Recalling an amazing electro-optic instrument interferometer

[mawyatt]

While posting a note about measurements wrt to Closed Loop Systems we recalled an instrument from back in ~1980 called XM-21, which was a Remote Sensing Chemical Agent detectors for battlefield use, and thought to be scientifically impossible at the time (another story if interested). Also the fielded instrument had to survive a point-blank thermo-nuclear detonation (another story if interested), which created a host of additional problems.

This XM-21 instrument housed a tiny handheld moving mirror interferometer operating in the 8-12um region. A masterpiece of engineering in both optics, mechanics and control electronics which utilized a case machined out of Invar for ~ Zero TC. Required bearings for the moving mirrors & motor were from Flex-Pivots for zero mechanical noise and ability to survive extreme shock (blast wave from Nuke see above). The fixed and moving mirrors were diamond turned corner cubes which had to be extremely flat and precise.

A Voice Coil type Motor was employed directly connected to a "Garden Swing" type arrangement for the moving mirror which was all suspended by Flex Pivots. The Garden Swing arrangement translated the Voice Coil Motor rotation into Linear Translation as required by the interferometer moving mirror.

Knowing exactly where the moving mirror optical path is in real time was required for the signal post processing, this was achieved by a clever tightly controlled servo loop which utilized a precise HeNe Laser (0.6328um) that was detected after bouncing around the optics path with a photodiode and transimpedance amp. The output was a sinusoidal "Fringe" with "Zero Crossings" at the optical path length in HeNe wavelengths of 0.6328um, however these were just Fringe counts without a start reference for "Zero count". This was solved by a small integer step of in the HeNe wavelength (recall it was 16) in the mirror corner and a spot white illumination beam was bounced off the mirror corner into a fixed photodetector.

Since the White is broadband it only correlates when the optical path difference thru the entire interferrometer is exactly Zero and thus indicates when the path of the moving mirror is exactly the same as the moving mirror which includes the optic lenses (ZnSe) and beam splitter, also compensated for thermal effects and aging as well. Thus the Impulse from the White Light Detector was used as a Start of Scan with a 16 Fringe count preempting the true Signal Optical Path Zero Mirror Position giving a precise preconditioning for the interferometer servo loop and subsequent signal processing.

In signal processing the optical signal from the interferrometer which was sensed by a custom cryogenically cooled (77K) HgCdTe detector in a custom ultra-low noise transimpedance amp (sensitivity exceeded theoretical at time, another story), this was digitized by a custom 18 bit ADC (dynamic gain ranging) which was clocked by the clock inferred from the HeNe Fringe clock. This created an overall synchronously tuned system with its inherent advantages with ADC samples taken at precise wavelengths and had excellent outside influence rejection from noise, vibration, humidity and temperature.

Very clever electro-optical/mechanical interferometer design concept, one of the most intriguing efforts we've been small part of or experienced during our career.

Hopefully some will find this interesting.

Best
   

[coppercone2]

mail to dave jones with fake mustache added to box please

next eevblog from undisclosed location in indonesia

[mawyatt]

mail to dave jones with fake mustache added to box please

next eevblog from undisclosed location in indonesia

What's that supposed to mean???

Best

[coppercone2]

it sounds like if eevblog did a tear down on youtube you might have to sneak out of the country in a fake mustache and hide in Indonesia

[mawyatt]

That's really old stuff from way back in ~1980. Nothing "classified" about the instrument, the algorithms were tho and maybe some of the detailed designs.

XM-21 worked on passive spectral radiometry which at the time was thought to be limited by various means, and thus the belief (presidential science adviser thought so!!) that the concept of passively detected chemical agents in the atmosphere at distance was impossible.

The HgCdTe IR detectors created a lot interest also, eventually later the semiconductor theory was revised to account for the "better than theoretical performance". Everyone was using these HgCdTe detectors but one one fab knew how to make them that sensitive.

Best

[coppercone2]

maybe you can use it to check exhausts of various combustion things

can it see if musks rockets are compliant with emissions?

[mawyatt]

Later the technique became popular for all sorts of things, including finding the Cocaine refineries buried in the Amazon jungles and elsewhere.

It's also used to chemically study the upper atmosphere from space.

Spotting the exhaust fumes from rockets and engines wouldn't be much of a challenge since the source gas temperature is high.

Passive Spectral Radiometry is a common tool in use today.

Best

[pardo-bsso]

I'm interested in all of your stories and technical posts.
Would contribute to a patreon or something like that if it helps.

[mawyatt]

I'm interested in all of your stories and technical posts.
Would contribute to a patreon or something like that if it helps.

There's more to the story of the XM-21 if interested (read reply and follow ups), and more around this some of which still can't be disclosed.

https://www.eevblog.com/forum/thermal-imaging/germanium-or-znse-lens-for-diy-hobbiyist-projects/msg3524760/#msg3524760

Surviving the point blank nuclear detonation proved difficult and tripled/quadrupled the instrument cost, at the time a totally ridiculous requirement IMO. The blast wavefront travels at the speed of sound and is a high density wavefront of very high temperature. To test for surviving this wavefront a special test chamber was utilized where fine Aluminum "dust" in a hopper is opened into a high velocity "Impulse" gas stream at ~1300kph impinging upon the instrument. The stream "Impulse" is oxygen, and just in front of the DUT Instrument the fine Aluminum dust is ignited by a large spark plug like device. This causes the aluminum dust to flash like a giant flashbulb from an old film camera.

During an early test the tripod back leg folded due to the pressure wavefront allowing the release latch to release the head which held the instrument, this propelled the instrument into the chamber side wall at ~1000kph shattering it into tiny pieces, requiring a new tripod design!!

The resulting test effects were all the paint from the instrument front was burnt off, and the needle in the analog voltmeter (used for external power assessment and had a Red, Yellow and Green area for power judgment) had curled up like a moth proboscis, but otherwise the instrument survived (using new tripod design).

The EMP from a detonation is just like a direct lightning strike and the usual electrical protection means applied (Transorbs, Gas Discharge and so on), however the ionizing radiation from the blast required sensing such and protecting the semiconductors.  Sensing was done with 3 axis large PN junction areas wired OR to high current/power SCRs which "Crow Bar" (short circuit to ground) all the Power Supply outputs. The general idea was to remove the Power Supply Energy Sources that would damage the circuit semiconductors since all the PN junctions become heavily forward biased due to the ionization photons. Testing for this was a special remote building/chamber which housed a large Flash X-Ray generator which could emulate the massive ionization photon flux from a nuclear blast.

Anyway, those were a few of the challenges faced when developing the XM-21 which was considered "Impossible" at the time, the Presidential Science Advisor thought so (another story briefly mentioned in the link above).

Best