Perytons and PSN in Microwave Ovens

In April of this year, widespread global publicity—both print and electronic, including even NPR in the U. S.---was given to a report that an Australian radio astronomy (RA) facility had suffered some RFI, dubbed Perytons, because of its mysterious mimicking of sought-after distant extra-galactic signals (FRB—fast radio bursts) and ---(Roll of Drums!!!!)---the culprit was found to be nearby microwave ovens. We extend the report here because the Australian RA scientists have inadvertently discovered (or for the first time recognized) a “new” type of microwave-oven noise—which I dub PSN, for “post-shutdown noise”.

First, we check with Wikipedia on the term “Peryton” and find:

“The Peryton is a mythological hybrid animal combining the physical features of a stag and a bird. “ And at the end “Science: the term , peryton, is also used for radio signals of terrestrial origin that mimic fast radio bursts, pulses that appear to be coming from outside of our galaxy[8]. These perytons were found to be the result of premature opening of microwave-oven doors, which releases a frequency-swept radio pulse, which mimics an FRB, as the magnetron turns off[9]”

(No doubt, these last two sentences were added in April or May of this year.) (We will cite below their reference [9] as our reference [1] that contains most of the scientific analysis of the microwave-oven connection)

There is much loose talk about microwave ovens and magnetrons in the general media (just google “Perytons Australia microwave ovens”)—including the judgment that the microwave oven or magnetron were abnormal ,i.e. “Rogue microwave ovens”,  or that somehow the offending Peryton signal (which is a frequency-swept—downward-- signal between 1500 and 1200 MHz, and ¼ second in length, including an important RA band) sneaked through the door as it was opened despite the expected action of the interlocks to shut-down the magnetron and the microwave oven..

There also is some misinformation in the scientific paper [1] that speculates on “lower-frequency modes” of the magnetron, possible phenomena associated with metal in the oven or arcing that could help explain how the peryton signals escaped by the microwave-oven door.  None of this is true.(Note that the scientific paper goes into theory of the FRB which will lose the interest of most IMPI people with the introduction of esoteric terms like “Dispersion Measure, DM” which has units of pc/cm3 , where pc stands for “parsecs”.)

What the scientific study [1] does correctly report (and I have done experiments to confirm with a modern microwave oven) is that the ovens in question (Japanese-made) were not abnormal and were behaving normally, despite their advanced age (> 10-20 years). It is true that the Perytons are not generated under normal full-power operation and normal timer-triggered shutdown, but only when the shutdown is premature through door interlock switches or a separate “stop’ switch (Also at the end of “on” periods during reduced-power operation in older ovens that rely on switches or relays without triacs.) The Perytons. however, appear only half of the time---i.e. 50-% probability.

To cut to the chase, I did some experiments (with high-voltage probes and storage oscilloscope) and found that whereas timer-triggered oven shutdown merely resulted in a quick (msecs) drop of magnetron anode voltage to zero, the “premature’ shutdown triggered a decaying anode voltage waveform that lasted several seconds--but only with 50% probability. (We speculate that this merely reflects the 50% probability at any instant when a switch may be activated, that the anode voltage is at ~4 kV in the normally expected waveform for the universally-employed half-wave doubler power supply of almost all microwave ovens.) The associated finding of noise during the post-shutdown voltage decay period is what we call PSN and is not associated with any magnetron anode mode but is explained as “Magnetron diode oscillations” [2].—which are voltage-tunable. (This was the subject of my Ph. D. thesis at Harvard in 1957.)---i.e. space-charge effects which also explain [3] the generation of sideband signals (e.g. at ~2.7 GHz) which greatly annoyed [4] RA people (including Sir Bernard Lovell) at the Jodrell Bank observatory about 40 years ago. We therefore expect that the Peryton signals, well below 2.45 GHz, are radiated from the microwave oven via the power supply leads and not through the magnetron anode and the oven cavity. Our experiments  (with 10 – 12 runs) ( with a 0-3 GHz spectrum analyzer) did indeed show much more noise is generated in this frequency range as PSN, than under normal operation where shutdown is immediate without decaying voltage waveform. (We note that the existing filter-box and power supply components affect the filtering of the output signal)

It is certain that eventually studies by scientists and engineers will elucidate many questions posed by recognition of PSN—e.g. are such signals generated during the normal rise and fall periods of anode voltage in steady-state operation. Since these rise and fall periods are very small, e.g. a few msecs, it is likely that the signal strengths are weak with esoteric account of oscillation buildup time etc.

Still, as researchers [5] begin to employ more sophisticated techniques, working towards real-time spectrum analysis, even these potentially small signals will be measured---but unlikely to compete with the PSN that is the basis for Perytons. (In [5], they detected signals at 1.9 GHz during the slower rise and fall at the 50 Hz rate but not during the main pulse where rise and fall are at a rate of 32 kHz for an inverter oven.)

In conclusion, the RA scientists in Australia have done a favor to the microwave-oven community and IMPI. Up to now, the PSN phenomenon has received no attention by the microwave oven community or IMPI. Thanks to their (the RA scientists) interest in “Perytons”, the members of IMPI have new worlds to explore:  Is PSN a threat to other services besides the exquisitely sensitive RA receivers and is PSN better or worse in inverter ovens compared to conventional 50/60Hz ovens?---and can PSN be eliminated?

Let the Games Begin !





  1. E. Petroff et al., “Identifying the source of perytons at the Parkes radio telescope,” Mon. Not. R. Astron. Soc., April 10, 2015.
  2. J. M. Osepchuk, “Magnetron Diode Oscillations” Chapter 4.6, pp. 275 – 317, in E. Okress et al (Eds.), Crossed-Field Microwave Devices, Vol. 1, Academic Press, New York, (1961)
  3. J. M. Osepchuk, “The Magnetron and the Microwave Oven: A Unique and Lasting Relationship,” Proc. IEEE Int. Conference on the Origins and Evolution of the Cavity Magnetron, pp. 46 – 51, April 2010.
  4. B. Anderson et al., “Spurious radiation from microwave ovens,” Nature Vol. 282, pp. 594 – 596, (06 December 1979.)
  5. Y. Yamanaka and T. Shinozuka, “Statistical Parameter Measurement of Unwanted Emission from Microwave Ovens,” Proc. IEEE Symposium on Electromagnetic Compatibility, pp. 57 – 61 (1995).


John M. Osepchuk, Ph. D.

Concord, MA

June 30, 2015.



Posted in Uncategorized.

Leave a Reply