A monthly column of technology rambling, rumination and reality

By: Jud Early, Corporate Vice President, Research, [TC]²

My introduction by Mike Fralix in last month’s newsletter promised our readers that I would share with you some thoughts on technology that are triggered by my eclectic taste in technical areas. It’s now time to deliver on that promise. Before we get to the topic for this issue, I’d like to provide a little insight into my background and experience. While not quite an old codger, I have reached an age where some nostalgic rambling is usually tolerated by polite people, and by some curious crossing of paths, may have learned something worth sharing. I hope to share with you in a way that provides interest, and possibly a little humor.

As a pre-teen, I found electronics to be fascinating. From crystal radios to short-wave receivers and FM tuners, I built many projects from hobbyist magazines and kits. Repairing television sets for friends and neighbors provided cash to support my car expense. By the time I was a junior in high school, I was teaching electronics and vacuum tube theory to sophomore and junior classes. This had the benefit of allowing me to miss some less interesting class subjects with the full permission of the teacher and principal.

Entering college, I declared electrical engineering as my major. I was also beginning to work in industry in an electrical production and development environment. After a couple of years, my interest shifted to mechanical engineering, which fit better with the work that I was doing in industry. At twenty-five, I owned my own business, a machine development corporation that designed and built equipment for packaging, paper, film and foil converting, manufactured parts for aircraft and Saturn rockets, and any other jobs that were within the capability of my employees.

After three years in business, and with inadequate capitalization, I decided to take a breather from the pressure of making the weekly payroll and selling our services in a difficult economic climate. Shuttering the business, with plans to start up again after the economic downturn of 1970 had run its course, I was offered a position in the R&D Department at Haggar Slacks by Joe Off. Joe had been a customer, and I had built machines for Haggar in my machine development business. My plans to revive my business were placed on hold as I found the stimulation of challenging technology and the opportunity to advance in a company where your ability is recognized and rewarded. It was twenty-one years later that I decided to leave Haggar to join [TC]². During my time at Haggar, I had a wonderful opportunity to observe how a well managed company is run, and to participate at various levels of management, in many different areas, but always with technology as my base.

Joining [TC]² in 1991, my first tasks were to re-evaluate the research direction, and to begin shifting from outsourced research services to the building of an in-house team of engineers and scientists. Our projects have been far-ranging; from sewing machine attachments to laser cutting and laser enhanced adhesive bonding, software and transport systems. Computers play a big role in the work here at [TC]², and are the foundation of 3D scanning and systems for the creation of patterns from 3D data. Digital printing, and the pioneering work in ink development and engineered print design here at [TC]² are rooted in a continuing personal interest in printing and converting of web goods. From this wide mix of experiential and intellectual curiosity I will bring to you each month a story that I hope will inform and perhaps, entertain.

A Walk Down Memory Lane

No, it’s not the paragraph above, but some elements from that history will be a part of this story.

Early computers had very limited memory. Does anyone remember the term “core dump”? What is a “core”? The first computer that I had the chance to work with had core memory. “Core” refers to the ferrite beads (tiny torroid shaped doughnuts of iron ferrite) that were strung together on strands of wire that passed through the hole in the bead, and were terminated at the end with a connection to driver circuits that would pass a current through the wire to temporarily magnetize the bead. A second wire, passing at right angles to the first, and also threaded through each bead in the matrix allowed a specific bead to be addressed by row and column. A third wire passed through each bead, and was the “read” wire. The reading of the bit (bead) took place when a bead that had been previously magnetized by a current of a given polarity was activated by a reverse direction of current flow, and the magnetism in the bead was reversed. The “read wire” sensed the reversal, and by knowing which row and column was the bead’s “address”, it could be determined if the bit was set to a “1” or a “0”. The act of reading is destructive, in that the bit polarity is reversed in order to read the bit.

These memory boards were a beautiful thing to see, having such symmetry and a delicate structure that was only possible by hand stringing of the thousands of beads on their respective row, column, and sensing wires.

A close-up view of memory cores and wires

Assembly operations were often performed by skilled apparel workers, using long needles to pass wires through each bead.

A larger section of core memory. These panels were often stacked to form a cube.

The cost, obviously, was very high, and operation was relatively slow. A further disadvantage was the complete loss of state in a power down situation. The computer could not even remember how to start itself! Enter the bootstrap loader… a paper tape, one inch wide, and perforated with holes every 1/10th inch. Eight holes could be punched in a one-inch wide paper tape. Seven holes provided addressing, and the eighth, was sometimes used as a parity bit. With hundreds, perhaps thousands of rows of punched holes, the tape contained the start-up instructions for the computer, and was read by a model 33 teletype console to set up the computer to run a job. So, we see in this use, the memory is in the form of a strip of paper with holes in it.

Punched paper tape showing eight data lines, and a ninth row for sprocket pins.

Solid state memory came on the scene with the advent of the transistor. Two transistors in a cross coupled configuration, could be set so that one transistor is turned on, and the other is off. Application of a signal causes a reversal of these on and off states. This flip-flop changes the output from a “1” to “0” or from a “0” to “1”. Thus is formed the classic flip-flop circuit. This type of circuit was also known in the vacuum tube world, but had the disadvantage of large space requirements, high power consumption and heat generation. Solid state flip-flops did not generate appreciable amounts of heat and were much smaller in size. Discrete transistor based flip-flops used as memory elements were not very practical, due to the many thousands of components needed, any one of which would result in a memory error, and might not be detected for some time.

A transistor flip-flop circuit

The development of the integrated circuit provided the breakthrough to allow a small package size, solid state operation, and low heat dissipation. The earliest were based on architecture known as TTL, transistor-transistor logic. These circuits combined the elements needed to connect the transistors in silicon chips together to make functional memory elements, which could be assembled in large numbers to create memory systems.

Microscopic view of an integrated circuit with tens of thousands of transistors

Early users of personal computers became familiar with the integrated circuit memory chip. Due to the cost of memory, the computers would come with memory cards with numerous sockets, and with only one-fourth or one-half of the sockets populated. As money became available, people needing additional memory could buy chips as discrete components, often packaged in an anti-static sleeve or foam sheet. These additional memory chips could be plugged into the open sockets on the board, and the system would recognize the additional memory. Companies such as Hercules made add-on memory boards that plugged into the machine’s bus, to add even more memory. Without an additional third-party memory manager, the limit on memory for an IBM PC or clone was 640Kilobytes. Today’s machines have 32 and 64 bit memory addressing and are capable of addressing terabytes of data

Along the way, as modern PC architecture evolved, memory chips packaged on single in line memory modules provided for higher density packaging. Dual in line memory modules now contain up to 4 gigabytes.

Most users will never have need for such large blocks of addressable memory. Due to the onward march of technology development, the ability for those who have a need for larger amounts of memory can meet that need with multiple off the shelf modules.

To this point, the memory that we have been discussing has been Random Access Memory, RAM. This memory, like the core memory and solid state memory does not hold data without power applied.

With complementary symmetry metal oxide semiconductor (CMOS) memory, power requirements are low enough that it can be maintained with a small battery for several years. However, this is still considered volatile memory, and at the very moment it is needed, a battery failure could result in loss of data. “Smart” devices have driven the development of flash memory. Flash has the benefit of holding information without power applied, but can be reprogrammed at will. This feature makes for easy upgrades of firmware that may reside in modems, cell phones, internet routers and other systems that are subject to version upgrades.

So, now that we have memory that can be changed at will, and does not lose data when power is removed, what good is it? Here, I will digress a bit. Before hard drives became a part of PC architecture, the computers of that day often had one or two floppy disk drives. If the machine had only one, the operating system was loaded from a floppy disk, then it was removed and the application loaded from the same drive. Then, the drive was used to store the data from the application you were using. Thus, you can see why the adoption of two floppies was quick to occur. Using the floppy disk, data became portable between machines, and the sneaker net was born.

The use of the floppy for moving data survived for years as the least expensive, although least capacity means of transferring data from place to place. There were Bernoulli drives, floptical drives, and Zip drives, but the floppy hung on. When CD burners became available for use in personal computers, those who could afford the technology were able to move much more data, although in the early days, making coasters was more the norm than making usable CD-ROM disks. Today, the CD has almost replaced the floppy for transporting data. CD burners and software are much more reliable, and the number of coasters has dropped to almost zero. However, there is a better way.

Enter the pen drive, cigar drive, thumb drive, etc. You get the picture… a small thumb shaped object that can store from 32 megabytes to 2 gigabytes of changeable data. If you are moving files of almost any type, this may be just the item for you. The long walk down memory lane has brought us to this point. Memory now is available in a pen sized module that can be plugged into the USB port of your desktop or notebook computer, and can be carried in a pocket or purse just like a key chain fob. Costing from $22 to more than $800 dollars, if data portability and integrity is important to your work, and if you need that presentation, word processing file, spreadsheet, or photo album, it can always be with you, and will not weigh you down with unnecessary hardware. For a backup of critical files, it can be removed and stored off-site. For CAD files being released to production, these thumb drives can be the medium of transport from CAD to CAM.

Typical Pen or Thumb Drives

The consumer can now carry with them a complete closet of clothing. No need to remember what color that little number is that was purchased last fall, just do a quick search for it with an embedded search tool. Measurements, too. If you were scanned or measured precisely, the measurements may allow selecting from a catalog or web site, or even uploading the measurements needed for a custom produced garment. For little more than the cost of a meal out, your data can be portable, and secure. And, when your memory fails, just reach into your pocket and pull out that thumb drive!

Next Month: What’s in a name?

Library Index | Home