Computer Collection Projects
About . Before computers . Mainframes . Minicomputers . Other stuff . Projects . Wish list . Information . Library


Projects office

Here is where I read tapes and scan documents. The 7-track tape drive is at the far left in the rack. In the middle is the document scanner and to its right is a tape cleaner, mounted on the cleaner is one of the tapes people send to be read. Out of the photo to the right is the large format scanner.

Reading Old Media

I am, from time to time, able to read several kinds of obsolete computer media to recover historic and/or interesting programs and data.

The equipment to do this is also old and can be somewhat temperamental, so the capability to read any particular format might come and go. For instance, at this writing the 7-track tape drive is in active use and working well (knock on wood) while the card reader is lost in a box somewhere (and didn't work last time I tried it), the paper tape reader I used to use was acting up and will be replaced with another device, even my fancy new document scanner is currently down because its PC won't boot. (later update - its working fine now.) Generally when a load of data piles up in a particular medium, and the previous work tapers off, I will repair the next device and work with it for a while.

All cases involve hooking up an older device (though often not nearly as old as the media to be read) to a modern PC. Often this involves a special interface of one sort or another, and software to run the old device and convert the incoming data to a byte-oriented file format.

7-track Magnetic Tape

I have a 1970's Pertec 7-track tape drive in pretty good operating condition. At first it was connected by a home-made but otherwise conventional digital interface to a PC, and I went through two generations of software to operate the tape drive and read the data. However, there were always tapes that would not read cleanly, especially high density (800 BPI) tapes.

Now the tape drive is connected through an analog-to-digital interface based on a stock Data Translation data acquisition board. This takes analog data directly from the tape head read amplifiers as the tape is spooled at a steady rate (25 inches per second) from one end to the other. The captured data is processed, compressed and burnt on CD-ROM, then processed more later to recover the original data from the tape. This technique has been very successful, with complete and apparently error-free recovery of all the data from most tapes with only one pass, even when there are mild drop-outs.

There is an interactive program for processing the analog data, that performs fully automatic processing followed, when necessary, by manual inspection and fine-tuning of processing parameters around difficult spots. Here is a screen shot and description of the program. And here is the the library entry for the tape drive.

My regular file format for 7-track tapes is one character per byte, with a filename extension of ".bcd". I'm now using this convention whether the data is BCD (even parity) or binary (odd parity.) In this format each character from the tape takes up one byte in the file, stored in the low 6 bits. The 7th bit is the parity, same as on the tape, and the 8th bit indicates the first character of a tape record. A file mark is indicated by a record consisting of the single byte hex 8F. There is more information and some tools here.

7-track tape images in the library. Note that only tapes read with the new analog setup appear in the library. Older data can be found in the Old IBM Software Project section below, until all the data there has been reread or verified with the new setup.

9-track Magnetic Tape

For reading 9-track "round" tapes there is a modern (more or less) Kennedy 9100 streaming tape drive currently connected by conventional SCSI to an old 25-MHz 486 PC. So far this drive has been able to read all the tapes I've tried on it with little trouble.

Although as I write this I've not made any 9-track data available, I plan to use one of two file formats, each storing bytes from the tape directly in the file.

The general format is the ".tap" file used by the simulators of the SIMH project. In this format each tape record is preceded and followed by a 4-byte count of the number of bytes in the record. Thus, between two records there will be the count for the previous record and the count for the next record.

In special cases where the record structure of the tape is not important the image will be presented as it comes straight from the tape with nothing between records. This is appropriate for many Unix and streamed tapes with fixed length records, and particularly for unlabelled tapes written in a well known format, such as .tar tapes.

9-track tape images in the library.

Punched Cards

Using a small desktop card reader with a home-made interface to the parallel port on a PC, I've read several card decks of IBM 709 diagnostics. That card reader is in storage and I've since acquired another promising unit, so future punched card projects might use a different setup. There are now many decks of IBM 650 programs waiting to be read.

Punched card decks are presented in files with the same general format as 7-track tapes, as if the cards had been read onto tape. Cards containing only text are stored in BCD in 80-character records. Binary card decks are stored in 160-character binary records, where each card column is stored in two adjacent characters. The first character has the bits for rows 4-9, with row 9 in the lowest bit. The second character has the bits for rows 12, 11, 0-3, with row 3 in the lowest bit. In either format if the record is short, omitted columns should be treated as blanks.

Punched card images in the library.

Paper Tape

I have two small paper tape readers with stepper-motor pin feed and optical sensing. One has a home-made parallel port interface and I used it to read several paper tapes. However the last times I used it there were problems where it obviously misread some of the data.

I have many Digital Equipment Corporation diagnostics and other programs on paper tape, as well as an entire library of tapes for the Bendix G-15, Royal Precision LGP-30 and (I think) Royal Precision RPC-4000. Some of the library tapes are in very bad condition, but I am confident that its possible to read them. I may build some special equipment for that task.

Paper tape images in the library.

Scanning Documents

The collection includes a large number of documents of many different kinds. It is my intent to scan and make available all of the unique documents I have. Many of them are loose or stapled pages approximately A- or B-size (8 1/2 x 11 inches or 11 x 17 inches) that can be easily and safely taken apart and put back together. For these I have a Bell & Howell drum scanner with a large capacity sheet feeder and gray-scale capability. For larger documents such as blueprints I have an older large format scanner.

I scan at 300dpi, gray scale. For the Bell & Howell I use software that automatically rotates pages into horizontal alignment and saves the scan with 32 levels of gray and white level clipping so they compress better. The large format scanner only produces 32 levels, I take that as it comes. I save the raw scans on CD-ROM and later process them to automatically detect photographs. Photographs are de-screened to 100dpi, everything else (text and line drawings) are converted to bilevel black and white.

For distribution I'm currently using Adobe PDF made directly from the processed scans. This is easy to view or print, and I have different processing options optimized for each case.

Documentation scans in the library.

Other Projects

My first big project was the Old IBM Software Project, described on its original pages. Inspired by a first load of tapes from a fabulous cache in Ohio, this was an attempt in 1998 to gather all the IBM 1400 and 709x series materials available and put them on a CDR.

Another ongoing (in fits and starts) project was inspired by several other collectors finding, apparently, IBM training devices for the earliest IBM vacuum-tube modules. The IBM pluggable units project pages contain information from the modules and documents in my collection.

Back to the Collection

On to the Wish List