Factors Suppliers Consider Before Sending a Memory Device to Space

Since the late 1950s, governments and agencies have devoted billions to interstellar exploration. Launch systems and software development, according to NASA, is constantly being innovated, especially by using open-source coding. This code relies on optimized memory—both volatile (DRAM, SRAM) and non-volatile (ROM, Flash, NVRAM)—to run properly.

Unfortunately, memory devices have commonly created problems for space missions. Engineers must follow special guidelines when designing this hardware and manufacturers are apt to solve these issues up front with new radiation-hardened (rad-hard) memory devices. 

For instance, in the past week, Microchip added yet another device to its radiation-tolerant COTS-based devices designed for space systems: the 64-Mbit parallel SuperFlash memory. 

 

Microchip says that even when Flash is still biased and operating, the new device is radiation-tolerant up to 50 kilorad TID. Image used courtesy of Microchip 

 

Using Microchip's rad-hard microprocessors and FPGAs, the new SST38LF6401RT SuperFlash device "offers the vital protection these space systems need for the most reliable digital processing where companion Flash memory is required to store the critical software code or bitstream that drives the complete system,” according to Bob Vampola, associate VP of Microchip’s aerospace and defense business unit. 

As a general rule, though, what considerations must manufacturers take into account when providing designers memory solutions for space applications? 

 

Types of Space Stressors on Circuit Memory

Though similarities exist between applications, memory units used in space should be galvanized. This is necessary because space travel creates unique stressors (downloads as PDF)—like higher-sustained levels of ionizing radiation and acute exposure events. Ionized particles carry more energy than their non-ionizing counterparts and pose greater risk to key electronics. These particles can disrupt the atomic integrity of those they come into contact with. 

Other specific threats include:

• Cosmic radiation: Caused by waves containing varied particle types
• Solar particle events: Large solar flares or particulate bursts into space, mostly protons
• Trapped radiation: Electrons, protons, and ions contained within Earth’s magnetic rings

 

Some sources of ionizing radiation. Image used courtesy of Analog Devices and NASA

 

How Does Radiation Impair Circuits?

Damage may either be temporary or permanent, depending on dosage and threat. A circuit may be impaired for some time before it can recover. A memory device might behave unpredictably during that period. 

This is called a soft error since it’s ultimately fixable. The real trouble comes with hard errors. Radiation can cause permanent changes in semiconductor structures—indefinitely hampering, if not destroying, that component’s functionality. 

Observed effects can include the following: 

• Logic state disturbances
• Amplifier disturbances
• Write errors stemming from false currents
• Data processing errors

 

Architectures and Countermeasures

Hardware units used in space—primarily CMOS circuits, alongside BJT and FPGA technologies—are built to withstand interstellar radiation. Circuits contained within are hardened against these invisible forces. Commercial electronics aren’t. 

While commercial components traditionally cost less, per-unit pricing isn’t a massive concern for novel, low-production systems. Furthermore, radiation-resistant memory boosts the overall safety of manned spaceflight (via reliability). 

 

Generations of rad-hard SRAM. Image used courtesy of BAE Systems (downloads as PDF)

 

Designers have to cram a lot of functionality into small spaces. Not only that, mechanisms with finite power supplies must be frugal with that power. This includes rovers, but crucially spaceships. Power is generated onboard and fuel is limited; loss of power can threaten safety and mission longevity. 

 

Manufacturers Gear Up Against Radiation

In a previous article, we discussed at length how radiation and electronics is a bad combination, with suppliers spearheading new manufacturing methods to combat the perils of space. 

Other manufacturers are addressing radiation malfunction by investing in memory research. BAE Systems, for example, has poured decades of R&D into creating SRAM units based on CMOS technology. Likewise, Cobham Semiconductor Solutions has introduced resistant MRAM devices to the space market. 

SRAM alternatives like GSI Technology, in contrast, utilize FPGAs, ADCs, and DACs. As previously mentioned, Microchip has also joined the fray by developing a radiation-tolerant flash memory device for code storage. This includes heavy protections against higher total-ionizing doses. 

 

Block diagram of the SST38LF6401RT. Image used courtesy of Microchip

 

This includes heavy protections against higher total-ionizing doses. 

 

Building Robust Memory Units

Space poses many threats to memory technology: mechanical stress, wild temperature swings, and heavy vibrations. Potting, shielding, and even insulated semiconductors can safeguard circuitry against the unknowns. 

While earth’s magnetic field provides decent protection against space’s unique threats, it can only do so much. This is also true as memory-dependent spacecraft venture further out into space.