This 8" hard drive doesn't seem to exist (or does it?)

[@adriansdigitalbasement]

Additional information from Patron Peter Kooiman: it appears this drive is the (or is similar to) the Priam DiskOS1070 and that the interface card is the Priam Smart Interface. There is documentation on it here: www.bitsavers.org/pdf/priam/DISCOS_1070_Product_Specification.pdf www.bitsavers.org/pdf/priam/SMART_Interface_Product_Specification_Nov80.pdf It does require a spin-up command so that explains the lack of activity! Computer Design 1981 volume 10 has an ad for the drive: US$1985 with the interface card included, power supply an extra US$295 Ad: imgur.com/XZPVZ9w

[@dauerwerbesendung7045]

I"m sold. All of that for as little as $1895 with Terminator included, of course!

[@Chris_In_Texas]

If you reference the #4 from the listing of defunct HDD companies, you will notice down in the tables that #6 lists that Priam used/contracted Hokushin Electric as the assembler for the HDDs itself in the early 80's. So that would make sense. 👍😁

[@mrfrenzy.]

That's a really great find. Reading the manual explains the hour counter, mean time to failure 10 000 hours which is just over one year. They do say that most repairs should be completed in ½ hour.

[@jacktheripper6716]

I found old adverts 1980 to 1982 they started at $1899.99 (matchmaker technology) other info i found was a page in mini-micro systems 1982 sad to say its digital scans are gone

[@kittyztigerz]

i recognized that pcb before it need pc to tell it power up to turn on motor if no communication from pc it does nothing it in standby mode saving power waiting for communication from pc when it get communication from pc it switch to power up mode i heard it from someone own it just trying save it self from not being worn out unlike other hdd just get power and spring up i think there is way troubleshooting it if you have hook up to pc with diagonal in slot can be come handy i dont know if they video still out there or not i take look n see what can find

[@JRoughgarden1]

That power supply was built in Hackensack NJ. I was the computer operator for their business applications. This unit was most likely early 80's as the later part of that decade RTE from Wisconsin purchased PowerMate rebranding was RTE/PowerMate. BTW that was not the original power cord.

[@nutzeeer]

The internet sure is small sometimes. Cool to know!

[@1993MAZDAMIATA]

Dang newjersey

[@gregm1457]

Ive used a few different PowerMate supplies, they are good stuff. Those arent a good match for the drive... I guess the last operator had them on hand, so thats what he used...

[@brucereynolds7009]

Related to the Roughgarden family of Midland Park? Paul, Nancy.

[@TombstoneChris]

That means you're old.

[@magicphred]

The "Computer History Museum" has a link to a "CD-8010P" Hard Drive (Inside an Advanced Computer Design PDQ-3 Computer) You can also find ( from "bitsavers" ) "PDQ-3 System Users Manual Ver 3.1 Apr81" and "PDQ-3 Hardware Users Manual Jan81" (I'm not sure if the weblinks I added in prior messages will get deleted)

[@alakani]

Makes sense, Computer History Museum is in Mountain View and that Priam board was made in San Jose a few minutes away. I bet that thing at some point passed through a place I used to work called WeirdStuff Warehouse in Sunnyvale, I miss that place heh

[@JamieStuff]

@@alakani Ah, WeirdStuff. The best place to buy swag from defunct companies. I still have my MediaVision screwdriver that I bought from that store. Across the street from the old Fry's that had "Enter" and "Escape" keys on the doors. Pre DotCom implosion.

[@neillthornton1149]

@@alakani man, Weird Stuff. Used to love just walking through there and seeing what they had.

[@paulstubbs7678]

It seems the only links KZbin leaves alone are links to other KZbin video's You can probably blame this on all the scam artists with links (and Telegram 'codes') to various get rich quick crowds (In reality, get poor quick, with their help)

[@tinytonymaloney7832]

Does that mean with the extra info we can have a part 2?? Basic knowledge only of electronics and logic and this was way above me but I was fascinated.

[@Nas_Atlas]

I love how this channel has gone from swapping chips on a commodore 64 to some serious hard core geek material. Awesome stuff.

[@crazyidiot5309]

I'm completely here for it too.

[@evaDrepuS]

For some of us it does not matter that you might not get something going, it is still enjoyable to watch you work through your thought process, as we still learn from it.

[@8bitrocketstudios]

100%

[@gmirwin]

Exactly.

[@koekone]

Enjoyed this episode really much, thank you.

[@Kassetkyn]

+1

[@halfabee]

In about 1983 I was working for Link House publications as a Analyst Programmer. The computer room was quite large. 60 foot by 40 foot. Filled with 4 foot square hard drives. The operator used to walk around the hard drives every morning checking the temperature of the air coming out of the hard drive, One morning he noticed one drive was hotter than the others hence he closed the computer system down and called an IBM engineer who changed a bearing on the drive.

[@lolbots]

a living resource monitor lmao

[@PeterSwinkels]

That thing is small considering the original harddrives were the size of a washing machine...

[@PeterSwinkels]

LOL - I accidentally replied to someone else's comment!

[@paulstubbs7678]

Wow, what a different world. These days the whole contents of that room would fit on a USB memory stick, kind of crazy

[@PeterSwinkels]

@@paulstubbs7678 : indeed

[@DEtchells]

A quick note on the white strips on the processor board: I agree that they’re for power distribution and decoupling. I forget the manufacturer (US, I think it started with an “R”, Rogers maybe?), but I used a product back around ~1990 for decoupling chips on problematic boards. They were thin flat packages that sat under the chips, with pins sticking down into the power and ground pads. You’d stuff them in the board, stuff the chips in on top of them and then solder. The big advantage was that they had essentially zero inductance because the pins commented directly to the flat electrodes of the capacitor. I used them on a 2-layer board that had noise problems and they did the trick. I think the same company also made longer units for power distribution. They made the board layout easier because you didn’t have to worry about power/ground routing, and also had the advantage of very low resistance and inductance (the conductors were flat strips of relatively heavy-gauge metal), and the bypass capacitance was fully distributed along the length of the bus.

[@DEtchells]

@SiliconValeyGeek Ah, I'd forgotten about the SIP versions; we had some as samples, but only used the flat packs in production. Good note on the 74Sxx parts - I never had to deal with them but think I used some 74AS, which were actually faster, but their speed-power product was less than a third that of 74S.

[@jamessmith1652]

I haven't a clue anything Adrian talks about. I just enjoy watching all his videos, pleasant and intelligent guy talking passionately about his interests.

[@ricardog2165]

Adrian is the Bob Ross of Retro Computing!

[@christinesmith734]

Same, I know very little about what goes on inside a computer. Adrian makes it so interesting.

[@ChildrenOfTheRosary]

@@ricardog2165 Happy little IC’s!

[@iteachtime]

Looking forward to part II of this very interesting hardware. Thank you Adrian for posting these even when there isn't an instant resolution. Gives me something to look forward to!

[@BersekViking]

The spikes you see on the 5v rail and other places is caused by the long ground lead on your scope probe. 1: It picks up noise from the board and that is "in series" with probe so it looks like it is on the signal. 2: Your grounding is far away from where you are probing. Ground is not ground when it comes to high frequency AC signals. Always try to use a short ground cable to the probe as close as possible to the place you are probing.,

[@gregm1457]

Yep ground noise will get you every time 👍

[@qviewq2071]

We used to maintain giant 14" Shugart hard drives sold as Altos 8000 CP/M and MP/M systems. Total capacity 15 MB. We learned that if you checked the heads for continuity with a multimeter it would burn out the heads as the wires were so fine - and they used to spin up too fast and the drive belt would spin off so Workshop Terry devised a plan to put a 1 ohm resister on the motor to slow it down at start up - but it got hot. He them mixed araldite and heat sink compound to try and stick the resistor to the casting that the drive was built on. He found his mixture combined the sticking power of heat sink compound with the thermal qualities of araldite. Fun times.

[@tim0steele]

Given that you mentioned seeing 4.7V on the board, I'm guessing the power supply was turned up to 5.3V to give 5V on the board.

[@adriansdigitalbasement]

Yes, that might be the case because of voltage drop. I wasn't checking the voltage at the chips

[@senilyDeluxe]

@@adriansdigitalbasement Does the PSU have a sense line? That would explain the voltage going up under load when measuring on the PSU.

[@garthhowe297]

Yes, 5.25", and 8" drives drew a ton of current on startup. I remember on Intel based servers, when the power supplies were getting a little weak, we would have to unplug the power connector from one, or more, 5.25" drives, and plug them in one at a time. Once they were all spun up, do a hard reset, and then you could load the OS. Once we got to 3.5" hd's, the current demands were less, but the problem remained due to a much larger number of drives. On the Compaq servers, they would sequentially power up the hard drives (by SCSI id order) as part of the POST, in order to minimize peak loads. Great video...not a failure at all.

[@tw11tube]

I still have a Compaq Prolinea server in my "digital basement" (of course not powered on 24/7), and you can clearly hear the drop of the +12V supply while the hard drive spins up, as the fan RPM goes down. And that's just with the single Fujitsu M2623F half-height 3.5inch drive that was the basic storage option. The peak spin-up current is 2.5A @ 12V. Don't believe marketing touting this drive as "high-performance [...] hard drives, which increase the performance levels of high-end personal computers, workstations and network file servers", though. They are solid drives, but at 4400rpm and less than snappy seek times, they are mid-level for workstations and bottom-of-the-barrel for servers. An official option for that system was the Compaq SMART-RAID controller with up to 7 drives connected to it, and you would most likely buy bigger and faster drives if you went the RAID route. Of course they did staggered spin-up by default.

[@gblackhead]

Hello Adrian Just a few side notes: Most of the chips are made in 1980 & 1981. Even the EPROM chips. But the CPU on one of the boards seems to be from 1984. It could be serviced/replaced some time after the production, or the the drive could be completed some time later, after those boards... The 5V line ripples could be caused by the surge in the logic, trying to do something, or by the output lines trying to drive the power transistors to actually "move" something... SASI stands for "Shugart Associates System Interface" For the SCSI there are single-ended and differential modes. The single-ended works with signal wires versus the ground. But for the differential, it works with two data lines, one driven up to the "+" signal, the other down to the "-" signal, changing the both wires polarity with the data changes. So the terminator connecting the two data lines could mean the differential mode. And there is only 8-bit data in the SCSI-1 (I wasn't able to find much about the SASI pinout). You could use some sort of the logic analyzer (f.e. one based on the Arduino) to trap just all of the address pins (together with ALE pin) to see the address at which the code is cycling... Then, with the EPROM code being disassembled you could see more of what's going on... And last but not least: THANK YOU for the very interesting content! ;-)

[@stevemcknelly5036]

I'm guessing that drive was made prior to 1983, since Hokushin Electric Works was merged with Yokugawa to become Yokugawa Hokushin Electric Works in 1983. Back in the 90's, there was a massive text list of HD drive info called "hdlist". I tried to find a copy of it to see if Hokushin was on it, but couldn't find a copy.

[@adriansdigitalbasement]

Yes indeed, the date codes on the ICs are all 1981.

[@jwhite5008]

Irregardless of this drive, it may be good idea to find and archive this document. If you looked at it in the past, you might remember some text that was definitely in there. Try to google all different things that could be in it in one search query. Did it just have names, or also some kind of parameters? Like heads/cyls/tracks? Or interface? Or size? If so, you could probably find some parameters somewhere else and google them all together. Try putting a dash between numbers that were close together. Although it could easily have been just an internal document that got passed about a few companies.

[@GrayRaceCat]

I used to work for circuit (PCB) makers in the late 70s early 80s, and I would bet that the E.C.1283 marking is the circuit Manufacturers initials followed by the month and 2-digit year, Dec 1983. It was a common practice back then. For example, a company I worked with in Fort Lauderdale, FL was called Tropical Circuits and they would have put TC1283 on their boards.

[@kevinchastain727]

The spin up problem maybe because the drive is set on " Start on Command " this is set with a jumper on the drive itself it will power up the board but will not spin the drive until it receives a command to start the clunk you hear on starting the drive is the head locking to parked position.

[@5nowChain5]

In the London UK area (86-91) we used to use Northamber (located just off the A3, near Kingston) who used to repair all industrial computer hard drives and floppy drives of all sizes, including the ones for ancient Mainframes with the early huge Cabinet harddrives.

[@rakslice]

The terminator pack you're looking at 41:26 is on the board directly connected to the drive, right? I feel like there is confusion here, if anything here is going to be a peripheral bus like scsi/sasi it's not going to be that, it's going to be the connector that goes from the other board -- presumably the drive controller -- to (whatever host adapter installed in) the computer. Like think about how ST506 drives were used with SCSI using ST506 to SCSI drive controllers like ACB-4000A you would then connect to your scsi bus, and those are later more miniaturized 5 1/4" hard drives.

[@mfree80286]

It's a proprietary bus from Priam, an NRZ serial type.

[@aCivilServant]

I remember some of the IBM servers we had used an option via the SCSI controller to either start up the attached SCSI drives instantly, or used a 'cascaded' start up based on the drives SCSI ID. This allowed them to start up one by one and reduce the load on the PSU. Pure speculation, but perhaps this drive is only programmed to spin up when it receives the correct command from an attached controller / host rather than when it's simply switched on?

[@DiskWizard001]

This option varies on different controllers . Remebering SCSI drives - there was a jumper option to enable/disable disc behavior on power on.

[@billmiller4800]

Since the start current on drives this size was massive, sequential start was important to not blow a breaker or smoke some wires. I have an old 5.25" SCSI that takes 15 Amps to start (at 12V) and is like a jet engine as it winds up, and this thing is even bigger, so I don't doubt the external start signal required. I worked with an old NCR mini in the 90's, and its power supply was good for 200 amps at 12V for the 4 drive chassis expansion module, running off 220V. That monster cabinet was something like 300 lbs and about the size of a large tower PC of the time to house 4 drives.

[@russellhltn1396]

The voltages seems a bit wonky. TTL chips need 5V +/- 0.25V. So, 4.7V is too low. You might want to adjust the 5V supply to 5.1V to allow for power loss in the cables/traces.

[@benbaselet2026]

if 5.25 is upper threshold of the spec then it should be safe to go around 5.3 on the PSU end from the start.

[@8bitwiz_]

Quick hint for disassemblers: 11K is 8085 code 0000-07FF 12K is 8085 code 2000-2FFF (same address space as 11K) L15 is 8085 code 0000-07FF

[@Breakfast_of_Champions]

11:55 there is a *nix console command "strings" to do just that

[@NicolaiDufva]

Was thinking the same. Looks like Adrian use Windows. Cygwin's binutils package has the same utility.

[@Breakfast_of_Champions]

@@NicolaiDufva I found my version was even one of the many Sysinternals tools

[@MurderMostFowl]

After buying my son a 512GB Micro SD card for Christmas for $49 to replace the 32GB one he had and promptly threw away really puts 10MB in perspective

[@RandR71]

DIP SWITCHES: it's a long shot, however some of them oxidizes, try flipping all of them, a couple of times, ending where they are now, and/or check their continuity, if not already done.. of course.

[@OtherWorldExplorers]

I really love this non-standard stuff. Trickle it in whenever you have the time to. I'll look forward to the next episode after I'm done with this one.

[@Farm_fab]

I visited a trade school where the computer lab had one of these. The instructor said he wanted to build a laptop with it. I had one similar to yours back in the mid 80's. I worked near a computer manufacturer that was competing against IBM, and found a lot of goodies in their trash dumpster, including the HD.

[@rvenden]

Hi Adrian - I love the forensic work you do old gear, such as this hard drive and controller. Your approach and thought process is very impressive and very educational. Thank you for the great work and even greater videos. I cannot imagine how much time this must all take. Thanks again.

[@FXGreggan.]

Had something similar in a PDP-11 at work, it was a laser marking semiconductor chips. I was the service engineer maintaining the machine and I remember the 10Meg Winchester HDD was very heavy and it needed several minutes to spin up and get warm before being able to use it...

[@chadhartsees]

This is fascinating. The controller board certainly looked different than others shown on the channel - those strips - I don't think we've seen something like that on the channel before. It's fascinating seeing different ways people made boards and solved problems.

[@jwhite5008]

You might like to look at main board of Centurion computer that "Usagi Electric" is restoring. That one has actual wires (not traces) INSIDE the board epoxy. The video is called "Minicomputer Part 4: In-depth Look at all the Computer PCBs" at timestamp 19:56

[@chadhartsees]

@@jwhite5008 ooh! I've watched some in that series, I don't know if I've watched that part. Thanks!

[@paulstubbs7678]

I remember seeing ad's for very similar strips, they were advertised as bringing low impedance DC and bypass caps right up to the IC:s. These days with multi layer circuit boards with a ground & supply plain this is no longer an issue, but back then two layers PCB's was all most had access to

[@davros1971]

Back in my dim dark past of the early 80's I was working on/supported some Basic 4 (en.wikipedia.org/wiki/MAI_Basic_Four) computers and S100 devices around Sydney For Rank industry's one fun job was that they had an s100 card system that was installed in the Sydney(Australia) container port that some ham-fisted dock worker attempted to reseat some cards one card had strips like that but mounted 90" to the surface of the board they supplied the power rails and also isolated each row of chips my muscled friend had managed to break a few connections and making them intermittent. a couple of hours with a magnifier and fine wire fixed the issue in those days it was weeks/months to get weird parts by ship or $ 1000s for air freight so it was mostly surgery, canablise or invent something to replace what was broken once knocked up a replacement lm741 op amp out of bits to get a industrial metal detector working while I waited a week for a replacement part

[@TheSimTetuChannel]

Happy Holidays Adrian! Thank you for all the great videos you've brought us throughout the year. I wish you health and happiness for 2023 and beyond!

[@Solanza]

Respect for including measurements like Kg as a sidenote on the screen, don't see allot of youtube channels puttin in the extra effort nowadays, keep it up Adrian!

[@tekvax01]

(29:25) Power resistors do not lower the voltage; they limit the current. Simply stated, when the current is limited, the measured potential difference (voltage drop) appears to be lower due to the input impedance of the load.

[@dariodzimbeg]

Thank you for including metric system as overlay text.

[@lohphat]

I worked for a systems integrator which specialized connecting SASI (SCSI precursor) to TRS-80 Mod II systems as we had one of these 8" drives -- it was either from Ricoh or Hitachi and I still remember the spin up sound almost 40 years later. IIRC it was a 300-350 MB drive. It was "huge" for its day.

[@Renville80]

Some of those older linear supplies used a honking big transistor in a ‘series pass’ configuration, and the actual regulator is a LM723 IC or similar.

[@PileOfEmptyTapes]

Yeah, very common setup for lab power supplies back and such in the day. Fun fact, the trusty '723 actually has a really low-noise voltage reference even by modern standards.

[@Bob-1802]

About 110 chips on the top board. Almost as many on the other board, this accounts for about 200 chips. Can you imagine these were most likely installed by hand before going to the soldering wave.

[@grahambird1570]

I worked at British Airways as an HVAC Engineer, 21 years ago at their Data Center area, at Bodicia Hse. In the background on this Huge site there were a couple of Condemned Buildings, one of them was called Comet House, which was infact the Airway Staff Medical Center and Hospital. The Building had been 'Ear Marked' for Demolition but there was alot of Politics involved because of Asbestos Presence and was Boarded up. But, we got through the Boarding and went on a journey of discovery. Now, me being interested in Computers and Technology found a Redundant Storage area of Dozens of old Form Computers, Boxes and Boxes, lines and lines of shelving containing electronic components. Now, I saw a few dozen of the Hard drive and boards that you are exhibiting. They certainly weren't screwed to a Board like yours and were in a Proper or appropriate Container in a similar configuration to the Machine you are showing us. I noted the 50 Pin configuration and here you have revealed this to us. Apparently the PC's then were used for and by the Pay Offices and Staff Records.

[@Tyle_smalcu]

Merry Christmas, Adrian! Thank you for another year of awesome videos! ^_^

[@hugoegon8148]

21:00 I just use car light bulbs as loads for power supply. Different combinations for different voltages or currents.

[@wkrick]

You might be able to find that UK ebay page in your browser cache.

[@Penfold8]

That looks like an IDE cable. I was actually thinking you were going to have an MFM cable when I saw the size of that drive.

[@wb5mct]

The speed sensor is very likely a variable reluctance type, as evidenced by there being only two wires feeding it. (Hall devices require power). Variable reluctance sensors are a essentially a magnet inside a coil of wire that generates an AC voltage when a ferrous metal comes closer/farther away. Very simple and robust!

[@timballam3675]

Seen some scanned Priam documentation on a polish site but they want payment for the PDFs they must be available elsewhere?

[@Alex-je6od]

I feel like everything is actually working as it should be. You just need to issue a spin-up command to start the drive. SCSI does something similar. (1B START STOP UNIT)

[@tjs114]

In the mid-1980s, we had Hokushin hard drives - 5mg and 10mg attached in an outboard cabinet to a hot rodded MicroVax and the interface was Shugart ST412.

[@richardchambers256]

Milligrams? Thanks Doc! 😆

[@joejohnson8132]

I'd really love to see an update on this now you have more info! Would be great to access the disk.

[@JoeMcLutz]

Merry Christmas Adrian! 🎄😊

[@unmanaged]

Hokushin Denki later became Yokogawa-Hokushin Denki, and now called Yokogawa Electric...

[@scottlarson1548]

In 40 years of owning computers, I've never owned a computer that had as many chips as that disk drive.

[@Rx7man]

when you ask "How complicated can it be to put 1's and 0's from the data bus onto physical media... well, this is the answer! Also interesting is the clock speed seems to be competing quite well with the clock speeds of the computers it was hooked up to!

[@scottlarson1548]

@@Rx7man The Ohio Scientific floppy disk interfaces were nothing but a PIA (6821) to control the head and a serial adapter (6850) to read and write the data. There was also a "data separator" to remove the clock from the data stream, but that was it! Three main chips and everything else was done in software. Hell, it wasn't much more complicated than their cassette tape interface.

[@null_carrier]

That could easily be a Control Data Corporation (CDC) SMD (Storage Module Drive) interface. Just recently I had at hand an 8'' Fujitsu HDD from 1984. using that interface. It was used in Boroughs (Unisys) mini and micro computers.

[@Distinctly.Average]

I’ve worked on those systems. I’ve also seen that drive here in the UK. I cannot remember if it was on an ICL system or a Boroughs/Unisys one as it was a long time ago. IIRC, we had to keep spares in the room with the mainframes so they were at the right humidity and temperature inside when they were fired up as not doing so reduced their life.

[@Torbjorn.Lindgren]

From what I remember SMD has the common shared control and per-disk data cable, and neither of them were 50-pin. Yeah, SMD was standardized as ANSI X3.91M and the Internet Archive had a copy. According to this SMD uses 60 or 75-pin control cable and 34 or 26-pin "read/write cable", so it's not standard SMD at least. Beyond that, anybody knows.

[@null_carrier]

@@Torbjorn.Lindgren I still have access to those drives (as they are now part of the company's 'mini museum' so I can have another look. You might be right about the cables and pin counts.

[@null_carrier]

@@Torbjorn.Lindgren According to new documments Adrian just posted in pinned comment it seems that his drive is of Winchester (or variation) type. SMD was an attempted improvement upon Winchester and an incompatible competition protocol so I'm afraid my experience wth Fujistu drives is of no use in this case. Ihope that someone else can help. It would be interesting to see the drive working again.

[@Torbjorn.Lindgren]

@@null_carrier SMD is just a protocol like SCSI, SATA, MFM and so on, there were even more of them them back in the day. Exactly how high or low-level each protocol was varied but they're still "cables and what runs over them". All modern disks drives are of "Winchester type", named after the 1973 IBM 3340 disk drive with code-name Winchester, meaning they are based on the formula this introduced (sealed unit with lightweight heads floating on a air cushion generated by the spinning disk). It was so influental that in a few years everyone was using the same basic idea and as a result was common to use Winchester or Winchester disk as the name for hard disks even in official documentation for pretty much all computer equipment. I strongly suspect every single 8" fixed hard disk was "of Winchester type" so it really doesn't narrow things that much. This use gradually faded away in the early 90s because no one built any disks that weren't "Winchester type" but it's not surprising that the document use that term. Heck, if you find the technical manuals for your Fujitsu it'll probably use that term too.

[@DavidRavenMoon]

That drive is HUGE! Was probably very expensive. I worked for a graphics company in the early 90s that had a big Scitex computer system with big 2MB drives like this. And tape backup on reel to reel. Looked like something from a movie.

[@Birb_of_Judge]

I wonder how people during that time would have reacted if you told them that you can buy a 4TB 3,5" hard drive for under 100$.

[@DavidRavenMoon]

@@Birb_of_Judge I was so excited when I got an external 1GB SCSI drive for my PowerMac 6100! Lol. When I got that in 1994 I added 8 MB of RAM for about $800! Then prices came down and I sold that to a co worker for like $300 and bought 16MB for $700! And don’t get me started in my $600 Iomega Jaz drive!

[@ivilmi]

How about connecting it to good old Adaptec 2940 scsi controller and see if gets recoqnized.

[@ZakHooiTM]

Or atleast map out the pins voltage of the drive board and the empty connector at the adapter board, then see if the pinout matches anything known

[@rasoirwolf]

I wanted to see the gigantic drive spin :c Oh well - I wonder if you could place this in your TRS-80 Model II, in place of the 8-inch floppy drive, if you could get that to work somehow, I think you said earlier hard drives were modeled on Floppy tech and that had an 8-inch Floppy - you can probably think of a baker's dozen reasons why that wouldn't work right away, but to a layperson like me - it's an interesting idea at least. 10MB in the late 70s and early 80s was probably quite a bit, you'd never fill it.

[@makskk6spm275]

I love such videos. Thanks Adrian! Would be great to post more videos like this where you do investigation of some old electronic artifact from the past.

[@annareismith6843]

I demilitarised took apart and destroyed the discs in many of those like that at Puget Sound Naval Shipyard I was working for. And stripped out many boards on computer server towers for precious metals of gold platinum and silver recovery in the late '80s and early '90s. So I have seen so many of those. I would see that one there a lot. Those and tape drives. A lot of optical hard drives too. I loved stripping them apart. I would have to smash and drill holes in so many Hard drive discs. Some really old and very big ones. I saw so much cool stuff most did not get to see or know what they are and even now I am still learning what they were. Some were so secret I will never know. And they are so outdated now though back then they were very high-tech. I saw the first CD ROMs before most did. A lot of things the average consumer could not get or afford I had to play around with. I even saw many Commador 64 used that were on ships. A lot of them came through where I worked in a large tall pile on many pallets for me to ship to be sold in DRMO DLA auctions. Many Printers Tape and Floppy Drives and Monitors for them too. I would grab the best ones and play games on them till they died I would have many of them at my work desk in case one died on me I could not fix myself when on break or at lunch or get there early to play them tell I start working. I used them to make documents I printed for shipping things too. I made my own program for it just to do that.

[@shadyweaver]

Very cool read, thank you for sharing this experience with us!

[@bobsbarnworkshop]

I just came across this video! I was in community college ET from 79-81. For my senior design and build project I made an all discrete LED mantle clock which consisted of 60 mini green LEDs for seconds, 60 mini yellow LEDs for minutes and 48 jumbo LEDs for quarter hours. It took over 25 ttl logic chips ( no clock chips, just counters dividers a decade drivers) all wire wrapped !

[@wmlindley]

I worked on some DTC (Data Technology Corp) SASI controllers, integrating them into the Heath/Zenith Z-100 ecosystem. The Z-100 was DOD's "Desktop One" back in the 1980s and our customers were mostly military installations, and folks from the Heathkit world. 50-pin is almost certainly SASI, but that host adapter (the upper board) is a weird one for sure! "EC" is probably Engineering Change and a date-code.

[@samuelcolvin4994]

I remember the zenith machines! Bulletproof monsters those were!

[@steveschulte8696]

SASI is short for Shugart Associates System Interface. The standard committee did not want a commercial reference in the standard name, and chose SCSI instead.

[@hessex1899]

My Sun 3/50, back in the day, had a second board on the hard drive, similar to what you have there. It was an ESDI to SCSI adapter.

[@Cinncinnatus]

this reminded me of a HD from an old 3B2 AT&T mini-mainframe and like you said an adapter board.

[@paulmichaelfreedman8334]

Did the ESDI interface have a 50 contact edge connector?

[@Cinncinnatus]

@@paulmichaelfreedman8334 tbh no clue been so many years since I touched gear like that. it might have in mainframe/backplane designed equipment.

[@paulmichaelfreedman8334]

@@Cinncinnatus I ask b/c I have a 5.25 half-height disk with such a connector. Wouldn't work with SCSI 8-bit controller. My guess is also that it's a proprietary disk belonging to some niche market process computer from the 1980s.

[@hessex1899]

@@paulmichaelfreedman8334 I honestly don't remember. it was over 30 years ago.

[@BowWowPewPewCQ]

Wow, first time viewer. This stirred a bunch of old memories. Glad I found you.

[@donaldcongdon9095]

That no-code ROM could implement a drive control state machine.

[@benbaselet2026]

I was thinking it might contain a hardcoded bad block map or something. It would be a pain to update it but I guess it's still doable.

[@robertharris2262]

Drat, I should have read further down the comments. Said basically the same thing higher up this list. Custom 8086 computer accessing SASI drive via special controller with state machine implemented from 4 x 4-bit PROMS. Hell on earth to troubleshoot

[@AureliusR]

Yeah, combinatorial logic in PROM was a very common trick back then. You use the address lines essentially as logic inputs, and the data lines as logic outputs. You can shrink a lot of 74 series chips into a PROM this way if you're clever about it. The result of course will be *very* random looking until you understand the logic function going on. I might take a poke around in that PROM and see if I can't figure out the logic equations used.

[@G7VFY]

It could be that the RELAY on the bottom board has dirty contacts. I think it might be some kind of 'soft start' for this drive. But I would start by testing the relay contacts.

[@jaredwright5917]

I'm curious whether or not the serial outputs on the 8085s are used for anything. They'd be great for diagnostic output in something like this.

[@michaelallen1432]

Many years ago I read the user manual for an 8-in Enterprise class hard drive. It was a little newer and a completely different interface. It was something called SMD however, one of the jumpers on it determined whether it would power itself up when you turn the power on or whether it required a command to start and what sequence they would start in or something like that

[@roymerkel8008]

Parish to ask, but you ever get this drive working, or is it still busted?

[@papafrank7094]

Adrian, I wish you and yours a very happy Christmas season, and I thank you for all of your videos and work you share with us.

[@crabby7668]

The cdc smd drives could be set to remote start by command from the host computer. Perhaps this has a similar setup and is waiting for a start command.

[@nigelman9506]

Hi, at 29:18, the two large resistors are frost heaters same as used in aerospace data recorders, those old drives do not automatically park the heads and is bad news if it loses power, the heads grind down into the disks, to me this drive could be an early attempt of a prototype flight recorder and failed as hard drives get damaged by knocks and vibration, look up flight recorders

[@InfiniteBrain]

Not disappointing at all. Pretty cool to start with no information at all and start building out all of the info that you've already collected. The collaborative effort that comes into play makes stuff like this even better as information that would never see the light of day suddenly stands out when people realize there is a "market" and that it's not dead technology.

[@johnjoyce]

I wonder if strings at that time might have been encoded differently than ascii and or even been half width katakana.

[@8bitwiz_]

Trust me, there are no strings in those ROMs. It's all code and data.

[@johnjoyce]

@@8bitwiz_ totally believable also. Japanese systems engineers are also historically very good at utilizing resources completely and leaving out strings here would make a lot of sense.

[@dragonfireproductions790]

The CPU loop is caused by it waiting from the mainframe to slap it back from its hibernation, it's used to save power and to prevent unnecessary wear to the motor bearings as it used to be brittle.

[@Andreas_Straub]

I had one of those 8" disk drives about 30 years ago. It used a SASI interface - kind of a predecessor of the SCSI interface coming several years later on the market. It should be realtively simple to interface to a parallel interface. I remember we added such a drive to a Motorola Exorciser at that time and running Versados from it.

[@geoffreed4199]

Shugart Associates System Interface, was either (0.5 or 1.0 MB/sec) then 2.5 MB/sec IIRC SCSI -1 was essentially the same interface at 5.0 MB/sec, Fast SCSI 10.0 MB/sec there are some older SCSI controllers that could be clocked down to SASI speeds to migrate your data off the old drives, or continue using them, but that wasn't technically supported.

[@StrangelyIronic]

Definitely not a disappointing video. It's always interesting just to see the older boards that are filled with logic chips over single packages. I've built my own emulation setups that use microcontrollers similar to how something like the C64 had a handful of special purpose-built chips to handle tasks. It's a bit silly when you could easily just emulate it all on a Pi nano or something, but I like the idea of this microcontroller handles io/storage access, this microcontroller acts as a main processor, another mc circuit handles graphics/sound, etc. It's cheaper and easier for me to breadboard that type of design with spare atmegas I used to use for custom keyboards or something over buying an fpga board and recreating the logic from scratch using schematics for something like an Apple ][. Of course, now something like the RP2040 would be way overkill and let you build a basic Apple][ in the size of a thumb drive almost.

[@chadhartsees]

What kind of metal was that chassis/frame around the hard drive? Steel? What makes it have that yellow patina?

[@Rob2]

Cadmium plated steel...

[@chadhartsees]

@@Rob2 Ah! I've seen a lot of that in heavy duty electronics, didn't know what it was. Why do they plate it with Cadmium?

[@Rob2]

@@chadhartsees Cadmium plating is similar to zinc plating, but it provides better protection. It is also very toxic so its use is avoided these days, except in situations where it can be well controlled (aviation, ships, etc).

[@Jimblefy]

Very cool. I'd love to see you get that drive running. Amazing retro tech.

[@salipander6570]

Adrian, do you take any ESD measures? It seems you touch all those old sram RAM chips with your fingers while moving around. This could damage them.

[@Derundurel]

I wondered about that too.

[@brunomoyano8727]

Sooo... what happened to it? it's been a year did it die or lived?

[@connecticutaggie]

Ah, the TMS2516, I recall when that was released and I know the story... or at least the outside view if it. The story starts first with Intel's first EPROM in 1971 - The 1702 2Kbit UV erasable EPROM. The was followed up in 2704 and 2708 in 1975. They revolutionized development since you could reprogram your ROM. Soon everyone (including TI) started copying it and racing to see if they could beat Intel to make a 2716 before they could. The one annoyance with the 1st and 2nd gen EPROM is to use them you needed more than just 5V. Well, Intel surprised EVERYONE when they introduced their 2716 which was 5V only (for reading). Also, it could reuse those supply pins for address pins so the 2716 an 2732 could use the same 24-pin package as the 2708. That made TI's 2716 inferior and incompatible with Intel's 2716 which forced it to change its part number. The TMS2516 was TI's "copy" of the Intel (5V only) 2716.

[@ovalteen4404]

I wonder if some of those 8085's were manufactured with built-in ROM like 6800's were? That could also account for missing code.

[@danielmantione]

What we are looking at here were the business computers of the 80s. People did still love and cherish their home computers after the 80 bit era did end, because they causes so many great moments for people. This wasn't the case with the business computers; once companies bought better computers, the old ones were turned into trash. There was no one trying to preserve the documentation because nobody had any love for those big machines. Especially proprietary hardware was simply erased from history. It's kind of a problem: When in say, 200 years people dive computer history, they will find a treasure chest of well preserved 8-bit computers and enough video on KZbin to spend years of time. However, the 80s were also the great decade of mainframes. Next to nothing has been preserved and this will result in a wrong picture of history.

[@jamescronin7742]

Certainly in the UK in the 80s qe have various drives that used their own interface, but also had a add on board to allow it to be hooked upto SASI. I wouldn't be surprised if that board does the same, ie translates the drives own proprietary interface to SASI or SCSI. Ive seen various systems use 40 pin cables for SASI connections to the host computer.

[@qviewq2071]

We saw a lot of Adaptec / Omti / Wester Digital / Control Data / Fujitsu / IMI - The controllers always seemed to be the most reliable part of any system.

[@tony359]

Another amazing video! I so much love those almost unknown things you show here!

[@lemon3rd800]

Not disappointing at all; you made a lot of progress finding possible errors or failures and open spots.

[@enzofitzhume7320]

Have you checked for continuity on the coils to the spindle motor?

[@machdaddy6451]

I worked at 3M Camarillo, where for a while we made an eight inch hard drive. This is not it, however it is similar in dimensions. Their capacities were twenty and sixty megabytes. The motor was a brushless motor. When the project was abandoed, I was able to get one.

[@cgourin]

If I'm not mistaking the hour counter is interesting being somewhat solid state and reversible if you invert the current. I can't remember the details but I believe a liquid metal moves slowly in a glass tube. Nice design.

[@zaprodk]

It's a column of mercury with a drop of some liquid in the middle that is pushed along as the mercury migrates across. Yes reversing the current makes it run the other way :)

[@cgourin]

@@zaprodk Thank you, I can't think of a more elegant and simple design to this simple important task.

[@indirektdjs]

With the serial number starting with CD would that suggest that this might be built under license from Control Data corporation? Maybe it would be substantially similar to one of their models…

[@donaldcongdon9095]

Or a clone of a CDC model.

[@rajatanpacelana]

Merry Christmas Adrian and everybody that is watching this channel....

[@michaelkelly1251]

I have a power supply that came out of an IBM digital computer that mostly used 74181 alu's an TTl logic chips. The computer was working when I pulled that power supply out around 1993-40 .. It had 5 10 inch floppy drives in it and was basically a data entry type computer used to record and save bowler info and score in the bowling center I worked in at the time. I had all the stepper motors and such at one point but sadly had to let go of those for lack of space. The unit is about 12x5x5 bulging at the transformer area with wires. Has the old TO case regulators would you be interested in it?

[@HudsonGTV]

That molex connector looks familiar. I think it's a standard pinout. I have seen it used to connect to 8" floppy drives. I think there is 24v on one of those wires.

[@scsirob]

Some of those bit patterns in the 12E EPROM might be a discrete state machine. There's quite a few 74LS273 latches there, it might be a discrete CPU-like circuit. I used to work for Intergraph who used huge boards with similar technology for graphic and disk processors.

[@wlbrobinson]

I worked for intergraph in the 80's and 90's, fujitsu made a drive like this and so did CDC. These were the standard replacements for the much larger CDC washing machine size drive, in 1982 they were 80 MB in 85 they were 300MB and the units you have in the video would fit side by side in a 19" standard rack, 300 mb each. It was a big jump in tech as a small room could hold a pretty powerful computer for the time.

[@williamgraham2468]

I built a homebrew i8085-based system. The instruction set is the same as the i8080A, with two additions: RIM and SIM (read/set interrupt mask). ROMs can also be used to make a state machine. Do some of the data lines wrap back to the address lines? It might be decoding a raw bit stream from the disk into data bytes. See Woz's Disk ][ interface design for example.

[@forbiddenera]

@12:25 is you assuming its is ascii? It could be a different character set though most likely it's going to just be binary instructions with no text anyway.

[@exidy-yt]

This huge clunker reminds me of the 20mb SCSI drive I bought for my Amiga 500 in the mid 80s. It was in it's own big box like Adrian's, connected by a ribbon to a sidecar that attached to the Zorro slot on the side, unlike the Commodore or GVC brands that were card and drive in the sidecar. The heads were driven by a stepper motor not a voice coil, and it was loud as a vacuum cleaner and typewriter combined when running, you could hear the beast 2 rooms away, and there was some kind of conflict where if you installed the boot ROM into the sidecar, it took over 30 seconds of gray screen before the thing would begin to boot. It was faster to boot Kickstart from floppy and use the HD purely as storage. Those were the days, my friends.

[@tubical71]

my first HD was a MFM 20MB drive i had for my A2000 came with a proprietary PC MFM controller which then was hooked up to another board which had a little contoller and two roms on it to translate the XT bus from that MFM card to the zorroII bus of the Amiga, it also had drivers for the ISA bus as well for the amiga and it also was bootable (!!)and it was fast back than.i measured 2.8MB/sec that was about trippled compared to what i could read on the PC side using a A2286 and also an AT computer.....and it was about 500bucks total. i later added another 80MB MFM drive and then i bought a A2091 and finally a GeForce 040/33 with its own scsi....all in all 15grand for a computer system...think about what you wpuld get for 15grand today....those had been the days of geeky hardware used by geeky people to get things up and running....;)

[@exidy-yt]

@@tubical71 I really miss those times, sometime. It was a great time to grow up loving tech.

[@TheRealUpuout]

Really Intresting, hope you find a Way to read the Data on the HD. I need to know whats on it.

[@ForgottenMachines]

Awesome video, as always! 27:33 There's a 4th option, I'm certain has already been discussed. Just sticky bearings, that often can be freed up by giving the chassis of the drive a quick and sudden "spin" by hand, in the direction perpendicular to the drive motor axle. Of course, this is easy with a 5.25" drive, much harder with an 8", but there's always a way! Has this been tried yet? Well, moving ahead to 30:38 I can see that you've ruled this out, as you've exposed part of the drive motor, and you could just spin it by hand there and see if it moves freely...and you've determined that the BLDC circuit is not even trying to spin the drive...anyway... LOVE this stuff, Adrian...we'll get the flux transitions off of this device yet for you!!!

[@ntsecrets]

I wonder if that weird rom has a map of good/bad sectors like the other drives of that time would have on a printout?

[@mrfrenzy.]

It says in the manual that bad sectors are stored on each track on the disk itself, and also available on paper printout. The eeprom is most like used to select different logic chips depending on which adress lines are selected. The data lines would be connected to select lines or similar on the many logic chips.

[@8bitwiz_]

Given how expensive the drive must have been, it's certainly possible that they used an EPROM instead of merely printing a bad track list on a sticker. But it's on the "adapter" board, when I would expect it to be on the main drive board.

[@bruceheadley7191]

I should add the 8085 uses a time-division multiplexed address/data bus. The 'standard' parts like ROMs require demultiplexing the ADn signals with ALE and a latch chip (Intel 8282 or 74xx373 or 74xx573) for address A0-A7. 8155 has an active low chip select. 8156 has an active high chip select.