Vintage IBM 5155 Portable Computer Teardown - The Electronics Inside

[@matthewfeurtado8921]

The 50 hz and 60hz lines on the floppy spindle are for use with a standard incandescent light bulb. Shine a lamp on it and the appropriate AC markings will stand still if the motor is at the correct speed

[@amirpourghoureiyan1637]

They used standard Shugart drives that were able to be used as separate AC devices (Commodore 1541), the meter isn't helpful if the drive is hooked up to DC like these internal drives are.

[@jankkhvej]

<a href="#" class="seekto" data-time="660">11:00</a> That socket is for FPU (Floating Point Unit), Intel 8087 or compatible (there was faster ones from other manufacturers). And ALU you mentioned is just a part of any CPU.

[@jankkhvej]

<a href="#" class="seekto" data-time="999">16:39</a> It's Async Card, clearly marked, with 8250 Serial I/O controller, also there is a DB-25 male serial connector, this is classic COM port.

[@matthewfeurtado8921]

The 5155 is based on the 5160 otherwise known as the XT. The 5150 is the original model and has only 5 ISA slots and a cassette port while the 5160 has 8 ISA slots and dropped the cassette.

[@richgolfs]

I don’t recall a cassette port on the IBM PC XTs. Also you need a fluorescent light to be able to cause a strobe affect on the disk drive so you can check the speed. Incandescence do not strobe.

[@martinleist6331]

I had one of these as my first IBM PC back in the 80's. I worked on IBM mainframe systems and bought one through our IBM rep who I was friendly with (they got a healthy discount) for about £800. Originally had two floppy drives but I got one swapped out for a 20Mb hard disk - massive in those days. The power switch on the PSU was exactly the same as the ones on the mainframe peripherals. Used it for a few years and wish I'd kept it now.

[@jankkhvej]

<a href="#" class="seekto" data-time="525">8:45</a> it's not for the tachometer, it's for naked eye, you need a lamp, preferably neon, which blinks at AC power frequency (hence the dual markings for 50 and 60 Hz) so the pattern will stay steady, many vinyl record players have same device. Imagine - you can fine-tune floppy drive with just your naked eye.

[@jankkhvej]

<a href="#" class="seekto" data-time="264">4:24</a> it's not a PS/2 connector, it's PC, PC/XT, PC/AT 5 pin DIN connector.

[@smartassist9700]

FORMER IBM’r here. I used that at work in the beginning!

[@richgolfs]

A couple of things. For one, IA stands for industry standard architecture not for industry standard adapter. Second one filling around inside of a PC with a CERT and it always discharge the CERT they can build up a charge overtime. Even when just sitting there. The discharging of the CERT was told us by our electronics teacher in college who had a TV repair class. Safety. 40,000 V and that’ll knock you off your seat.

[@RottnRobbie]

I was willing to let "IA" slide as a simple typo on "ISA", but then you typed "CERT" - twice. I had no idea what a "CERT" is in this context, so I had to think about it. I considered re-watching the video to see what you were referring to, but eventually realized you meant the *C*athode *R*ay *T*ube that David referred to multiple times by the extremely common acronym "CRT". And at that point, you lost a lot of credibility regarding correcting others on the meaning of acronyms...

[@robertpowell7319]

The empty socket next to the CPU is for the 8087 math coprocessor. When installed it sped up all of the floating-point. My old CAD software just flew with the math coprocessor installed.

[@muffenme]

I pause the video at <a href="#" class="seekto" data-time="152">2:32</a> to guest at each port. From left to right, 5 pin din keyboard connector, 25 pin com port, external SCSI, parallel port, and 9 pin mono/CGA port. That keyboard port isn't PS/2. PS/2 are 6 pins.

[@amirpourghoureiyan1637]

That's an external floppy port, not SCSI. It's the same card from the 5150. The 5 pin connector used the same protocol as the 5150 and 5160, this was called the XT keyboard port.

[@richgolfs]

Definitely would not be a PS2 connector. PS two connectors came out with PS twos. They were not around for the PC series of computers by IBM or anybody else. No they did not come with SCSI connectors. In fact the entire PC had internal or external SCSI connectors.

[@muffenme]

@@amirpourghoureiyan1637 It does look like one, I was just guesting.

[@muffenme]

@@richgolfs I was just guesting without much info.

[@stank_tater]

the first luggable computer wasn't the xerox notetaker it was the ibm 5100. it came out 3 years earlier

[@mechanicbot2834]

That isn’t a PS/2 port, it’s an AT keyboard port. The PS/2 port came out with the IBM Personal System 2, which released in 1987, while this system released in 1984.

About the double stacked chips, these are OKI37564 chips. There built as double stacked ICs by design! (I had never seen that before). Maybe each chip handles a different part of the bits?.

[@dirething]

I do not know about that unit specifically, but in designed for stacking elements like that the simplest thing is to invert the enable on one of them internally

Maybe the chip on socket is a PAL (or a GAL). These were programmed externally and then plugged to the board. They were used on some motherboards as part of the bus control circuitry to avoid having custom chips (and to allow upgrades in a way).

[@jankkhvej]

<a href="#" class="seekto" data-time="1067">17:47</a> And this card is have Parallel Port, DB-25 female socket, and clearly no serial controller is present on card, just some usual buffers and latches with address decoding logic.

[@dykodesigns]

The aftermarket floppy drive looks like a Panasonic, judging by the design of the bezel. I was expecting the orginal drive to be a YE-DATA (IBM used those with a characteristic bezel) on several models. The chip stacking must have increased the manufacturing cost signifcantly, but I guess it wasn't much of a problem for the typical IBM consumer, since this computer was kind of aimed at the executive buisness suite wearing types. Kaypro also had a luggable computer, around the same era, but that one used CP/M.

[@amirpourghoureiyan1637]

Annoying that the previous owner cut into the case to make it fit instead of just buying a different drive.

[@jafirelkurd]

At the time this system was released, IBM was still using full height MPI or Tandon drives on the PC and XT. It wasn’t until a few years later that IBM used half height drives on XT. The half height drives in these are the ones from the PC Jr if I recall correctly.

[@smartassist9700]

Very funny. Those were the days!

[@DaxVJacobson]

It's not a PS/2 keyboard that won't hit the market till 1987 (with the IBM PS/2) (This 5155 computer debuted in 1984, the year the Mac came out) I think it's just called the keyboard port

[@NiHaoMike64]

Perhaps replace the broken PSU with a common 12V module and some DC/DC converters for the other voltages? An interesting idea would be to make a card that plugs into one of the slots and interfaces a Raspberry Pi to it. Would make it quite a bit more useful without detracting too much from the vintage look and feel, and can be easily removed to return it to its original condition.

[@jafirelkurd]

That async card is actually a serial card, not a parallel card. In normal PC convention male 9 or 25 pin ports are serial and parallel are female 25 pin ports. Not every female 25 is parallel though, because there is also SCSI that used that and probably others.

[@robertpowell7319]

The empty socket was for some type of memory addressing chip. One could insert the chip, un-solder all the 16K memory chips and replace them with 64K memory chips.... at a great expense of time AND money. :-S

[@Chriva]

The stacked chips is just to increase capacity. They did the same on some old edo (and presumably other 72-pin and earlier 30-pin modules), sdram and ddr1 sdram modules :) I don't know the exact mechanism here but the newer modules use two slightly different chips as to offset their working address range or by having row and col selection on different pins.

[@user-qf6yt3id3w]

I could see stacking SRAM chips if you had some that needed and active low strobe and some that needed an active high one. That way you could switch between two banks - you'd always get data from one of the other. However those chips are Oki MSM3764A which are pin compatible with the classic 4164 64k*1 bit DRAM. Asynchronous DRAMs are hard to drive using 80s technology because you need to present a row address, drive /RAS active low, present a colum address and then drive /CAS active low and I can find no evidence that there was an analogous scheme where some DRAM chips used active low /RAS and /CAS and some used active high RAS and CAS. It's a slightly mindbending idea. The only way I can rationalize it is that the IBM engineers needed to use two chips in parallel to drive the bus signals faster something. Even for IBM that seems very wasteful - DRAM chips were a significant percentage of the board bill of materials cost.

[@Chriva]

@@user-qf6yt3id3w count the number of chips, then check ram size of these machines. I assure you it's for memory size reasons :)

[@user-qf6yt3id3w]

@@Chriva Look up the datasheet for the MSM3764 and explain how putting two chips on top of each other would double capacity, even if you had some 'active high' variant which a) I can't find any evidence of and b) I'm not sure is even possible to make. You can see just by looking at them that both chips share the same /RAS line.

[@Chriva]

@@user-qf6yt3id3w You are the one that is stuck on different active states. How do you know what the chip under the labeled one is? Count the number of chips and explain to my why the machine has 512K of ram

[@user-qf6yt3id3w]

@@Chriva I'm just saying I've got no idea how the board works. Which means there's something interesting going on with it.

[@smartassist9700]

Correct on your history!

[@edwingonzalez700]

Z