find one on eBay to buy, but they’re rarer than hen’s teeth nowadays.
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.
@@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.
For washing, you could use an ultrasonic cleaner and put isopropyl alcohol in it. This may also be interesting for mechanical components, especially floppy drives, as it will clean cavities otherwise unreachable. I cleaned a modern GPU with it, but I bought it for cleaning resin-type 3D prints.
I wonder what that 2nd socket for a 486 CPU is for, you can see it next to the main 486DX cpu on the motherboard... 486DX'es have FPUs in them so it should'nt be for a math co-processor... Either I missed something, it's a 486SX with no FPU, or it's somehow got support for a 2nd CPU, which I don't think a 486DX supports...
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 )
You may have found this but here is a PDF on the machine including the PSU pin out if you need that in the future. www.elhvb.com/mobokive/Archive/Intel/professionalworkstationtps.pdf Cool machine though, I like it! (And great content you make!)
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.
@@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.
As another old hardware flunky (cuz they do make us carry their chips) .. very interesting. I'd never seen one of these. My solution for the data dump would be somewhat different... recently had a similar situation, client's Mac of about that same era, totally dead other than the HD. Found a modern SCSI card ($10 on eBay) that supports this class of drive (didn't even need an adapter cable), plugged the whole assembly into my modern PCLinuxOS system, and just copied the data off the drive. But it's gonna be fun to see how you get that thing to network. I have faith. :)
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 :/
Cool machine :) the easyest way to backup the data from the scsi HDD is to connect it to other modern PC or add IDE HDD to this one and copy all files to it :)
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 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.
The way I would deal with measuring dangerous voltages inside a PSU, would be to attach some wires while the machine is off, bringing the points to test out of the case. Ensure the ends of the wires are appropriately anchored, with plenty of space where nothing can snag them. While you're still dealing with dangerous voltages, you can manage the possible threats. You might even consider attaching the negative lead of your multimeter to the power supply chassis so you only need to manoeuvre a single probe while the power supply is on. Make sure to think about how you'll turn power off, before you turn it on. Of course, no part of this procedure is a replacement for understanding the dangers of a power supply. Don't try this at home.
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
I've always heard that Intel did the port of NeXTSTEP to the i386.. I wonder if it works out of the box so to speak. It'd explain the EISA support in NS for no apparent reason.
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.
This is a PERFECT example of fixing something that ain't broke and where the "fix" breaks it. This is not a vintage radio with 300 volts running across wax capacitors with a bunch of tubes heating up the enclosure. This a modern solid state device with a high quality power supply.
... 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.
+1 For the Montreal Shirt! :) if you're from Quebec let me know... I do have something that you could make a video with! have you ever heard of Ogivar?
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.
Rare workstation and nowhere to be found hardware specs.... :-( i'm mostly interested in graphics chip, id like to see what intel considered workstation worthy at that time and not a single word about it.....
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'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.
underrated TH-camr, I'm shocked people didn't even come to see the second part of the SLS Linux video and forgot you after that Reddit post... :(
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.
My inside geek is in love with this Channel!
When I worked for a NeXT VAR/OEM in 1993, we had a few of those workstations at our booth at the NeXTworld Expo provided by NeXT. The (alleged) 'story' is that Jobs was working on NeXTstep for Intel by 91 and pressured Intel to produce a pizzabox design similar to the NeXTstation. Our booth was a handful that had NeXTstep 486 (later just NeXTstep and eventually Openstep when NeXT's hardware was phased out) machines running on those workstations. They did the job great and ran very fast compared to some of the other 486 computers demonstrating product around the show.
Speakers in the power supply are common on post-WWII electronics from the 60-70s. Probably had a former military contractor working on it. Only reason I know for putting it there is to make use of dead air and move a little air around.
Cool find. We used to have one of these at home bought 2nd hand when I was a child. Up until now I wasn't able to find anything about it anywhere and assumed it was made by IBM as ours came with an IBM monitor and an IBM Model M keyboard.
I remember these systems when I worked at Intel. They were great systems for the time.
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.
So, I have one of these (and EISA adapters like Token Ring) - We should talk. Are you still looking for a 'Type 1' (DB-9 to hermaphrodite connector) Token Ring cable?
I contacted you by the Facebook :)
This system is a clone of the NCR corporation workstation used typically for point of sale only a bit more powerful over your typical POS system since it's also where you would have stored things like inventory databases and such. It was a perfect solution for modernizing small business that didn't have things like a back office or not much of one. A company I worked for out of high school deployed these around town so I was quite familiar with them since I also serviced them. I am glad you found the EISA configuration diskette too as that's probably just as useful as the bios firmware. Often times if the clock battery dies, so will the critical EISA configuration data as well throwing up a flag as it would with IBM's Microchannel. Interestingly you needed a utility diskette for that bus too. I used to keep around 30 or 40 diskettes around just for these types of systems.
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.
Interesting hardware you got there. Great video
Although EISA cards promised easier integration and didn't have to worry about IRQ conflicts, the big problems with EISA is that the boards couldn't self-configure; you had to have the magic floppy that came with each card. At the time, Apples and even Amigas had self-configuring cards for their slots. A classic overthink by IBM.
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.
6:45 is see an defective electrolyte ( or ist an shadoww on top ? ), did you replace
him ??? I´ll think there are more to replace (prevent).
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 ...
What’s the big prominent socket on the opposite side of the CPU and the FPU socket good for? I can’t remember seeing anything like that on a motherboard before.
The back of that case reminds me of late model Sun equipment, IPX era stuff.
Wish I had found you before I threw away my Commodore VIC 20 and Commodore 64. Like the video.
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.
Wow I remember these systems and yes there isnt a lot of information on them same with there servers from the same era no intel badging there usually in strange un marked cases
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?
those little captions in the video get blocked out by the youtube captions (tried moving them but they always go back to the bottom of the screen), could u consider moving those in-video captions to the top of the screen in future videos?
I noticed this problem when TH-cam 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 TH-cam 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.
I had this machine in 1995 for running NEXTSTEP 486, it was a great little workstation. I have been trying to find one on eBay to buy, but they’re rarer than hen’s teeth nowadays.
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.
Is that an EISA sound card? Wow!
It's not, just a standard ISA one but its not a soundblaster compatible one either :)
For cleaning of large surfaces with isopropyl alcohol it makes it easier and faster to use a toothbrush instead of qtips
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. :-)
Lucky it has a DS1287. Many EISA systems have the DS1385 Realtime RAMified chip with more battery backed RAM in them which is long EOLed.
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.
"looks like an easy solution to get the files out" - I bet attaching the disk to a modern machine and using ddrescue would be far more easy :D
14:29 essentially CMOS before they just made them removable coin batteries
For washing, you could use an ultrasonic cleaner and put isopropyl alcohol in it. This may also be interesting for mechanical components, especially floppy drives, as it will clean cavities otherwise unreachable. I cleaned a modern GPU with it, but I bought it for cleaning resin-type 3D prints.
I wonder what that 2nd socket for a 486 CPU is for, you can see it next to the main 486DX cpu on the motherboard...
486DX'es have FPUs in them so it should'nt be for a math co-processor...
Either I missed something, it's a 486SX with no FPU, or it's somehow got support for a 2nd CPU, which I don't think a 486DX supports...
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 )
You may have found this but here is a PDF on the machine including the PSU pin out if you need that in the future. www.elhvb.com/mobokive/Archive/Intel/professionalworkstationtps.pdf
Cool machine though, I like it! (And great content you make!)
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!
7:05 That black cap you are cleaning around looks distinctly bulged at the top - a.k.a. failed. Maybe just an optical illusion on the vid though.
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.
Speaker inside power supply? That's like your ear being on your knee.
As another old hardware flunky (cuz they do make us carry their chips) .. very interesting. I'd never seen one of these.
My solution for the data dump would be somewhat different... recently had a similar situation, client's Mac of about that same era, totally dead other than the HD. Found a modern SCSI card ($10 on eBay) that supports this class of drive (didn't even need an adapter cable), plugged the whole assembly into my modern PCLinuxOS system, and just copied the data off the drive.
But it's gonna be fun to see how you get that thing to network. I have faith. :)
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 :/
Cool machine :) the easyest way to backup the data from the scsi HDD is to connect it to other modern PC or add IDE HDD to this one and copy all files to it :)
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.
The way I would deal with measuring dangerous voltages inside a PSU, would be to attach some wires while the machine is off, bringing the points to test out of the case. Ensure the ends of the wires are appropriately anchored, with plenty of space where nothing can snag them. While you're still dealing with dangerous voltages, you can manage the possible threats. You might even consider attaching the negative lead of your multimeter to the power supply chassis so you only need to manoeuvre a single probe while the power supply is on. Make sure to think about how you'll turn power off, before you turn it on. Of course, no part of this procedure is a replacement for understanding the dangers of a power supply. Don't try this at home.
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
I've always heard that Intel did the port of NeXTSTEP to the i386.. I wonder if it works out of the box so to speak. It'd explain the EISA support in NS for no apparent reason.
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.
This is a PERFECT example of fixing something that ain't broke and where the "fix" breaks it.
This is not a vintage radio with 300 volts running across wax capacitors with a bunch of tubes heating up the enclosure. This a modern solid state device with a high quality power supply.
... 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.
+1 For the Montreal Shirt! :) if you're from Quebec let me know... I do have something that you could make a video with! have you ever heard of Ogivar?
drill the screws out
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.
be glad you didnt get a server there server's where 70 pounds without the hardware installed in them
Rare workstation and nowhere to be found hardware specs.... :-( i'm mostly interested in graphics chip, id like to see what intel considered workstation worthy at that time and not a single word about it.....
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.
Do NOT go inside an psu when you have no knowledge of electronics. It can kill 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.
just hose it all down and let it dry out in the sun. Waste of energy to use cotton buds like that
I live in New Jersey. The sun only works if you don't have a cloud of Newark blocking you :)