Having been a long-time manufacturer of oxygen concentrator equipment, when I saw an opportunity to bid on a job to provide an oxygen concentrator system (OCS) for a satellite project I jumped at the chance. The specifications were normal for this type of system and of course electromagnetic compatibility (EMC) requirements were included. Having experience with meeting EMC requirements for many projects, a review of the requirements showed that the list was much like the standard set of EMC test and evaluation concerns my company had met previously for commercial and general aviation approval. A few of the specific requirements included more detailed information than I could recall from history where we were required to meet the standard, but I didn’t consider the specifics to be excessive.
The technology was not unique but required a bit more operation time to collect enough air in the space environment to support the separation of oxygen and of course the collected oxygen needed to be compressed to maintain a steady supply for the crew and for extravehicular activities (EVAs). The valves and sensors would function the same as used on ground systems, so the technology adaptation would be minor and well within our capability.
I submitted a proposal that included funding for a contracted EMC specialist to support the project and write the required documents since our engineering expertise was focused on the oxygen system. I realized that actual number of purchased systems would be very low, so I factored in a lot of non-recurring engineering (NRE) and considered the cost of some research and development.
During the proposal development I often felt like I was a dog chasing a car—a lot of hope to expand the business but reality told me that I would never catch it.
Caught It—Now what?
As fate predicts, I won the project and during the celebration my engineering team raised the “now what” question and that night’s sleep vanished. My mind raced with more questions than I could note and by 3:00 a.m., I was putting together a plan to manage the project figuring out each team members role. By 7:00 a.m., the EMC dilemma was looming because I did not have the expert identified—the solution, find somebody.
I reached out to the test lab I had always used for support to find that testing was the focus and they did not have resources to support the design process. However, three freelance consultants were quickly identified. I contacted each to find that one was working full time on a project. I discussed this project with the other two, and were positioned to take on the lead role. I selected the one that appeared to fit best with my team and had impressive qualifications in the EMC community and iNARTE certification.
By now a week had passed, and it would be another week to get the EMC specialist aboard with the time to deliver an EMI control plan (EMICP) rapidly being consumed—the contract required delivery of the EMICP within 60 days of contract award. It seems that the satellite manufacturer wanted assurance that we understood the requirements and had an approach that could be integrated into the satellite and be compatible with all other systems.
I scheduled a team introduction meeting for the EMC specialist to provide insight on how the OCS functions and to detail the various components of the system to monitor the environment and control operation based on conditions. The EMC requirements were reviewed, and the design concept was presented. A meeting scheduled in two days to allow her to digest the requirements and examine the design.
Understanding the Requirements
She opened the meeting with a short introduction to EMC and the basis for some of the unique requirements that were contained in the specification. The EMC triad of culprit-path-victim concept supported the system integration plan identifying some high-risk compatibility concerns prompting tailoring of many requirement limits and test levels to minimize the risk of interference. This briefing made us aware that some parts of our system could be the culprit and other parts could be the victim.
The satellite manufacturer passed along the tailored EMC requirements after they had determined the choice of:
- Hardening the victim—making it less sensitive, limiting the reception of interference, etc.
- Severing the path—isolating the circuits and wiring, shielding, etc.
- Suppressing the source—limiting speed, currents, transitions etc.
Historically our approach to EMC design had been to follow the basic guidance from previous work where a shield or filter seemed to resolve the issues identified during test. We were now faced with more challenges on proximity to sensitive equipment, nearby antennas, severe weight limitations, and outgassing—things that we could easily solve in a ground-based environment.
We had to rethink our design to incorporate solutions from the ground up (pun intended). This is where the control plan became a necessity for our own design guidance and for the contract submittal.
The next few days seemed to be focused on EMC with the entire design team in one brainstorming session after the other, feeding product information to and receiving guidance from our EMC specialist. It seemed like every idea was met with a restriction and we had to gather a lot of satellite information that was mostly in the satellite integration plan. The research really helped the design team understand why the EMC requirements had been tailored, and understanding helped each to contribute more. Things like:
- Why the radiated immunity (susceptibility) test level was so high in the 2-3 GHz frequency range—because the launch site used that frequency range for radio communications during missions.
- Why was the RE102 limit in the 25-200 MHz lower than the standard limit—another system on the satellite is used for mapping cosmic radio noise radiated nearby the constellation of Sagittarius.
- Why had we not used a shielded power cable—wiring to distribute power is defined by the satellite manufacturer for all systems without shields for weight limitations.
The list of reasons provided offered many more details to consider as the design concepts matured into an approach for our OCS providing us a design map. During preparation, she identified a requirement conflict that we were able to resolve by a contract modification instead of having to take exception to being able to meet the specification.
The control plan was prepared following the content outline contained in DI-EMCS-80199 describing the overall system with emphasis on EMC electrical and mechanical control measures. The requirements were documented to show that we understood what was needed and that we had an approach for each function of the OCS. The final control plan was approved by each department having a stake in the project because the control plan identified the role and responsibilities for the EMC design.
We met our submittal date, and after clarifying a few minor concerns, we received approval in time to support the preliminary design review. We certainly realized that the control plan would be amended throughout the design process to keep the trade-offs incorporated in the document as we took care of issues related to availability of space related parts approval. Many common parts we normally used failed to meet the requirements for workmanship and outgassing, limiting the ability to survive the vacuum environment or tolerate the extreme vibration associated with launch.
Our EMC specialist continued work putting together the qualification test procedure following DI-EMCS-80201 but was interrupted several times to assemble procedures for component evaluations. We needed to be sure that many of the modules would be compatible with the OCS with respect to EMC. Module level testing was continually being accomplished throughout the design—we had to get the system right on the first design iteration or face a complete rework to solve failures.
Preparing the test procedure ran into a small obstacle. Although many tests were specified, and a few limits were tailored we were uncertain on the test standard. In the applicable documents section of the contract more than one standard was cited, so we had to determine the precedence since the tests contain conflicts and variances in the test configuration.
The satellite was to support many types of programs with each of the stakeholders citing the need for compatibility based on their standard. The OCS supported the entire satellite, so each of the stakeholders had an interest in making sure that their system compatibility needs were satisfied. The three standards cited in the contract were:
- AIAA S-121A-2017, Electromagnetic Compatibility Requirements for Space Equipment and Systems
- GSFC-STD-7000A General Environmental Verification Standard for GSFC Flight Programs and Projects
- SMC Standard SMC-S-008 Electromagnetic Compatibility Requirements for Space Equipment and Systems
This set of requirements placed us in an awkward position of having to integrate each of the standards into a composite test procedure considering the worst-case of each. In some cases, the test configuration variance demanded that we accomplish some tests with two different setups.
To help with the task of integrating multiple standards, she prepared a test list with notes to organize the process. The common issue that caused integration difficulty was the configuration variances for each of the standards. We decided that we would select the AIAA standard configuration and apply that to all the tests. We requested a deviation approval by the integrator and that deviation was approved. This allowed her to merge the tests based on the most restrictive limits and test levels. Her notes are provided in the following tables to show the variances of the standards.
|Table 1: Power bus conducted interference, load induced, audio frequency|
|AIAA||MIL-STD-461G CE101 Method||30 Hz–150 kHz||5 μH LISN + 5,000 μF||100 → 56 dBμA||Limit adjust for current|
|GSFC||MIL-STD-461F CE101 Method||30 Hz–150 kHz||10 μF + 10,000 μF||104 → 62 dBμA||Differential mode test|
|SMC||MIL-STD-461F CE101 Method||30 Hz–150 kHz||5 μH LISN + 10 μF||100 → 42.6 dBμA||0.1–30 Hz not specified|
|Table 2: Power bus conducted interference, load induced, radio frequency|
|AIAA||MIL-STD-461G CE102 Method||150 kHz–20 MHz||5 μH LISN + 5,000 μF||70 → 60 dBμV||LISN port measurement|
|GSFC||MIL-STD-461F CE101 Method||150 kHz–50 MHz||10 μF + 10,000 μF||62 → 26 dBμA||Differential mode test|
Tailored limit 25-200 MHz
|SMC||MIL-STD-461F CE101 Method||150 kHz–50 MHz||5 μH LISN + 10 μF||42.6 → 26 dBμA||0.1–30 Hz not specified|
Tailored limit 25-200 MHz
|Table 3: RF common mode conducted emissions, power and signal cables|
|AIAA||MIL-STD-461G CE101 Method||150 kHz–20 MHz||5 μH LISN + 5000 μF||63 → 20 dBμA||Tailored limit 25-200 MHz|
|GSFC||MIL-STD-461F CE101 Method||30 Hz–200 MHz||10 μF + 10,000 μF||50 dBμA||Tailored limit 25-200 MHz|
|SMC||MIL-STD-461F CE101 Method||30 Hz–50 MHz||5 μH LISN + 10 μF||50 → 20 dBμA||0.1–30 Hz not specified|
Tailored limit 25-200 MHz
|Table 4: Conducted Emissions, Antenna Terminals|
Not applicable—no antenna ports
|Table 5: Conducted Emissions, Differential Mode, Time Domain, Load Induced Voltage Transients|
|AIAA||AIAA standard||Transients||5 μH LISN + 5000 μF||50 μS >10% power||External switching|
Voltage measure at LISN
|GSFC||GSFC standard||Transients||10 μF + 10,000 μF||1 μS → 20mS|
% of steady state current
|5 μH LISN + 10 μF||Various||Periodic, aperiodic, ripple, in-rush|
|Table 6: Audio Frequency Conducted Susceptibility, power leads|
|AIAA||MIL-STD-461G CS101 Method||30 Hz–150 kHz||5 μH LISN + 5,000 μF||1V/4A||Current limited|
|GSFC||MIL-STD-461F CS101 Method||30 Hz–150 kHz||10 μF + 10,000 μF||0.1V → 1 V 0.02–2W||Power limited|
|SMC||MIL-STD-461F CS101 Method||30 Hz–150 kHz||5 μH LISN + 10 μF||2V → 1V|
|Table 7: Conducted Susceptibility, Antenna Port, Intermodulation|
Not applicable—no antenna ports
|Table 8: Conducted Susceptibility, Antenna Port, Rejection of Undesired Signals|
Not applicable—no antenna ports
|Table 9: Conducted Susceptibility, Antenna Port, Cross-Modulation|
Not applicable—no antenna ports
|Table 10: Conducted Susceptibility, Bulk Cable Injection, Swept Frequency|
|AIAA||MIL-STD-461G CS114 Method||10 kHz–200 MHz||5 μH LISN + 5,000 μF||Curve 3||CS114 standard method|
|GSFC||MIL-STD-461F CS114 Method||10 kHz–200 MHz||10 μF + 10,000 μF||70 dBμA Calibration||Power limited|
|GSFC||MIL-STD-461F CS114 Method||150 kHz–50 MHz||10 μF + 10,000 μF||83.5 dBμA Calibration||Power line test|
|SMC||MIL-STD-461C CS02 Method||150 kHz–50 MHz||5 μH LISN + 10 μF||1V (1W limited)||Capacitor coupling|
|Table 11: Conducted Susceptibility, Bulk Cable Injection, Impulse Excitation|
|AIAA||MIL-STD-461G CS115 Method||Impulse||5 μH LISN + 5,000 μF||5A||CS115 standard method|
|GSFC||MIL-STD-461F CS115 Method||Impulse||10 μF + 10,000 μF||5A||CS115 standard method|
|GSFC||MIL-STD-461C CS06 Method||Impulse||10 μF + 10,000 μF||200V||CS06 method|
|SMC||MIL-STD-461F CS115 Method||Impulse||5 μH LISN + 10 μF||5A||CS115 standard method|
|SMC||MIL-STD-461C CS06 Method||Impulse||5 μH LISN + 10 μF||200V||CS06 method|
|Table 12: Conducted Susceptibility, Damped Sinusoidal Transients|
|AIAA||MIL-STD-461G CS116 Method||10 kHz–100 MHz||5 μH LISN + 5,000 μF||0.1V → 10A||Current limited|
|SMC||MIL-STD-461F CS116 Method||10 kHz–100 MHz||5 μH LISN + 10 μF||0.1V → 10A||Current limited|
Power off test included
|Table 13: Conducted Susceptibility, Ground Plane Injection, Spike|
|AIAA||AIAA standard||Impulse||5 μH LISN + 5,000 μF||10 μS, 8V peak||Applied to isolated chassis|
16 A limit
|SMC||SMC standard||Impulse||5 μH LISN + 10 μF||10 μS, 8V peak||Applied to isolated chassis|
16 A limit
|Table 14: Conducted Susceptibility, Ground Plane Injection, Audio Frequency|
|AIAA||AIAA standard||30 Hz–150 kHz||5 μH LISN + 5000 μF||1V (10 → 0.1 A current limit||Applied to isolated chassis|
|SMC||SMC standard||30 Hz–150 kHz||5 μH LISN + 10 μF||1V (10 → 0.1 A current limit||Applied to isolated chassis|
|Table 15: Conducted Susceptibility, Ground Plane Injection, Radio Frequency|
|AIAA||AIAA standard||150 kHz–100 MHz||5 μH LISN + 5,000 μF||1V (280mA current limit||Applied to isolated chassis|
|SMC||SMC standard||150 kHz–100 MHz||5 μH LISN + 10 μF||1V (280mA current limit||Applied to isolated chassis|
|Table 16: Susceptibility to Switching Transients, Power Leads, Time Domain|
|AIAA||AIAA standard||Transient||5 μH LISN + 5000 μF||0V, 100 μS||Voltage sag|
|Table 17: Radiated Emissions, Magnetic Field|
|AIAA||MIL-STD-461G RE101 Method||30 Hz–100 kHz||5 μH LISN + 5000 μF||RE101-2 curve||Pump motor may emit|
|GSFC||MIL-STD-461F RE101 Method||30 Hz–100 kHz||10 μF + 10,000 μF||Not specified||Default to AIAA standard limit|
|SMC||MIL-STD-461F RE101 Method||30 Hz–100 kHz||5 μH LISN + 10 μF||RS101—20 dB||Pump motor may emit|
|Table 18: Radiated Emissions, Electric Field|
|AIAA||MIL-STD-461G RE102 Method||20 MHz–18 GHz||5 μH LISN + 5,000 μF||Tailored limit||Notch frequencies|
|GSFC||MIL-STD-461F RE102 Method||200 MHz–18 GHz||10 μF + 10,000 μF||Tailored limit||Extended frequency range|
|SMC||MIL-STD-461F RE102 Method||14 kHz–18 GHz||5 μH LISN + 10 μF||Tailored limit||Notch frequencies|
|Table 19: Radiated Emissions, antenna spurious and harmonic outputs|
Not applicable—no antennas
|Table 20: Radiated Susceptibility, Magnetic Field|
|AIAA||MIL-STD-461G RS101 Method||30 Hz–100 kHz||5 μH LISN + 5,000 μF||180 → 116 dBpT||Standard test|
|GSFC||MIL-STD-461F RS101 Method||30 Hz–100 kHz||10 μF + 10,000 μF||Not specified||Default limit to AIAA standard limit|
|SMC||MIL-STD-461F RS101 Method||30 Hz–100 kHz||5 μH LISN + 10 μF||180 → 116 dBpT||Standard test|
|Table 21: Radiated Susceptibility, Electric Field|
|AIAA||MIL-STD-461G RS103 Method||2 MHz–18 GHz||5 μH LISN + 5,000 μF||Tailored limit||Notch bands with elevated test levels|
|GSFC||MIL-STD-461F RS103 Method||2 MHz–18 GHz||10 μF + 10,000 μF||Tailored limit||Notch bands with elevated test levels|
|SMC||MIL-STD-461F RS103 Method||10 kHz–40 GHz||5 μH LISN + 10 μF||Tailored limit||Notch bands with elevated test levels|
|Table 22: Conducted Susceptibility, Lightning Induced Transients, cables and power leads|
|AIAA||MIL-STD-461G CS117 Method||Transient||5 μH LISN + 5,000 μF||Tailored limit||Survival— not operational|
|SMC||RTCA DO-160 Section 22||Transient||5 μH LISN + 10 μF||Tailored limit||Survival— not operational|
|Table 23: Electrostatic Discharge Susceptibility, Personnel Borne|
|AIAA||MIL-STD-461G CS118 Method||Transient||5 μH LISN + 5,000 μF||8 kV/15 kV||IEC 61000-4-2 equivalent|
|SMC||IEC 61000-4-2||Transient||5 μH LISN + 10 μF||Level 4||MIL-STD-461G CS118 equivalent|
|Table 24: Radiated Susceptibility, Magnetic and Electric (Induction) Fields|
|SMC||SMC Standard||Transient||5 μH LISN + 10 μF||200 V, 10 μS||Spike tests cables and chassis|
The testing seemed to last forever, and we encountered some issues. Overall things went rather smoothly. The in-depth analysis and control plan work guided us into developing a product that could meet the EMC requirements and the integrated test procedure answered questions from the test laboratory before the question was asked.
We received an education in EMC and realized that we needed the support of the specialist and that will continue to be true—after all we are product designers that have a basic introduction into EMC. The specialist has the skills to work the details and hopefully we will be able to bring such a knowledgeable person aboard for future projects.
The product in this discussion is fictional but the process is realistic. In most cases you would not encounter the requirement for multiple standards—that would be resolved by the procurement agency. However, approaching the requirements should follow an organized method paying attention to the details.
Hopefully, you will find this information useful and I welcome questions. If you have a topic associated with EMC that you would like to have reviewed, let me know and I will try to place it in the queue for future articles.