The 1992 briefing that priced PMAD at up to 220 kg per kW

What the work claims

This is a program-status briefing, presented in vugraph form, covering NASA Lewis's PMAD technology work in the context of a piloted Mars nuclear electric propulsion vehicle. Its central quantitative claim is a mass database finding: PMAD mass across existing and state-of-the-art space systems ranges from 40 to more than 220 kg/kW, with state-of-the-art power processing quoted at 25 to 100 kg/kW and a 5 to 10 kg/kW figure presented for the piloted Mars NEP vehicle total.¹ It also claims a series of component results: a demonstrated 0.24 kg/kW, 50 kW soft-switched converter using a coaxially-wound transformer; a depletion-mode silicon-carbide MOSFET demonstrated at an operational temperature of 500 °C; and a component characterization base spanning magnetics, capacitors, and switches tested to 200-300 °C.¹ All results are bench-level; nothing here is flight hardware.

How it works

The briefing's logic is that PMAD mass is set by passive components and thermal limits, not by switch topology cleverness. Power conversion at 50 kW requires magnetics that store energy without saturating and capacitors that filter ripple without degrading, and both get heavier and less efficient as temperature rises. The reported program therefore attacked materials rather than circuits: soft magnetic alloys (80-20 nickel-iron, 50-50 nickel-iron, silicon-iron, and amorphous alloys) characterized to 300 °C with B-H hysteresis loops recorded, against a program objective of 50 kHz excitation; 50 samples of samarium-cobalt (Sm2Co17) permanent magnets from five vendors aged at 300 °C; and ceramic and Teflon capacitors thermally aged at 200 °C for 2000 hours, with life testing under full electrical bias still underway at the time of the briefing.¹

On the semiconductor side, the briefing reports radiation-effects testing of power bipolar transistors, MOSFETs, and static induction transistors at neutron fluences above 10^13 neutrons per square centimeter, tracking current-gain degradation against neutron fluence and gate threshold-voltage shift against gamma dose.¹ Its stated high-capacity-power objectives (the CSTI program) were systems above 100 kW with fault tolerance and alternator sharing, tolerant of gamma doses above 1 Mrad.¹ The thermal-electric coupling is stated explicitly: 200 °C baseplate electronics, survivable in severe environments, would exceed the 130 °C lunar surface temperature environment and reduce radiator area by more than a factor of two.¹ Supporting diagnostics were fiber-optic current sensors operating from -65 to +125 °C and surviving 17 g vibration.¹

The strongest case for it

The steelman is that this 1992 briefing identified, correctly and early, every constraint that still binds space power electronics. It treats converter temperature as a radiator-sizing variable (200 °C baseplates halving radiator area), which is exactly the argument behind today's push of converter losses out of the electronics bay and into hotter, lighter packages. It identifies magnetics and capacitors, not switches, as the mass and reliability wall. And its 0.24 kg/kW coaxial-transformer 50 kW converter, demonstrated at University of Wisconsin under a NASA grant, shows the mass target was being approached with hardware, not just projected: 0.24 kg/kW is nearly three orders of magnitude below the worst end of the quoted 40 to over 220 kg/kW system-level range, because system-level PMAD mass is dominated by everything around the converter.¹

Where a skeptic should push

The load-bearing assumption is that component-level results scale to system-level mass, and the briefing itself supplies the counter-evidence: even with a 0.24 kg/kW converter demonstrated, the system-level database still spans 40 to over 220 kg/kW, because harnesses, connectors, fault protection, redundancy, and thermal hardware dominate. The 5 to 10 kg/kW "piloted Mars NEP vehicle total" figure is a goal-state number whose provenance on the slide is ambiguous (the vugraph layout does not cleanly separate it from the 25 to 100 kg/kW state-of-the-art band), so it should be read as an aspiration, not a measurement. Several quoted test endpoints are goals rather than results (95 percent inverter efficiency at 125 °C is a stated target; the 50 kHz magnetics objective exceeds the tested range). The gamma-dose axis in the radiation tests is partially illegible in the scan, and no error bars or sample counts beyond the magnet lot sizes appear. Finally, all of this is 1992 state of the art: silicon bipolar-era devices, pre-wide-bandgap commercialization, with the SiC result an early research device rather than a product.¹

What 1992 PMAD mass data says about power in space

The non-obvious implication is that the mass of conditioning a watt can exceed the mass of generating it, so any specific-power roadmap that stops at the array or reactor output is telling you less than half the story. The briefing's own numbers make the case: a 50 kW converter at 0.24 kg/kW sat inside systems whose total PMAD mass ran 40 to over 220 kg/kW, meaning up to nearly three orders of magnitude of harness, protection, and thermal overhead between a working converter and a working power system.¹ Modern kilowatt-class spacecraft still discover this the expensive way: the converter is a solved subproblem, the system is not.

The genuine threat is nuclear-electric coupling: the briefing's radiation data (current-gain degradation at neutron fluences above 10^13 n/cm2) exists because an NEP vehicle parks its electronics near a reactor, and shielding or distance both cost mass that lands back on the PMAD and thermal budget it was meant to protect.¹ The opportunity is the one the briefing states most plainly and the industry still underuses: raising electronics operating temperature is a radiator program. The 200 °C baseplate work, justified explicitly by a more-than-two-fold radiator-area reduction in a 130 °C lunar environment, means every degree of converter temperature tolerance bought back is heat-rejection mass not launched.¹ That trade, not switch efficiency, is where power-system mass is actually won.

The bottom line

Established by the record: system-level PMAD mass of 40 to over 220 kg/kW across the surveyed systems, a demonstrated 0.24 kg/kW 50 kW converter, SiC MOSFET operation at 500 °C on the bench, and component aging data to 300 °C. Established as goals, not results: 95 percent inverter efficiency at 125 °C, 5 to 10 kg/kW system-level PMAD for a Mars NEP vehicle. The core diagnostic, that magnetics, capacitors, harness overhead, and thermal coupling set PMAD mass, has aged well and matches present-day practice. Confidence: high in the quoted measurements as 1992 bench results; moderate in their transfer to modern wide-bandgap systems, where the same passive-component and thermal constraints apply but at higher switching frequencies and lower masses. What would break the diagnosis: system-level PMAD masses routinely falling toward the converter-level floor on flown high-power spacecraft.