The Space Reviewin association with SpaceNews

Stacksat prior to integration atop its Atlas launch vehicle. Stacksat involved three small, lightweight satellites built for science and engineering purposes at minimal cost and launched in April 1990 after some drama. (credit: Joel Powell)

The Stacksat saga story

Bookmark and Share

Most satellites end up as nothing more than footnotes in unread history, long forgotten by all but those who built and flew them. But even the forgotten and obscure satellites can tell part of a larger story. In April 1990, the US Air Force launched a trio of small experimental spacecraft. Although not particularly notable, they typified a then-emerging trend for very small satellites, a trend that has reappeared and faded several times throughout the history of the American space program.

Bigger is better, right?

There is a general rule for spacecraft, whether they are in Earth orbit or flying around Mars: bigger satellites are better satellites, or at least certainly more capable satellites. They can collect more data, relay more signals, or perform more functions. While this might seem axiomatic, the important thing to understand is that capability doesn’t simply increase directly with mass, particularly at the lower end of the satellite mass range. For instance, a 2,000-kilogram satellite is often more than twice as capable as a 1,000-kilogram satellite, and satellites that are really small are often capable of very little at all. Even worse, at the lower range, launch costs start to eat up a larger and larger percentage of the overall mission cost, meaning that you’re spending a lot of money simply to get into orbit, not to actually, well, do anything.

The concept of the smallsat is one that has come and gone over the decades.

Of course, budget limitations prevent—or at least should prevent—developing ever-larger spacecraft simply to reap the relative performance benefits that they provide. A 1,000-kilogram satellite may be far more capable than one weighing half as much, but it may not be affordable. This is one reason why spacecraft designers have, from time to time, sought ways to break out of this relationship and to look at increasing the capabilities of small satellites.

The concept of the smallsat is one that has come and gone over the decades. Small satellites have existed since the beginning of the American space program. The Transit navigation satellites were highly successful small satellites. Both the US Air Force and Navy operated small signals intelligence satellites throughout the 1960s, ’70s and ’80s. But, starting in the 1980s, some within the US Congress, industry, and the military advocated a “smaller is better” approach to satellites, arguing that advances in technology would allow them to break out of the relationship they long had with much heavier and expensive satellites. New technology could enable them to squeeze more utility out of a relatively small payload, refuting the claim that small satellites are not capable of much.

During the late 1980s, the Defense Advanced Research Projects Agency (DARPA) explored the “Lightsat” concept for a communications “switchboard in the sky.” Pushed partly by persons involved in the Strategic Defense Initiative, around the same time several organizations within the US Department of Defense experimented with small satellites to determine if developments in electronics and instrument miniaturization had made them feasible. Stacksat emerged from this movement.


The designation “Stacksat” actually referred to three independent small satellites launched atop an Atlas E rocket from Vandenberg Air Force Base on April 11, 1990. The satellites were part of the Air Force’s Space Test Program and hence they also had the designation P87-2, or the second Space Test Program project initiated in 1987. The satellites were designed from the beginning to be small, cheap, and to operate independent of the military satellite control network. This last point was key, because reserving time on the Air Force’s satellite communications net was difficult, especially for small experimental satellites.

Stacksat started in 1987 with a concept developed by Defense Systems Incorporated, a small company located in McLean, Virginia, just outside Washington, DC. DSI’s idea was to control small satellites from ground stations using personal computers, a technology which was then only a little more than a decade old. Although it is not clear if this was inspired by DARPA’s Lightsat project, DSI was also involved in that work around the same time. The Air Force decided to fund the project, with the Office of Naval Research providing contracting services. Initially the idea was to build four experiments for launch aboard a Scout rocket, apparently either building entirely new spacecraft or refurbishing surplus Navy Transit satellites. Finally, the Air Force settled on building two satellites of an entirely new design.

Representative Aspin got the HASC staff to do a special investigation on the uses of small satellites, and the result of that was shocking: they concluded that “small satellites are a solution looking for a problem.”

As the development continued, the few remaining Scout rockets were allocated to other payloads. The Air Force therefore assigned a new rocket to the project, a refurbished Atlas E ICBM. This was Atlas 28E, originally based at Forbes Air Force Base in 1961. Because one of 28E’s engines was not capable of producing as much thrust as a normal Atlas engine, this rocket was not useable for other, more typical Atlas missions, and it was therefore assigned to carry a lighter payload. One unique aspect of the rocket choice was that the mission retained the Altair upper stage originally planned as the fourth stage of the Scout rocket, since the two satellites were designed for that system. Altair used a Star-20A solid rocket motor that generated 32,600 newtons (7,330 pounds-force) of thrust in vacuum. But an Atlas was far more powerful than the Scout, and program managers took advantage of this by adding a third spacecraft atop the Altair.

The three satellites carried four primary experiments. They were nearly identical except for differences in their experimental equipment. They were gravity-gradient stabilized and weighed 54–68 kilograms (120–150 pounds) and had average on-orbit power requirements of 10–15 watts. They were designated POGS, TEX, and SCE.

According to Wayne Eleazer, who was the Atlas launch director at Vandenberg Air Force Base in the later 1980s when Stacksat was started, and later went to the Pentagon, Stacksat had a spate of bad luck and bureaucracy. Eleazer said that by the second half of the 1980s Congress had taken an interest in small satellites. In particular, Les Aspin, the chairman of the House Armed Services Committee, wanted to see the Air Force build smaller satellites.

Part of the problem was that when the Air Force committed to launching all of its payloads on the Space Shuttle, spacecraft designers could make their spacecraft bigger, and hence their cost increased as well. Even after the Challenger accident resulted in the transfer of national security payloads to expendable launch vehicles, the Air Force could not immediately downsize their spacecraft. At the Pentagon, Eleazer found himself trying to explain to Congress that spacecraft were designed around mission requirements, “and that, in turn, drove the size of the payload,” he wrote.

“Representative Aspin got the HASC staff to do a special investigation on the uses of small satellites, and the result of that was shocking: they concluded that ‘small satellites are a solution looking for a problem,’” Eleazer remembered. Part of the push for smaller satellites had come from entrepreneurial companies looking for an entry into the satellite market: they could not build big, complicated, expensive satellites, “so they made a virtue out of a necessity and insisted that small satellites were better… just because.”

According to Eleazer, Aspin’s special investigation had a surprising result—it changed Aspin’s mind. From then on, the Air Force had to provide “special justification” for building a small satellite. This reversal may also have made Stacksat vulnerable to what happened next.

Eleazer speculated that a 60 Minutes story about “Radio Shack parts” in Peacekeeper missiles, which had a ripple effect throughout the aerospace industry, may have prompted an employee to claim that similar parts were installed in some satellites then under construction, including Stacksat. The Defense Criminal Investigative Service (DCIS) raided Stacksat’s manufacturer, Defense Systems Incorporated, and had a Navy satellite and possibly also one or more of the Stacksat satellites partially dismantled in search of the bogus, non-space-rated parts. “Well, not only did they not find any such parts, they found out that nobody gave a rat’s rump if there were any used anyway, since they were both low cost research satellites that were not going to be used operationally,” Eleazer recalled. “DCIS frantically tried to save their case through various absurd methods, and someone even produced a cartoon of an Atlas lifting off while a DCIS patrol car drove up with lights flashing. I understand that DCIS even requested special independent analysis of the spacecraft telemetry to try to prove that bad parts were screwing things up. It cost $1 million to fix the damage to the spacecraft that DCIS had done and the launch schedule was screwed up for some time.” Eleazer remembered that the Air Force sent a strong protest letter to DCIS, “about the strongest letter I have ever seen from one government agency to another,” Eleazer said. “But I thought that DCIS deserved an even stronger letter,” he remembered. After all, it was idiotic to pursue such an investigation for some small satellites that never cost much to begin with.


On April 11, 1990, the Atlas Altair placed them into 740-kilometer (400-nautical-mile) (polar orbits after launching from the SLC-3 complex at Vandenberg. The satellites were connected to each other by four tension bolts. These were severed when the Altair reached orbit, and four springs on each satellite pushed them apart. About thirty days after launch the satellites started gathering data. They had a one-year lifetime requirement and a three-year goal.

In many ways, Stacksat was an exception to the smallsat story, not the rule.

The Polar Orbiting Geomagnetic Survey, or POGS, carried two experiment packages. One was a magnetometer and the other was a Solid State Recorder (SSR) manufactured by Fairchild Space Company in Germantown, Maryland. The SSR recorded data on the Earth’s magnetic field collected by the magnetometer. The data was used for the latest version of the Defense Mapping Agency’s geomagnetic charts, regularly used by mariners for navigation and updated on a five-year basis. Although the data was collected for the Naval Oceanography Command, it was being used in a model constructed by a joint US/British Geological Survey project.

The Transceiver Experiment, or TEX, used a variable-power transmitter for studying irregularities in the ionosphere that disrupt high-frequency radio signals. A variable power transmitter determined the power level required to overcome such disruptions.

The Selective Communications Experiment, or SCE, involved a variable-frequency transmitter to determine which frequencies are affected by irregularities in the ionosphere. SCE also employed a “store-forward” technique. Short messages were sent from the ground, stored aboard the spacecraft, and retrieved using various frequencies.

In addition to POGS, TEX, and SCE, Stacksat carried a secondary payload known as the Prototype Deployment Device, or PDD, attached to the payload adaptor situated between the Altair upper stage and the bottom spacecraft. PDD was a mechanical latch designed for use in zero-gravity.

The Stacksat UHF ground stations were located near Washington, DC, at DSI’s facility. Once the satellites checked out okay, the ground stations were turned over to the experimenter teams. Even in the days before cellphones, there were a lot of transmitters in the Washington area which created interference problems for the ground stations.

Problems with the satellites were solved, but the TEX experiment was troublesome. The ground stations could not establish consistent communications with the satellites, but eventually these were solved for SCE and POGS, and later TEX. Ground stations were established at Edinburgh University in Scotland, at Fairbanks, Alaska, and at Navy Oceanography Command’s headquarters in Bay St. Louis, Mississippi.

The past does not repeat, but it echoes

In many ways, Stacksat was an exception to the smallsat story, not the rule. The satellites got a cheap ride to orbit thanks to a leftover Atlas that was not capable of doing much else. Although most of the Air Force’s small experimental test satellites tended to have more instruments and experiments, Stacksat’s small satellites did not really need to be any bigger and more sophisticated than they were. In 1991, seven of DSI’s “microsats” were launched aboard a Pegasus rocket to test the Lightsat idea, but they were placed in improper orbits, limiting the effectiveness of the experiment. Within another year, the short-lived trend in small satellites for military purposes was dying out. Later in the 1990s a few people, including some who had been involved in the earlier effort, would again push for small satellites, eventually leading to the concept of Operationally Responsive Space (ORS). By this time, sensor technology had improved to the point where it was possible for relatively small satellites to perform militarily useful missions, demonstrating that they could at least accomplish something of value, and moving the argument to the cost-benefit of those capabilities. But the argument is by no means over, as demonstrated by recent efforts to eliminate ORS from the Air Force budget.