Infinite loop: the Sinclair ZX Microdrive story

Joining drives

It was only the first drive that connected to the IF1 with a ribbon cable. Subsequent drives were connected by a connector like this: http://www.clive.nl/files/imagecache/detail/images/pict0840.jpg

and a plastic plate on the base of the microdrive was used to secure the drives together.

Still got mine, still working, along with the +D and various drives.

Sinclair QL

Interesting article. Sinclair was approached by floppy makers to use their drive in the QL but stuck with the microdrive. Who knows what might have been but it's still an enjoyable machine with a great community.

Interface 1 & ROM compatability.

The article mentions that the Interface 1 had fixes for bugs in the Spectrum ROM. I seem to remember this caused problems with some programs. I definitely remember seeing code samples in Speccy mags about detecting if an Interface 1 was present so you could take account of the ROM changes.

The microdrive wasn't that bad. I used one for ages and it did me proud.

Combined with the Romantic Robot Multiface One it was a godsend - load up a game, get past the annoying copy protection (ELITE with LensLok I am looking firmly at YOU!) then dump the memory to MD or Floppy and life was all good.

By the time I got my +3 though with 3" disks it got consigned to the bin

Happy times with a Microdrive

I only picked up a Spectrum in 1984, but it was a bumper "sale pack" with an Interface 1 and two Microdrives

Two things made it incredibly useful: a piece of third party software "guaranteed" to format a cartridge to 96k, and Romantic Robot's Multiface 1, which dumped the Spectrum RAM to Microdrive and let you restore it later. For games with no save point, or awful copy protection like Elite's "Lenslok", or just going back a couple of days in "Doomdark's Revenge", it was almost indispensable.

When I finally got rid of the Spectrum, I had over 50 cartridges, and had only two die on me. For all that it was a kludge, the Microdrive served me well.

Does anybody remember the some cool hardware...

- which let me use a real 3.5" floppy drive and could dump the memory contents to it at the press of a button, for virtually instantaneous loading of all my games at the press of a button - was it called 'snapshot'? - advertised in some of the mags at the time.

I didn't dream it - I've still got it in my 'legacy' spares box

Rose tinted hindsight?

All these articles on 80's computers begs the question, was computing more fun then, or has my memory donned rose tinted specs?

I loved my Beeb, it outlasted Atari ST and my first IBM clone, I've lost count of the number of Speccys I fixed, C64s as well (though melted power supplies was their biggest problem), I've had a QL and an Oric and they were all fun in some way shape or form.

Modern reliability is nice, admitted, but somehow modern computing seems boring in comparison to the mid 80's to early 90's. Maybe that's why the Raspberry Pi has taken off?

21st cenutury technology on 20th century kit

The good thing is though, if you've still got your original kit you can still do fun things with them, such as all these CF and SD card adapters and virtual disks. Having a 256Mb SD card in my Beeb on a board that cost less than a tenner to make and having all those free disk images from the 'net on there is wonderful.

What about the offspring?

I think that mention should be made of the offspring of this project, the ICL One Per Desk aka BT Tonto. Basically AFAIR a QL built into an early attempt at en executiev workstation. I used on when I was in BT Marketing, it incorporated a telephone and modem, with two lines. Apart from the normal functions of word processin et al I was able to interrogate a distant mainframe running RAMIS Reporter to obtain the relevant stats.

Infinite Loop.

If the Microdrive cartridges were on an infinite loop, does that mean that Apple were inspired by copied Sinclair with their choice of headquarters address?

OPD

The other machine that used microdrives was the ICL OPD (One Per Desk), a collaborative project between ICL, Sinclair Research and British Telecom

Simon Webb

Museum of Computing

www.museumofcomputing.org.uk

I had one

I had one, and despite it's flaws it was truly revolutionary compared to waiting 5 mins for data (read data as games) to load.

Of course the afformentioned games didn't come on microdrive so with the little Z80 assembler I had I managed to hack a way of loading a game in while avoiding the microdrive memory map, dumping that back to microdrive and then write a small loader program for each game that would load and shift everything to the right memory locations after everything had loaded. I even used the memory area reserved for the video as a buffer for some games - if you ever wanted to see attribute screwup then that was a prime example.

I think fondly of my microdrive. It had it's problems, but it was also lightning fast as far as I was concerned.

Ahh those were the days

This happened to me(*)

I'd built a crude mouse and circuitry to hook it up the the RS-232 interface. Then it took me weeks to write the mouse driver interface, saving it to the Microdrive. Alas the tape failed...

Flappity Floppity Flip

The mouse on the Mobius strip

The strip revolved, the mouse dissolved

in a chronodimensional skip!

(*) well, not really; I just wanted to post this rhyme.

Wafa drive

.... Still got mine somewhere.... And talk about recovering ancient data, the sound of it came back to me reading this article. Sort of 'shuffa-shuffa-shuffa chuck-chuck"

Oh yeah! An my GF at the time insisted on pronouncing it "waffer" drive.

Not just tape that still stores ancient crap eh? Brain 3.7.2 -

Microdrive technology

The article has a photograph of a spooled infinite loop and later describes this. My Microdrive was different. I describe my experience of the two methods of running a continuous loop to provide a bit of history.

The spool :

In 1947 I met continuous display from 16 mm film. A unit was provided that mounted on the projector. This pulled the film from the inside of a coil and fed the film back to the outside of the coil. The axis of the coil was vertical, the film edge rested on rollers so it circulated without much friction. The theory was that pulling the film out from the inside caused little or no friction as the movement was always away from the neighbouring surface. It worked almost perfectly. At an exhibition a film would run for a total of say 50 hours, and be pretty horrible at the end.

The esses :

In 1970 I met more continuous loop of film in microfilm duplication. This threw film into a chamber that contained it and drew film out from the bottom of the chamber. Within the chamber the film forms itself into a stack of esses. We would run a loop of several hundred feet at 120 feet per minute. Of course there was degradation, but not much. Silver emulsion survived several hundred passes, tougher materials were available for intermediates. At computer fairs there were big versions of this on display handling wide mag tape to drive the demos.

In 1984 I found Microdrives would fail after ten accesses, reported way above.

I broke open a carttridge and discovered that the method of recycling the loop was the esses but horizontal. The tape rested on its edge. I conjecture that this was the original design and the spool described in the article was a later version to provide a useable product. Several contributors above speak well of the Microdrive, perhaps their cartridges were version 2 .

Microdrives never contained EPROMs (or ROMs).

The article incorrectly states that the first 1000 or so microdrives had EPROMs. They only had one IC in them and it wasn't a ROM or EPROM. The Interface 1 had a ROM (8 Kbyte?) and I have previously heard that early units had EPROMs rather than masked ROMs.

"..exclusive to Sinclair..."

and that was the problem. With disks you can transfer data to and from other systems, with microdrives your data is locked into a pile of decaying plastic.

Speed benefits when compared to tapes?

I was a kid in Calcutta, India growing up with a Sinclair ZX Spectrum, and I could not get my well meaning father to invest in a microdrive, after he had already plonked some hard earned rupees on a "Compluter friendly" casette recorder, where you could vary the speed of the playback motor, to load programs slightly faster, provided the reading head and the capstan were clean. Can anyone tell me the loading time speed benefits of the micro-drive vis-a-vis std data cassettes in loading up say "Daley Thompson's Decathlon" or "Horace goes Skiing", which btw we all copied from each other using analog tape to tape.

Not revolutionary at all.

I didn't see anything revolutionary about the Microdrive, and I don't think anyone else in the comptuer world did at the time. There were several similar products on the market years before the ZX-80 even existed - such as the Exetron Stringy Floppy. Sinclair didn't invent the printer either; the ZX Printer was a very small, low cost and completely rubbish version of what came before. They did the same with the Microdrive; small, cheap and rubbish yet it seems to have aquired mythical status with those of a certain age. Other tape loop drives worked really well, but by about 1981 they lost any price advantage they once had to cheaper, smaller (5.25") and higher capacity floppy disk units, and the access times of tape loop could not complete.

ICL OPD

I've still got a small box of micro drives for mine.

Can't remember them ever working...

I really struggle to get my head around how the Microdrive infinite tape loop actually WORKED. So Alan Firminger's post above was fascinating. I still haven't quite grasped how the inside of the tape loop lets you draw the tape out without friction. Anyone got a link to a nice diagram?

Mine was pretty damn reliable (and I too remember the joy of having Elite load in 7s without sodding Lenslok). So perhaps I had a v2 also.

An actual 8-track floppy/tape implementation

1977's Compucolor 8001 had an /actual/ 8-track cartridge tape system (which they called "Floppy Tape Memory") available for it. It was capable of storing 1MB (that was a lot of storage back then) which could be read at 4800baud.

Fun Times

I had the pleasure of working for Sinclair Research for almost two years in the 80's firstly commisioning technical software for the QL, then programming on the last Sinclair version of the Spectrum which introduced a screen editor to replace the single key press function entry (ZX128).

This article kicked into action some memory neurons...

I had successfully used the Interface 1 and microdrives on a Spectrum for games development, compiling on a Cromenco CP/M machine and then downloading images onto the Spectrum over the RS232 i/f onto the Spectrum, saving the images off to microdrive. Testing and reloading images from microdrive worked a treat...far better than tape or serial download.

The later version of Spectrum Microdrives worked reasonably well - the real problem for the QL was the capacity/speed increase from 85K to ~110K combined with the lack of a cassette tape interface. This meant that all programs had to be supplied on Microdrive, with the problem that a program written on Microdrive A had a high chance of failing to load on the customers Microdrive B. This was eventually solved but way to late for commercial success.

The really sad thing about the QL was that the software, operating system and Basic were ground breaking for the time but the hardware configuration decisions just let the system down.

Changing the tone...I remember the first action of the engineers on receiving a trial One-per-desk was to make it swear at them.

The really annoying thing was that before the Amstrad sellout, Sinclair engineering developed a product called Tyche, QL with more memory and integrated floppy disk which would have been a fine competitor to the Atari ST and BBC Micro but never saw the light of day.

-- Andrew

Teesside Poly pimped the QL

In 1985 every student on the HND Information Systems at Teesside Poly were loaned a QL and matching 14" crt monitor for the two year course. I still have the welts on my hands from carrying them the to the train station.

Assignments were to be brought in on the damn microdrive tapes. If I'd have been a bit more savvy I would have realised that I didn't have to do the assignments, just blame the tapes. I ended up doing the assignments and invariably losing them.

Our machine code lecturer was a big cheese in the 68008 world, he worked part time for Motorola (more likely a research contract). Seems odd to have someone genuinely good at such a 'ahem' 2nd tier 'university'. Maybe he was a degenerate.

MGT disciple +d Ingo Truppel MB-02+ DMA 3½" hi/extended density?

no mention of the MGT disciple +D or the MB-02+ from Ingo Truppel with hi density and DMA

1.8MB on a 3½"floppy

what happens when someone adapts an IDE to compact flash and twin SD to fit inside the microdrive?

rom reprogram?

amazing it didn't have 16kb of ram in the rom area - ideal for bbc basic etc