Timeline of double sided floppy disk drives

[per]

I had a talk with Erik Solhjell over the easter vacation, and during the conversation he mentioned he was involved in implementing the first use of the double-sided floppy disk drive in Europe through his work at Tandberg Data. I know from the Tandberg Data newsletters the machine in question containing the double-sided disk drive was released in April 1980, but given that double-sided floppy disk drives were quite widespread already in the early-mid the 80s, I thought this might have been a bit late for the introduction of that whole technology.

Now there is always the chance that things were widely adapted quickly, across all floppy-disk sizes of the time, after a working solution was introduced on one of the form factors. I was as such a bit curious if some of you know any more details about the timeline on how this all went down. I.e. when double-sided drives were available were first demonstrated for the different disk sizes, and when they were finally available in end-products on the open market.

Another bit of trivia he mentioned, according to employees he talked with at IBM back in the day, he claimed that internally within IBM the use of 8" floppies to replace punched cards in general predated the use for 360/370-firmware updating, arguing the secretaries being so fed up with constantly having to handle big piles of these cards as being the main driving-force behind the invention. According to this story, there was initially a lot of skepticism and pushback for the technology and it took some time before it saw use in a product going to customers. He also mentioned that after 8" floppies gained more use with word-processing, a problem arose where a 8" disk would not neatly fit in a purse and eventually that pushed for what would become the 5.25" disks. Apparently there had been discussions to market these as "ladies diskettes" or something like that, but I have no idea if this was an internal urban legend at IBM at the time or not. Erik was involved with the floppy-disk stuff, negotiating with IBM and Shugart Associates between ca 1976 to 1980, so that was a few years after all of that would have happened. But if someone knows more, I am eager to hear.

 

[krebizfan]

Looking at Disk Trends, the earliest double sided floppy drive I can find is the IBM 4964 from 11/76 with the Shugart 850 listed as first shipping in 5/77. The first announced double sided 5.25" drive was the Shugart SA450 from October 1977. Other 5.25" manufacturers followed with Siemens listed as 5/78 and YE Data at 1/79.

Disk Trends used announced for shipping dates in the list of all drives available. The 1978 issue lacks that list so some quickly canceled drives may not be listed.

IBM had a lot of code names for what seemed to have been floppy disk projects. Being IBM, multiple codenames may refer to the same device while a given codename might be for a project developing multiple devices. The 4" Demidiskette probably started in the late 70s given how it was canceled in 1983 after several years of work.

 

[Eudimorphodon]

Just randomly thumbing through the Byte magazine archive the first one I picked, the June 1978 issue, had multiple ads selling double-sided 8” disk drives from several manufacturers. (Including a double-sided version of the fairly infamous PerSci dual disk unit that was the heart of the Helios II disk system that basically killed Processor Tech, maker of the SOL-20.) Skipping ahead a year there are listings for double-sided 5.25” drives packaged for computers like the TRS-80. So the tech was definitely in active circulation well before April 1980. There might be some hairs you could split about who was the first computer company to make them “standard equipment”, I suppose, but they were an option going way back.

 

[per]

Thanks for the responses. But yes, he did put quite a bit of emphasis on "in Europe", so they were absolutely not first in the world.

Given what was mentioned here, I am pretty sure he meant as standard fitted in an end-product for sale. In an interview Erik did with the Computer History Museum a few years ago he talks a bit about the challenges of working with the early SA-850 drives, making a double-density controller for it. Some context to this story, Tandberg Radiofrabrikk had been shipping this computer with a single-sided drive and single-density controller since late 1976 or early 1977, then went bankrupt in December 1978 due to severely tanking sales on the unrelated consumer color-TV product line. Siemens became interested in selling an OEM of the Tandberg computer shortly before the bankruptcy, and were very quick to buy out majority share in the Tandberg computer division to not disrupt negotiations and development of this OEM computer during the bankruptcy process. The release of the Siemens OEM machine was in late 1979, but I don't know if Siemens shipped this with the double-sided disk as standard or option from the very start. According to the story in the interview, the double-sided double-density disk controller was completed in the window between the Siemens takeover and the launch of the OEM machine at least. It wouldn't surprise me if Siemens wanted first dibs on this. The April 1980 launch was Tandberg's own non-OEM machine with the double-sided disk as the main feature (otherwise mostly identical to the older single-sided disk version from earlier).

 

[krebizfan]

Note that Siemens purchased disk drive manufacturers around the same time: 8" GSI in 1978 and 5.25" Oric/Wangco/(Perkin Elmer) in 1979.

The problems with the early double sided drives led Apple to try the ill-fated Twiggy drive using a pair of single sided head mechanisms.

 

[per]

Ah! That tracks..

In the case of SA-850, it apparently was so bad that it caused bad blood and burnt bridges between a Shugart Associates and a lot of their customers. It was mentioned that they "solved" it way too late, by ultimately buying the head-assemblies directly from Tandon (literally one of their biggest competitors).

The Tandberg double-density controller solves the SA-850 issues by having a separate read-timing daughtercard. This has a quite intricate PLL, synchronizing some circuit pushing time-deltas straight into a hardware-FIFO between this daughtercard and the controller's TTL-logic based processor on the main PCB of the expansion card. This loosens the timing-requirements of the microcode firmware running the thing, to not be locked to the timing-variations of one drive in particular.

 

[MauriceH]

I have a Danish computer Piccolo RC702 with 2 drives 8" RC762 , (aug1980)
a follow-up from the RC701 system with 8" out off nov1979
Drive itself is a YE-Data ,YD-174
The manual stated "Engineering release 5.30..,1977
A Double Density, 2 sided drive.
Mostly used in a school environment.