Talk:9-track tape

IBM 2440
2021 September LSSM in Pittsburg has a 2440 together with two 9370 and a 3490E tape unit.

2400 7-track option
It's my recollection that 2400's with the 7-track option could only process 7-track tapes.--agr 14:24, 12 October 2006 (UTC)

Rename to 9-Track?
While not wishing to slight IBM's invention and popularization of 9-track tape, it ended up being a standard interchange medium. I think there ended up being only a few companies other than IBM making the drives - See bitsavers.org. Consolidating specifications for the various drives on one page, showing the evolution of the tech would be good. ie 6250 GCR drives... Comments? RDBrown 08:10, 14 December 2006 (UTC)
 * I'm hesitant to do so because I don't know that much about all the 9-track formats. I think there is an ANSI 9 Track standard that grew out of the IBM standard, but I don't know the details.  My understanding is that 9 track tape was used for 25+ years, but there are myriad standards and densities.  User:RTC seems to know a lot about the old IBM tape drives... Maybe if we ask nicely he'll add more.  If you know the info or can filter the stuff on bitsavers that would be much appreciated! -- Austin Murphy 18:33, 15 December 2006 (UTC)

I only know the 800 NRZI, 1600 PE, 3200 PE (Cipher or Kennedy drives?), 6240 GCR (though HP also had a compression option) - all but the 3200 mentioned on the Mag tape Data page - but I've only used HP drives on HP 3000s, 9000/800, a Cipher attached to a Xenix box, maybe a drive on a small DG AOS/VS box - over the period 1980-199x. I'm probably wrong about the number of drive manufacturers too - I'd guess that the BUNCH would each have had their own drives too. RDBrown 22:27, 15 December 2006 (UTC)


 * What I know is from reading documents I have access to at the Computer History Museum while working on the IBM 1401 restoration (it used 729 drives), what others on the restoration project that know these drives have told me, and the BItsavers documents. I've put in everything I've found out so far on the subject, that seemed interesting. -- RTC 07:07, 16 December 2006 (UTC)


 * It really should be renamed to drop the "IBM" designation. IBM may have "owned" the format at one time, but it was used by DEC and others&mdash;really by anyone who made mainframes, including Honeywell, Xerox, and all the rest. 9-track was the universal data interchange format (for anything that wouldn't fit on a floppy) until cartridge tapes became more common. They're still being used, so far as I know. (For one thing, the oil and gas exploration industry used them up until the 1990s for their "seismic" data tapes, which were supplanted by 3480.)
 * And I too have fond (not) memories of Cipher and Kennedy drives; not to mention Anritsu and M9 (both top-of-the-line desktop units), as well as the little bitty Qualstars; anyone else ever use those? +ILike2BeAnonymous 03:10, 21 January 2007 (UTC)


 * The problem with 9-track as a name is that it is vague. There may have been other 9-track recording formats. Even when other companies used 1/2 inch 9-track data tape, they were commonly referred to as IBM format. The wide use by other companies should be explained in the article.--agr 23:32, 21 January 2007 (UTC)


 * No other 9-track formats that I know of, so no danger of confusing it with, say, 8-track or other audio recording formats. I could be wrong, but I have worked with the stuff and have never heard of any other formats. We used to just call it "9-track tape". +ILike2BeAnonymous 23:43, 21 January 2007 (UTC)


 * Well, what can I say? I wuz bold and moved this page to 9 track tape. Created ,lotsa redirects, too, so nobody will have trouble finding this article. +ILike2BeAnonymous 23:54, 21 January 2007 (UTC)


 * Actually, I really do have fond memories of the Kennedy drives, specifically the Model 9100. Unlike other brands I worked on that used optical sensors in the vacuum columns resulting in the jerky supply and takeup motion so commonly depicted in movies, Kennedy used a strip that sensed the tape that fed a circuit (which I can't clearly remember specifics about, but I think the strip was capacitive and the circuit was a frequency to voltage converter) resulting in a variable "error" voltage that drove the motors much more smoothly, resulting in far fewer tape breaks as the transport components aged.  Really they were the least headache to fix.  Maybe I should edit this article expanding the technical operation section before my memories of those days (<1995) fade, eh?Murasaki66 (talk) 05:30, 21 November 2007 (UTC)


 * Please do! One of the most interesting things about older computing machinery is the novel ways in which strange problems are solved.  -- Austin Murphy (talk) 14:01, 21 November 2007 (UTC)

Types of drives
The earlier IBM drives were vacuum column, were they designed as start/stop or was streaming something seen as important early? I used HP7970 tension arm drives from 1980 on, I think they're designed as start/stop. The Pertec article mentions them OEM-ing drives for other companies, ... would that have been DEC & DG? Don't remember seeing an HP autoloader until the 7980, are all the tabletop style streaming drives from the mid .. late 1980s autoloaders? HP drives after the 7970 used HP-IB interfaces, but the 7980 OEM version supports Pertec and SCSI interfaces. Looks like most of the streaming drives provide SCSI & Pertec interfaces, with SCSI sometimes provided using a SCSI=>Pertec converter card as part of the drive. The HP journal article linked mentions that the 7976 drive could autoload 10.5" reels using an EZ-LOAD cartridge. Is that the seal around the example 10.5" tape in the article - was that an IBM autoloader tool? 203.17.47.178 14:19, 8 February 2007 (UTC)


 * I can answer a couple of points, not all: the IBM vacuum-column drives were designed for start/stop use; that's why they had the vacuum column, to maintain enough slack in the tape to allow for rapid repositioning.
 * At least one drive I know of had a Pertec-to-SCSI adapter: the M4 drives. (Which, by the way, were the absolutely most reliable of any desktop drives we used, and we tested most of them available at the time.) +66.52.186.119 19:46, 8 February 2007 (UTC)

Current media manufacturers?
This article Says that Media Mastr started manufacturing 9-track tape ("NineTrax" brand) in 2002 after Graham/eMag stopped. Real 2 Reel Computer Tape, Inc. has a very minimal web page (modification date 2007-11-07) claiming to still be manufacturing NineTrax brand. You can buy it here.

What I'm not sure of is if that's old stock or it's currently being manufactured. 71.41.210.146 (talk) 14:39, 15 March 2009 (UTC)

IBM 2440
I have not been able to find any references to the IBM 2440 tape drive. Does anyone have an authorative source for it? If not, I will delete it. John Sauter (talk) 04:50, 19 October 2014 (UTC)

tape handling during rewind and unload
The article states &ldquo;...the capstan and head assemblies were always in contact with the tape, even during fast forward and rewind operations.&rdquo; However, GA22-6866-5 (referenced in the article) states in a footnote on page 51 &ldquo;Rewind unload takes less time than rewind in the [IBM 2401] model 1 or 4 because tape is wound to load point at high speed. Rewind requires reloading tape. With models 2, 3, 5, 6 and 8, tape is reloaded during a rewind-unload at the end of high-speed rewind and rewound to load point at low speed; it is then unloaded again (rewind does not require final unloading).&rdquo;

While it is clear from the text that the tape is not in the vacuum columns during high-speed rewind, it is less clear whether the tape remained in contact with the head assembly. However, the picture on page 47 shows two high-speed rewind idlers. They appear to guide the tape directly between the reels, bypassing the vacuum columns and the head assembly, which is lifted out of the way. This isn't very good evidence, but I think it is good enough to alter the article to say that the capstan and head assemblies were always in contact with the tape except during high-speed rewind operations. If there is no disagreement I will modify the article. John Sauter (talk) 21:15, 25 May 2017 (UTC)

I have updated the article as described above. 17:16, 27 May 2017 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on 9 track tape. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20081029172933/http://bitsavers.org/pdf/fujitsu/B03P-5325-0100A_244X_Jun87.pdf to http://bitsavers.org/pdf/fujitsu/B03P-5325-0100A_244X_Jun87.pdf

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 14:35, 23 June 2017 (UTC)

blocks
The article mentions inter-record gap and inter-block gap. As well as I know, this is, more or less, right. From the hardware point of view, there are records. Hardware doesn't know if records are blocked. From the software point of view, there are records, blocks, and inter-block gap. Similar to the confusion in networking between packet (software) and frame (hardware). Gah4 (talk) 02:19, 14 February 2023 (UTC)

CRC or parity?
I thought the article might mention if there is a CRC or parity byte at the end of a block. I haven't thought about this in years, and this seemed a good place to look. Gah4 (talk) 02:20, 14 February 2023 (UTC)