IBM (Zenith) Thinkpad 300

[solidpro]

Hi

I've got a 300 which won't power up. I do, amazingly have a companion Zenith Z Note 433 which is the same machine but in a different form factor - but it does share the same obscure PSU.

The Zenith machine boots. It also has a removable top which contains what looks like 3 little ram boards.

So - the 300 won't power on at all. I've disassembled and there is nothing obvious, although I've yet to actually start tracing the PCBs. I'm curious as to whether it's something simple. The 300 also has a removable plastic panel above the top row of the keyboard and underneath it is a square slot for a maths co-processor and 3 empty slots:

http://toughbookparts.co.uk/IMG_4184.jpg

Does anyone know what these slots do and whether they're necessary for the machine to actually boot up?

Thanks

[RealBlackStuff]

According to the 300 HMM (tpvol1.pdf from HMM link above), they are 3 Memory Module Connectors.

[solidpro]

Ah, as I suspected (any BTW the 3 memory connectors in the Zenith are a different shape to the 3 in the 300)...

Are they necessary for the machine to power up or is it dead for a different reason?

Or to ask in another way, does the 300 have on-board memory...?

Thanks

[ThinkDan]

My ThinkPad 300 also fails to boot. Got further with the battery inserted, but still no POST. Disconnected power supply as it was becoming rather warm behind the screen bezel on the left of the display. Even when current they had an awful failure rate, including significant numbers being DOA from the factory. Perhaps failed 300's make for a more accurate historical record than working ones... which will fail. Soon. Maybe already. Check yours again now. And now.

As an aside, I'd had mine for about 8 years before I used a bit of lateral thinking and went looking for a Zenith 386 laptop power supply. I picked up a power supply model 232-423 (intended for a Zenith Note 325L) quite cheaply; it seems the Note 325L was a standard issue laptop for BT engineers in the UK (telephone monopoly) and quite a few were still kicking around. I can only assume that the Note 325L was more reliable than the ThinkPad 300...



From the HMM, it appears that the system has 3 memory slots and no onboard memory. Both the HMM and the Announcement Letter are somewhat confusing about RAM options; standard RAM is 4MB but you could only order FRU's or options for 2MB ($245) or 8MB ($795) modules. I presume standard configuration was 2x2MB modules and one slot free.

Oh, the standard unit was mono, with an 8" TFT upgrade available to order (presumably for fitment by a service engineer); base laptop price was $2375, the TFT upgrade part (lid assembly and system board) cost £3045 - plus fitting.



ThinkWiki.de has a useful summary and pictures, but they incorrectly talk about a grey retail model and a black business model. There is no evidence in the Announcement or HMM of alternative part numbers for colour options or case parts, as there were for the contemporary grey or black ThinkPad 700C.

[solidpro]

as it was becoming rather warm behind the screen bezel on the left of the display.

Did you feel the heat on the left side of your face?

I've got a few of these. The plastic they're made from feels like it came from Петропавловск-Камчатский. I've also got that matching Zenith 325L, which does miraculously work. I dare not even breathe near it for fear of it giving up. I think they're terrific examples of IBM's path and how the weak, ugly, crap examples of the 'journey' to the 755 or the T43 or the R61 or whatever, happened and were allowed to die. Along with Zenith themselves.

the TFT upgrade part (lid assembly and system board) cost £3045 - plus fitting.

Just for the part?? You had to buy the $2375 laptop first?

A 'business' 300? Isn't that the 700? Sounds like someone who doesn't know about thinkpads got something mixed up.

Just as an aside, when I was collecting (read 'hording') Toughbooks, most of the CF-27 machines I got my hands on had BT branding and a BT bios password. Which i've now forgotten.

[ThinkDan]

the TFT upgrade part (lid assembly and system board) cost £3045 - plus fitting.

Just for the part?? You had to buy the $2375 laptop first?

Yeah, that's the thing, $5420 in parts and then the service engineer's fee to swap the system board and screen/lid assembly. Close to $6000 all in? Can't find the launch price of a 700C at the moment, but circa $5000 probably?

A 'business' 300? Isn't that the 700? Sounds like someone who doesn't know about thinkpads got something mixed up.

Just as an aside, when I was collecting (read 'hording') Toughbooks, most of the CF-27 machines I got my hands on had BT branding and a BT bios password. Which i've now forgotten.

Ah, the thinkwiki.de pages are generally good but you can see here they've relied on anecdote or mis-recollections. It's easy to do at this distance

For the BT laptops, have you tried 'esmeralda' or 'ology' on those laptops? Check here or here for some password reminders...

[solidpro]

It was a word I don't remember but something to do with mythology or astrology or something. Crimson or Starfield or something like that. I will have written it down somewhere.

[ThinkDan]

Does anyone know what these slots do and whether they're necessary for the machine to actually boot up?

According to the HMM (tpvol1.pdf), the 300 doesn't do POST beep errors, so no audible indication of a RAM failure if the display is also duff. If it completed POST and the display was working, an error 200308 would indicate RAM card problems, seemingly even with no RAM fitted or functioning. Oddly 200301-7 are listed in the HMM as 'memory error' but with actions relating to BIOS update and/or system board replacement, i.e. nothing to do with what's in the RAM slots. Further to that, most miscellaneous 'does not power on' symptoms relate to the "video / power board".

I'm guessing you've been through this, but mention it just for completeness...

[LaptopRetrospective]

ThinkWiki.de has a useful summary and pictures, but they incorrectly talk about a grey retail model and a black business model. There is no evidence in the Announcement or HMM of alternative part numbers for colour options or case parts, as there were for the contemporary grey or black ThinkPad 700C.

A Charcoal grey option was available for the ThinkPad 300 monochrome, I suspect that is what they are talking about. You can see the reference to it in the Withdrawn Book.

https://psref.lenovo.com/syspool/Sys/PD ... twbook.pdf

[ThinkDan]

A Charcoal grey option was available for the ThinkPad 300 monochrome, I suspect that is what they are talking about. You can see the reference to it in the Withdrawn Book.

https://psref.lenovo.com/syspool/Sys/PD ... twbook.pdf

Oh yes indeed charcoal grey, but what I was getting at was that there doesn't appear to have been two colours of casing - grey vs. black. The announcement letter states nothing about colour, nor indeed a business vs. retail model, and if there had been two distinct colours then the HMM should have listed these, as it did for the lighter colours of 700/720/750 - where duplicates of all external plastics are listed "for germany" with distinct part numbers.

If you look at the 300 photos at the thinkwiki.de page, the lighting is clearly very different between the grey and "black" models. I think it's a great site and refer to it frequently as it adds to or refreshes much of my recollection of the older models; look at it another way, if this is all they've got wrong then they've done a bloody good job!

I'll email them about this and see what they say: either I'll learn something new about the German models, or it'll help improve the information available

[ThinkDan]

Tracked down the US list price at launch for a 700C as $4825 for 10" screen, 486 CPU, and MCA bus, compared with $5420 (plus fitting) for the upgraded 300C, with 8" screen, 386 CPU, and ISA bus. Oh, and the 300C still didn't have a TrackPoint

[goldeneagle]

My ThinkPad 300 also fails to boot. Got further with the battery inserted, but still no POST. Disconnected power supply as it was becoming rather warm behind the screen bezel on the left of the display. Even when current they had an awful failure rate, including significant numbers being DOA from the factory. Perhaps failed 300's make for a more accurate historical record than working ones... which will fail. Soon. Maybe already. Check yours again now. And now.

I have three of these, plus the companion Zenith as well. The one I've had since about 1998 - the LCD went bad on it. I think I found an LCD on ebay a few years ago, but never got around to testing/replacing it. I've noticed the LCD/polarizers on these go bad fairly quickly.

For the curious - this max out at 12MB of RAM, and the ethernet port on the side will work with an Apple AUI (AAUI) adapter. It's recognized natively in WFW 3.11.

From the HMM, it appears that the system has 3 memory slots and no onboard memory. Both the HMM and the Announcement Letter are somewhat confusing about RAM options; standard RAM is 4MB but you could only order FRU's or options for 2MB ($245) or 8MB ($795) modules. I presume standard configuration was 2x2MB modules and one slot free.

Oh, the standard unit was mono, with an 8" TFT upgrade available to order (presumably for fitment by a service engineer); base laptop price was $2375, the TFT upgrade part (lid assembly and system board) cost £3045 - plus fitting.




If that's your system, I'm TOTALLY jealous. Always wanted to find a 300C, but wondered if they even existed.

I also didn't realize who you were at first; I downloaded software from your website about 20 years ago for my 730TE (that I still have). Now if I could only find battery packs for the 730TE. Actually, I'd like to rebuild at least one of my packs for the 300 as well, when I show it off next year at VCF Southwest.

[goldeneagle]

Next up will be diagnosing LCD flickering/contrast. I have a suspicion it's the same scenario as my 720C.
I don't remember if the Zenith powered up or not (only found a Zenith AC adapter ironically), but three of the 300's fire up, though the screens look like crap (but still work).

Forgot if I answered it before, but these come with 2MB soldered onto the system board. Mine that I've had since about 1996 or 1997 has the 386SL coprocessor and 12MB of RAM. I was surprised that the 340MB hard drive I put in still works (but that's getting replaced with a CF card).

[ThinkDan]

Forgot if I answered it before, but these come with 2MB soldered onto the system board. Mine that I've had since about 1996 or 1997 has the 386SL coprocessor and 12MB of RAM.

Thanks, that's useful to know as I couldn't find any reference to soldered RAM in the HMM, the logical inference being that all RAM was installed in the sockets under the cover.

[ThinkDan]

Reading back over this thread, the fact that the HMM recommends system board replacement for RAM errors ties in with Goldeneagle's recollection of RAM soldered to the motherboard.

I was photographing my ThinkPad 300 today ahead of putting it up for sale, and thought more pictures might make a good addition to this thread.

Here it is beside a 700C, showing the clear differences in case colour, style, and detail. Note how some of the 700 design details are carried over, such as the rolled edge at the front of the keyboard, screen sliders and lid latches, etc. Really this is just dressing up a generic laptop to look like an IBM design. Bear in mind that these common design details were also inherited by the 700C from the prior PS/2 Note and PS/Note models, such as the N33, N51, 425SL, and in this respect the Zenith-made 300 is as much a tribute to the PS/2 Note designs as it was to the then brand new ThinkPad 700. The keyboard is completely different to the PS/Note or 700 layouts, and it has a different typing feel.



From memory, the only other IBM notebook that I've encountered in this slate grey colour was the PS/1 Note 2141-182. It's nothing like the cream of the German 700/720, nor the grey of the German 750. It's also not a flat uniform colour, but more like what would be termed 'grey marl' for a T shirt:



General views:


Panel at the top right of the keyboard covers the 387 co-pro socket and RAM sockets:


VGA, port replicator connector, PS/2 mouse, parallel, serial, power ports on rear:


I never did see a port replicator for one of these, and Google brings up nothing now, but I did manage to find an image of a Zenith 325L with what looks like a port replicator fitted... a 2" deep slice that presumably replicates at least parallel, serial, vga, power, and mouse ports for quick desktop docking:


Battery latch and slide-out battery on left side, plus lid latch:


Nothing on the front, battery and PC speaker (beep) visible on underside:


Lid latch, 1.44MB Floppy drive, panel covering modem module socket with removable blank for modem connector (modem not present), uncovered LAN port (as goldeneagle referred to above, connecting with an Apple AUI LAN Adapter):


ID sticker on the underside is very non-standard, but in classic IBM straightforwardness it is "Manufactured for IBM" not "Manufactured by IBM"; "Shimane, Japan" would have been Zenith's manufacturing plant:


Fujitsu owns a large PC manufacturing plant in Shimane that opened in October 1990, initially for manufacturing desktop PCs but more latterly manufacturing just PC laptops, according to that 2008 press release. I wonder if this was the plant that made the 300?

The type-model-serial number sticker at the bottom is pre-printed up to to the 55 of the serial number, with the real serial number 24681 overprinted in a slightly different font. 55- indicates Greenock, UK as plant of manufacture. This is a UK model, so would probably have been processed through Greenock even though originating from Zenith.... probably checked then boxed up with destination country appropriate manuals, power cord & power supply, etc. This happened on "made for IBM" products, such as the PS/2 wedge mouse made by Logitech but branded with a 23- serial number suggesting it was manufactured in Raleigh, USA.

With the Zenith power supply mentioned above:


Enjoy!

[solidpro]

It's a crying shame IBM didn't subcontract out all manufacturing and development of the Thinkpad at this point to the team behind the 300. I see the 300 and I see the beginning of the machine we could have had.

[ThinkDan]

Irony?!

[solidpro]

Trying to start a heated debate based on my castle full of sand!

[ThinkDan]

Phew. You had me worried there for a moment!

There's a fascinating and detailed group test of 31 contemporary 386 and 486 laptops in the March 1993 issue of PC Magazine, starting on page 144. The ThinkPads 300, 700, and 700C are all reviewed in there. It's interesting to see how the price, spec, and general physical design compare. I looked it up the other day thinking that the 300 was overpriced for a 386 mono laptop in 1992/93, but actually it was in the same price range as contemporaries of the same spec.

[solidpro]

I had totally forgotten or not appreciated the 300 was a 386 and the 700/700C was a 486. A world apart.

And so now I am getting confused. The thinkpad I got on ebay last week (which may or may not go anywhere) has an i387, which, as far as I know was a maths co-processor for the 386. If only the 700T came with a 386 CPU, why is there an i387 in this 486 Thinkpad 700C?

Also, I've only just realised, I only own 700 machines so a 700C is actually quite important to my collection!

**edit**. Oh, hang on, my geeky 13 year old brain seems to remember a 486SL is a 486 without a maths co-processor and a DX has one?....

[astral]

700C uses a Cyrix 486SLC (IBM-fabbed) that is basically a 386 with 486 instructions. Would make sense as such that it still uses a 387. 486SLC performs more similar to a 386 than a true 486. from what I remember, 486SL is a mobile-optimized full 486. 700C doesn't use it though.

[ThinkDan]

Astral - I think IBM later acquired the rights to use/develop Cyrix designs, and those subsequent chips differed from the 486slc in the 700/720. There were also later IBM 486 designs, known as Blue Lightning, but I'm really vague on those. Sometimes they were installed as whole new motherboards, e.g. for some 386-era PS/2 systems, sometimes they were pin-compatible upgrades to 486 systems.

solidpro - are you sure you want to go down this rabbit hole?! My brain hurts just thinking back to this time of (in)compatibility and highly optimistic promises of next-gen performance in last-gen hardware. I tried reading the Wikipedia pages to make sure I wasn't talking rubbish and had to go and lie down. I'm back now

OK, IBM licensed the 386SX chip design from intel, to produce their own 386SX chips - IBM of course had huge chip fabrication capabilities then. They were permitted to develop the design and added in extra features - particularly power-management and cache, which became the IBM 386slc. Due to the cache this performed broadly equivalent to a 386DX of the same clock speed, but cheaper; the power management meant it was also well suited to laptop and notebook designs. The IBM 486slc was the IBM 386slc to which IBM then added 486 extensions and possibly more cache. It was limited by its 386SX origins to 16MB RAM addressability, and similarly had no floating point unit, so used a 387 maths co-pro. Thus the ThinkPad 700 & 720 are technically 486 systems, in that they will run 32-bit software compiled for 486 processors, although architecturally they were more akin to 386 processors / motherboards.

In the 386 world, SX = 16MB max RAM and 16-bit data bus crippling, DX = max 4GB RAM and full 32-bit data bus. SX was therefore less good at large scale data throughput. All were 32 bit chips (instruction set / architecture), neither SX nor DX had floating point unit (FPU) - hence 387 co-pro option (if dedicated socket provided). Can't remember if they had any/much cache. IIRC intel later added a 386SL which was their own power-managed SX variant.

In the 486 world, SX = no FPU, DX = FPU. All were 32 bit architecture with 32-bit data bus and 4GB RAM addressability. Extra instructions ("486 instructions") were added, architectural improvements increased performance / throughput, and cache became standard (although variable to further differentiate SX & DX, IIRC). The 486SL was a low-power 486DX, confusingly. DX2 etc. came along later as clock speed multipliers, usually with lower power too.

intel offered a 487 co-pro that was essentially a 486DX with the pin packaging altered to fit in a co-pro socket if the 486SX motherboard had one. Oh, and then there were the optional OverDrive sockets on later 486 boards that allowed for specially packaged Pentium chips to be fitted to a 486 motherboard. I'm sure there were also interposers that were needed in some instances to allow certain upgrade chips to fit in sockets on earlier boards... it was quite messy.

(One job I enjoyed whilst with IBM was to document the 386, 486, and early Pentium systems' upgrade paths for the HelpCentre, including OverDrive capabilities but also RAM, disks, video RAM, etc.. We were getting a lot of calls about OverDrive in particular, since intel were heavily marketing it. Very little in the way of laptop upgrades due to absence of additional sockets bar 387s, but for the PS/1, Aptiva, and ValuePoint lines there were a range of possibilities, cludges, and limitations. IIRC, the later PS/2 desktop systems (76,77) might have had some upgrade path too. Somewhere I still have the results of all of this work, probably on a fading Zip100 or 1.44MB floppy disk, used at the time as internal support documents in OS/2 .INF compiled help file format - totally different to Windows driver INF text file format, of course!)

For extra brain frying, now go and read up on the Cyrix, Texas Instruments, and AMD 486 and Pentium compatible chips. Wasn't there something called a WinChip too, in a sort of "486 plus pentium instructions but minimal cache" kind of model?

[solidpro]

One job I enjoyed whilst with IBM was to document the 386, 486, and early Pentium systems' upgrade paths for the HelpCentre

I think this might be some insight into how your brain works.

I remember having a P60 Pentium (which was a faulty P75) as my first desktop computer and forever pouring over the possibility of Overdriving it up to something like P133. It seemed quite a minefield at the time as to what would work, what wouldn't and so it never happened. I ended up buying an IBM Thinkpad 365 with something like a P120 (and there my love of Thinkpads begun!) and the P60 was left in the dust by a tiny black bento box.

All of this does come flooding back about SL, SLC, SX, DX and so on but you have helped re-cement it in my mind. Thanks!

I did also own a Cyrix 5x86 (I think it was badged) at some point. Never had an AMD until the Athlon started whipping Pentium 3s (under a huge heatsink and fan!).

[ThinkDan]

One job I enjoyed whilst with IBM was to document the 386, 486, and early Pentium systems' upgrade paths for the HelpCentre

I think this might be some insight into how your brain works.

Thanks

The information was scattered around user guides, Technical Reference Guides, Hardware Maintenance Manuals, and anecdote, and resulted in a lot of unnecessary call escalations and call-backs. Remember: no google. Ended up talking to the product specialists in Greenock and Raleigh, also had a contact at intel UK - I remember he sent in a whole big bag of 'intel overdrive' keyrings with chips embedded which I gave out to the HelpCentre agents.

The satisfaction was in pulling it all together into one quick "I'll check while you're on the phone" reference, learning to use a new tool (IBM HyperWise) and markup language/system, and setting it up as a hyperlinked document that compiled into a native OS/2 help file... and getting paid for having all this fun!

Given that HTML's origins were heavily influenced by IBM's markup language(s), it was only a small step then to do real HTML in 1998 when I started my PC110 webpage, initially in a text editor but eventually using the WYSIWYG editor in the Netscape Communicator package.

[astral]

Astral - I think IBM later acquired the rights to use/develop Cyrix designs, and those subsequent chips differed from the 486slc in the 700/720. There were also later IBM 486 designs, known as Blue Lightning, but I'm really vague on those. Sometimes they were installed as whole new motherboards, e.g. for some 386-era PS/2 systems, sometimes they were pin-compatible upgrades to 486 systems.

I previously thought it was that Cyrix designed the 486SLC, then IBM and TI produced it as Cyrix didn't own their own fab. I looked it up just now - I was wrong. If wikipedia is to believed here, the Cyrix SLC chips were produced by TI and SGS-Thomson, then IBM ALSO made their OWN chip, based on Intel's core, and also called it the 486SLC... confusing. IBM then made clock doubled and tripled versions (SLC2 and SLC3), then the blue lightning chip was similar but actually ran on a 32-bit bus.

[solidpro]

Netscape Communicator package.

I remember when there was only one browser worth using....

https://preview.redd.it/cqvu8d84wnl61.g ... 6cfd0687a3

[ThinkDan]

Yeah, it's so difficult to pick through the competing and overlapping 386, 486, Pentium compatible designs to make sense of it. This page gives a quick snapshot of how convoluted it got, and this page gives as compelling a description of Blue Lightning variants as I can find.

[ThinkDan]

Found a couple of old pictures of a 300C in my backups...



2nd image appears to show 3x 2MB memory cards installed.

Oh, and my old 300 is back on ebay again.

[axur-delmeria]

Astral - I think IBM later acquired the rights to use/develop Cyrix designs, and those subsequent chips differed from the 486slc in the 700/720. There were also later IBM 486 designs, known as Blue Lightning, but I'm really vague on those. Sometimes they were installed as whole new motherboards, e.g. for some 386-era PS/2 systems, sometimes they were pin-compatible upgrades to 486 systems.

I previously thought it was that Cyrix designed the 486SLC, then IBM and TI produced it as Cyrix didn't own their own fab. I looked it up just now - I was wrong. If wikipedia is to believed here, the Cyrix SLC chips were produced by TI and SGS-Thomson, then IBM ALSO made their OWN chip, based on Intel's core, and also called it the 486SLC... confusing. IBM then made clock doubled and tripled versions (SLC2 and SLC3), then the blue lightning chip was similar but actually ran on a 32-bit bus.

Here's what I know: IBM has an x86 license from Intel that allows it to use and modify Intel designs. The IBM 386SLC is the first one, based on an Intel 386SX but with 8KB internal cache. Because of the cache, a 386SLC@25MHz can keep up with an Intel 386DX@33MHz. Do note that the IBM 386SLC is not pin-compatible with the 386SX and thus cannot be used as a drop-in replacement. Furthermore, the terms of the licensing deal prevents IBM from selling bare processors on the market, this means it's only available as part of a motherboard, or sometimes as an upgrade kit (small daughterboard that plugs into the CPU socket).

The IBM 486SLC and 486SLC2 are minor improvements of the earlier 386SLC (still stuck to 16-bit data bus and 24-bit address bus), with 16KB internal cache and clock doubling.

The IBM 486BL Blue Lightning is based on an Intel 386DX with full 32-bit data and address bus, but with 16KB write-back cache and clock tripling. Note that these IBM designs required an external FPU (math coprocessor) with made it much slower than the 486DX in this regard. OTOH most software at that time only used integer math and thus didn't matter much.

Cyrix had its own design, but the name similar Cx486SLC and thus causes confusion. To muddy the waters even more, Cyrix and IBM later got into a manufacturing deal that allowed IBM to sell Cyrix processors under the IBM branding. This means any socketed IBM x86 processor is a Cyrix design, while the soldered ones are more likely to be Intel-derived ones.