by Michael Schwartz
Today, less and less test equipment comes standard with a GPIB communication interface. Many don’t even offer it as an option. It wasn’t that long ago manufacturers included a GPIB interface standard on most test equipment. If you wanted to sell your electronic test equipment and get top dollar, it had to be GPIB compatible. But not anymore!
For more than 40 years, GPIB was the defacto standard for instrumentation control. It was simple to set up, easy to use, cross platform compatible. It controlled everything—not just test equipment; printers, hard drives, and all kinds of computer peripherals had GPIP versions at one time. At 8MB/S, it was also screaming fast running, supporting up to 15 instruments at one time and communicating with a device as far as 20 meters away. It also supported bi-directional communications, as well as instrument to instrument communications.
When I was first introduced to the General Purpose Interface Bus (GPIB), I thought it was amazing. I was new to automation, and computers at that time were not user friendly. Low level programming in C was the norm, and instrument control commands were more like cryptography. What has stuck in my mind all these years was when an engineer I was working with told me when they created the HPIB standard back in the late 60s, 8MB/S was theoretical. At that time, I was programming on a 286 computers with a top speed of 6 MHz.
GPIB gave us more than just communications; its simplicity allowed engineers to easily rack up a set of test equipment. Before, most calibration labs had all their standards located on their work bench. GPIB allowed us to take test equipment from different manufacturer’s, rack them up into a single system, and have it all work.
But in today’s fast paced world standards, GPIB is slow. Today’s computers are running Quad Core 3.4GHz microprocessor—multi-tasking, streaming videos, surfing the web, and crunching numbers—then they have to change gears, by slowing down to push data through GPIB cable that is 400 times slower. If you have ever tried to upload a signal to an arbitrary waveform generator or pull a trace from an oscilloscope, you know what I am talking about.
The demise of GPIB is not just going to be speed; economics is also one of the driving forces. GPIB chips are expensive, because they lack the economy of scale compared to other more predominate communication chips like TCP/IP. The average GPIB cards cost over $350 dollars, whereas a USB and network interface cards are under $5 bucks.
Today, we are witnessing a change in instrumentation standards and computer technologies. As test equipment and computers become more hybridized, we will continue to see more of the innovations in computing technologies spill over into test equipment. USB and TCP/IP is just the start of this evolution—soon, things like Blue Tooth and even WiFi enable test equipment will be common. Everything new in computing technology will eventually be integrated into test equipment.
As communication standards like LXI and USB-UTMC are adopted by the industry, GPIB is slowly becoming a relic of the past. Technicians of tomorrow will be both measurement experts and information technology professionals. They will have to understand TCP/IP configuration for both LAN & WAN configurations—security, virus protection, computer backup operations, as well as imaging a full computer system. They will need full administrative privileges to install USB instruments. In a sense, your calibration labs will become a microcosm of computer support specialists.