In embedded systems, there is a need to permanently store information such as program code and calibration data so that it remains available after power is removed from the system. Multiple memory devices and technologies are available to accomplish this such as PROMs, EPROMs, EEPROMs and flash memory. When the memory device is first manufactured it is blank, so it must be programmed before the system in which it’s installed can be used.
Even though current generation microcontrollers typically include flash memory and have the capability to reprogram small blocks of memory once their main code is installed, for programming “empty” devices, an external programmer is still required.
Other integrated circuits, known by acronyms such as PAL, PLA, PLD, CPLD, GAL, FPGA, are programmable logic devices where the data to be entered defines a combinational logic circuit rather than software instructions; these devices must also be programmed before they can be used. The equipment to accomplish this can be called a number of names such as a “device programmer,” a “chip programmer,” “circuit programmer,” “IC programmer,” or “EPROM burner.” No matter what it is called, however, it is simply a piece of hardware for transferring data into programmable ICs.
Figure 1: A universal programmer for FPGAs. Source: Lattice Semiconductor)
There are four general types of device programmers:
1) Gang programmers: for programming multiple circuits in mass production.
2) Universal programmers: for development and small-series production.3) Pocket programmers: portable programmers for development and field service.
4) Specialized programmers: for certain circuit types only, e.g. EPROM programmers.In this article, we’ll specifically address universal programmers, which are capable of programming multiple types of devices. This is a very useful ability because over the decades—the first EPROM came out in 1971—there have been many different programmable devices and numerous ways to program them.
Programming methods all follow one basic procedure: the device to be programmed is connected to the programmer, either by plugging it into a socket on the programmer or connecting the programmer via an adapter to the board that contains the device. Once that occurs, data is then transferred to the device by applying signals to the connecting pins using pin-driver circuits.
Within this basic procedure, though, there exist many differences between devices. First, there is no one standard pinout for the programming pins. Next, some devices require data via serial inputs, some parallel. Devices run on different supply voltages and use other voltages for programming.
The result is that each universal programmer I/O pin must be able to apply voltages in a range of 0 to 25 V, clock rates of up to 40 MHz, and logic inputs with adjustable thresholds. And, of course, let’s not forget about the ever-growing number of different packages! These are usually accommodated with socket adapters as shown below.
Figure 2: Universal programmer socket adapters. (Source: Lattice Semiconductor)
As for software, there are several different file formats for binary data – the Motorola s-record, the Intel HEX format, and others – that the universal programmer must also take into account. Many manufacturers make universal programmers for all of the devices in their product line, and designers can also obtain universal programmers that work with devices from different vendors. One such device claims to work with 92,000 devices from 332 IC manufacturers!
Since there are such a large number of possible combinations, a universal programmer is an essential tool in any lab that uses a range of different memory technologies.