Re-Envisioning the Telescope

by Paul Gilster on November 6, 2012

An inventor named Tom Ditto has been casting a serious look at diffraction gratings as large primary collectors for telescopes, work that has been getting a bit of a buzz on the Internet. See, for example, An Old Idea Gives Telescopes a New Twist, and ponder how much the eponymous Dittoscope sounds like something out of a Tom Swift story. Nonetheless, an instrument based on a diffraction grating as its primary light-gathering source may prove useful in a variety of astronomical settings, including the ability to produce extremely high quality spectroscopic information for radial velocity exoplanet searches.

Diffraction happens when a small obstacle or opening causes a wave of light to interfere with itself, creating patterns that depend on the size of the diffracting object and the size of the wave. A diffraction grating, in this case a flat surface with a regular pattern of grooves, can be used to separate different wavelengths of light, which will interfere at different angles. The use of diffraction gratings in astronomy dates from 1786, when the American clockmaker and astronomer David Rittenhouse ran experiments on the behavior of diffracted light. Joseph von Fraunhofer would later use dispersion by diffraction grating to resolve atomic lines of sunlight and starlight. Spectroscopy has made use of diffraction gratings to study the stars ever since.

Image: Diffraction at work. The grooves of a compact disc can act as a grating and produce iridescent reflections. Credit: Wikimedia Commons.

In a presentation to the NASA Institute for Advanced Concepts back in 2007, Ditto showed that his Dittoscope could be conceptualized as a standard telescope capable of spectroscopy aimed at a flat grating. The standard telescope’s spectrometer has its own grating and slit. The use of the second spectrometer eliminates the overlapping spectra from the flat primary grating. Each object is thus imaged at a single wavelength at any unique angle of incidence. There are no moving parts other than the rotating Earth, with the instrument oriented east to west.

The ground based Dittoscope, then, takes advantage of the Earth’s rotation, as described in the presentation for Ditto’s Phase I study: “The precession of objects in the night sky causes their incident angles to rotate. For any incident angle there is a corresponding wavelength, so an entire spectrogram can be assembled over the course of a night.” With this enormous field of view — a 40 degree arc — millions of stars are placed within view simultaneously:

Since the output of the telescope is spectrographic, multiple object spectrometer problems are no longer the vexing issue of conventional telescopy. Every object has its spectrum taken. Stars do not need to be localized in advance of taking their spectra. There are spectral signatures for all objects in sight from first light.

Rather than tracking individual stars, the Dittoscope would simply let stars pass over the grating to detect the spectral changes of different wavelengths. As an online paper on this work notes: “There are expenses associated with the mounts, but the mechanical complexity is reduced to a single axis which is static during observation runs…” From a practical standpoint, small early experiments can be expanded in incremental steps until a full observatory is constructed.

Image: The roof is coming off the observatory. Gone are the domes, the sliding hatch doors and the rotating walls. A Dittoscope can lay flat to the ground. Its roof may be the primary objective. Wind resistance is negligible. The secondary optics are buried in a trough, and the ray paths can be protected within a pacified atmosphere, even a vacuum. Credit: Tom Ditto.

Note the other advantages that accrue from Ditto’s implementation of the idea. Because the diffracting grating primary collector is flat, many of the size constraints imposed on standard telescopes are eased, especially the requirement to contain heavy mirrors. Ditto’s paper argues that tolerance specifications for flatness in the axis of diffraction may not be prohibitive — he even talks about plate glass as a possible medium. All this points to the ability to construct enormous collecting surfaces at relatively low cost. One possibility Ditto mentions is a lunar observatory at the Moon’s equator that has no moving parts, returns detailed spectra for all objects along the zenith, and operates with a service life of decades.

Now think of this in terms of other space-based platforms. Theoretically, it would be possible to develop a collecting surface that can be stowed on board a launch vehicle in the form of rolls of kilometer-length membrane — Ditto evidently experimented in his prototypes with methods for printing diffraction gratings out on various kinds of industrial material packaging. Once in space, the rolls of collector membrane could be deployed by induced centrifugal forces.

Moreover, the Dittoscope can be segmented so that collectors of an arbitrarily large size can be constructed. As the system is deployed, it can be added to incrementally to achieve the needed specifications. Ditto explains the construction of a large, ground-based instrument: :

A very large collector can be constructed from thousands of identical smaller gratings on individual supports… Each table could support glass substrates for the grating elements. Piezo positioners or other micro-adjusters would be used to slide gratings into alignment relative to each other. Alignment of disparate segments could be achieved using a laser that targeted a test patch on each grating. The laser would work like a collimating laser. The adjusters would be used so that a calibration wave length appeared superimposed at the same position of the secondary sensors.

Huge projects thus become possible:

This suggests planet finder projects where telescopes grow incrementally, starting with demonstration models in the 100 sq. meters, followed by working versions for bright stars in the 10,000 sq. meter scale, and finally imaging faint targets using square kilometers of collector built from modules as small as 1 sq. meter.

Ditto sees his instrument as an exoplanet detection tool that would provide the spectrographic information needed to do radial velocity work of extraordinarily high quality. Because the secondary can be a conventional telescope, the primary collector — the diffraction grating — can be located in a different place. We can envision a secondary telescope based on the ground being fed by a primary collector in Earth orbit, a system that would allow the collector to be shared among many existing ground-based telescopes.

Ditto’s intriguing proposal to the NASA Institute for Advanced Concepts was funded in 2006, allowing the construction of early prototypes. A Phase II grant proposal is now in the works. The Phase I proposal is “Primary Objective Grating Astronomical Telescope,” available at the NIAC site. Tom Ditto’s paper “The Dittoscope” explains the concept online. Thanks to John Kilis and Jason Wentworth, who provided invaluable pointers to this work.

tzf_img_post

{ 14 comments }

Schneider November 7, 2012 at 8:19

Another type of telescope based on diffraction gratings, more precisely on Fresnel lense, has been designed and proposed to the European Space Ageny by Laurent Keochlin (Toulouse, France).
See for, instance some of his papers:

The fresnel interferometric imager
KOECHLIN L., SERRE D., DEBRA P., PELLA R., PEILLON Ch., DUCHON P., GOMEZ de CASTRO A., KAROVSKA A., DESERT J.-M., EHRENREICH D., HEBRARD G., LECAVELIER DES ETANGS A., FERLET R., SING D. & VIDAL-MADJAR A.
Exp. Astro., 23, 379

Fresnel interferometric arrays for space-based imaging: testbed results
SERRE D., KOECHLIN L. & DEBA P.
http://fr.arxiv.org/abs/0808.0652

The Fresnel interferometric imager
KOECHLIN L., SERRE D. & DEBA P.
Astrophy. & Spa. Sci., 320, 225

High resolution imaging with Fresnel interferometric arrays: suitability for exoplanet detection
KOECHLIN L., SERRE D. & DUCHON P.
Astron. & Astrophys., 443, 709

Alex Tolley November 7, 2012 at 11:08

A remarkably simple and clever idea. I’d be interested in the astronomers’ thoughts about this approach. Even if a space based grating is not deployed, the technique should be quite feasible on the ground. The cost being the structure to protect the grating and its careful leveling. Yet another example of Moore’s law making low cost sensors feasible.

jkittle November 8, 2012 at 13:57

this is probably what we have been waiting for in the next generation of astronomy sensors… I expect rapid progress. they are really made much more powerful by the great leaps in CCD technology , whichis tied to simiconductor progress
imagine these in space based systems… full spectrum analysis from 5,000 to 200 nanometer spectrum.. or even more extensive spread..

Eniac November 9, 2012 at 0:41

This is an exciting idea, at first glance, but I think it has very serious problems with light gathering power. From the author’s description, at any given time, for any given target, all light except that of a narrow spectral window is simply discarded. In addition, it seems that the concept of “grazing exodus” on which the concept relies heavily, ought to take off a few additional orders of magnitudes from sensitivity. In the NIAC presentation a potential sensitivity problem is acknowledged with reference to “integration time”, but not quantitated or otherwise seriously pursued. In the paper it appears to be missing entirely. The paper also appears to not have passed peer review and reads like the draft of a patent.

I wouldn’t hold my breath for a breakthrough here. Catching the few photons from a planet requires exquisite sensitivity. Throwing away any light, much less orders of magnitude, is a clear no-no.

Ron S November 9, 2012 at 15:43

Eniac, this paper may be even more deficient than you suspect. For example, note the following passage on page 6: “Without a textbook to guide him, the author here attempts a description of the error introduced by an uneven grating surface.”

There seem to be a variety of issues with the proposal, however I don’t feel motivated to look deeper. I glanced at it only because your negative assessment intrigued me. Perhaps I missed it, but I saw no reference to any prototype-level instruments to quantify the potential or the problems. This smells like a fishing expedition.

coolstar November 10, 2012 at 5:59

I’ve tried to read Ditto’s papers and frankly find them pretty much impenetrable. I’m extremely doubtful (I’ve been doing stellar spectroscopy for decades) that any workable instruments will come from this.
Thanks to Jean Schneider for the references to Koechlin and Serre’s work as I had missed that entirely. Geoff Anderson at the U.S. Air Force Academy has done similar interesting work using Fresnel zone plates as “photon sieve telescopes”: http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA531869. That reference mentions the cube sat prototype for solar imaging that’s being built at the USAFL. Dr. Anderson has also done intriguing work on holographically corrected telescope optics.

Tom Ditto November 18, 2012 at 23:51

There is a version of the Dittoscope designed for exoplanet detection that addresses some of the concerns expressed above about sensitivity. In this special case, rather than survey millions of stars, it concentrates all of the light from one planetary system. Two papers were published by SPIE in their Astronomical Sensors and Instrumentation conference this year and NIAC funded a study in the coming year (2012-13) for HOMES (Holographic Optical Method for Exoplanet Spectroscopy).

The remarkably relaxed tolerance of a primary objective grating to surface figure error has now been noted by Andersen whose diffractive primary objective telescope concept is being explored by DARPA under the project name MOIRE. The primary is a anti-hole photon sieve, somewhat similar to HOMES but without the secondary that makes the Dittoscope and HOMES full bandwidth instruments. Andersen has published an effective derivation to show how the surface tolerance can be correlated to resolving power. There is a more general derivation in many of my papers dating to 2002 when the issue was first raised. Untutored though I am, my derivation appears to hold as a first approximation.

The flatness tolerance equation is one of the dependencies that appear as a result of this novel architecture. There are many parameters to be examined in the diffraction primary objective telescope that do not appear in the literature. While surface flatness is not as daunting as one might think, pitch accuracy is.

Just as a general point, it may be noted that when you start a line of inquiry such as this novel telescope architecture, new ground must be broken. I welcome a dialogue. It is somewhat disheartening but hardly unprecedented in my experience to hear this concept dismissed out-of-hand, as in a couple of comments above, but I take courage in Paul Gilster’s accurate and articulate coverage.

Ron S November 19, 2012 at 12:14

Tom: “There are many parameters to be examined in the diffraction primary objective telescope…”

Have you or are you examining them? This is your proposal after all.

If you are disheartened by the negative sentiment of a few anonymous commenters I think you have bigger problems.

Tom Ditto November 19, 2012 at 22:21

Ron,

Parameters are filling in, but this novel architecture alters the analysis from convention.

When my optical engineering software failed, the software company told me that the optical device I had modeled could not exist even as I was looking at it operating on a bench. Eventually I found work-around solutions in the software by hand entering values in a matrix of parameters, and the CAD model began to conform to behavior of the working device. The engineering software uses equations. Those equations will be work in this case, but it takes time to identify all the relationships and to modify the equations in this context. That’s a start. Then there are performance parameters not modeled in the software. They are being characterized now.

My concept has been dismissed with the following mantra, “I don’t understand what you’ve said, and I don’t have time to study what you’ve written, but your idea seems so simple that someone else surely thought of it, figured out why it is fatally flawed, and that is why it has never appeared before in the literature.” Criticism like this has come down from on high more than once. I sense it here:

“coolstar November 10, 2012 at 5:59

I’ve tried to read Ditto’s papers and frankly find them pretty much impenetrable. I’m extremely doubtful (I’ve been doing stellar spectroscopy for decades) that any workable instruments will come from this.”

You write, “If you are disheartened by the negative sentiment of a few anonymous commenters I think you have bigger problems.” An odd turn of phrase. The biggest problem is bringing the telescope into existence. That is what we should be discussing here, and it’s heartening when I see it.

Ron S November 20, 2012 at 11:38

Tom, the way to silence critics is to build it, on a small scale, to demonstrate that it works and does what you claim it can do. Only then can you credibly invite plans and funding to scale it up into an astronomical instrument. I deal with this sort of dilemma all the time in the high-tech sector, and have done so from both sides of the table.

Tom Ditto November 21, 2012 at 1:35

Ron,

The critics were not silenced by a lab experiment. It was built on a bench in 2006 under a NIAC (NASA Institute for Advanced Concepts) grant. 2 inches. I presented the results. You’ll see them in the NIAC Report. That year NASA funding to NIAC was cut, and it went out of existence. Nonetheless, I built one out of decorative plastic plane grating material and published the result in 2009. The primary was 8 inches. Both bench models were too weak to use for astronomical observations, but the double dispersion architecture and grazing exodus configuration were proven experimentally and conformed to prediction.

NIAC was restored in 2011. To build one outdoors for observing to 12th Magnitude I proposed a 3 x 0.75 m version, but the $500K, 2 year project proposal was declined. I will reapply in July 2013 with a rebuttal to the reviewers’ claims that the resolution of the primary is determined by the diameter of the secondary. There is resistance, but the idea is neither “fatally flawed” nor the instrument useless as this peer review maintained. Hopefully another round of review, and an astronomical version will go into fabrication next year.

Meantime I continue to do research, now on a terrestrial exoplanet finder. This is even more exotic, trading the plane grating primary for a holographic optical element. The ramp up to demonstration is more difficult. The nice thing about the original Dittoscope, now renamed The MOST (The Million Object Spectrographic Telescope), was that it only requires a simple plane grating as the primary along with a readily available secondary – a typical spectrographic astronomical telescope. I can make HOE’s as in the Habitable Planet Finder but not on the scale of an astronomical telescope primary objective.

In the decade pursuing the concept, I learned that Newton’s first mirror was about 1.3 inches, and his surviving demonstration version made for the Royal Society was 6 inches. My bench models are a tad bigger at 2 and 8 inches. What they most have in common with Newton’s experiments is their provenance. It was Newton who used double dispersion to such great effect when he demonstrated that monochromatic light could not be redispersed back into a broader band of color. He called double dispersion his “Crucial Experiment.” My 2 inch demonstration was a crucial demonstration that the primary objective grating architecture works. Take a look at the NIAC Report to be found here:

http://www.3dewitt.com/tele.html

Ron S November 21, 2012 at 14:48

Does the NIAC have a written, publicly-accessible report (what your refer to as peer review) on your proposal?

Tom Ditto November 24, 2012 at 0:27

The original NIAC (NASA Institute for Advanced Concepts) was a division of the Universities Space Research Association. The original NIAC closed August 31, 2007. My Report was submitted in August 2007, but was not included in the web site http://www.niac.usra.edu/ due to the rush to close down the Institute. In the interim, my Report has been resident on my web page.

NIAC has started up again as NASA Innovative Advanced Concepts under the Office of the Chief Technologist, I have asked the NIAC administrators to put my 2007 Report on line, and they have agreed to do so on their own web site http://www.nasa.gov/offices/oct/stp/niac/index.html. I have correspondence intimating that it will take months to reconstitute the original NIAC site to the extent that all Final Reports are open to the public, but they will reissue, including mine sometime next year.

Peer review is a process that takes place when a proposal is made. Over 600 proposals were made for a NIAC Phase I grant this year and 18 issued. There were two levels of review. The first cut was made by the NIAC administration, and they narrowed the proposals to less than 100. These went out to peer review panels. Each proposer received a review. These are privately sent to the proposer. The proposals themselves are not published either, although you are welcome to read mine as well as the two SPIE conference papers that derived from my proposal. I will be putting them on my web site shortly, and SPIE has them on-line and in print under the most recent Astronomical Telescopes and Instrumentation Conference which took place in Amsterdam in July.

At the conclusion of a NIAC grant period, a researcher provides NIAC with a Report which is published on-line by NIAC. Mine will be submitted at the end of August 2013. The Reports are not peer reviewed, but they are the basis for proposals which are peer reviewed for a Phase II project.

My present grant is hardly my first application for support, and I can recite some interesting comments that helped hone my follow-on proposals. Presently I am faced with seeking to counter a criticism that the large primary objective used in the Dittoscope does not determine its resolution, as I claim, but the diameter of the secondary mirror that collects diffracted light from the primary does. There are several ways to demonstrate otherwise, and I welcome the opportunity to disabuse the critics of their misconceptions. I will probably do this through a peer reviewed journal such as Optical Engineering.

ljk April 15, 2013 at 23:23

15 Eerie Abandoned Observatories

The outlook on (or from) abandoned observatories remains negative, mainly because there’s no one there to look up and no working optics to see through. Casualties of war, funding, politics and light pollution, these 15 eerie abandoned observatories have definitely seen better days… and nights.

http://weburbanist.com/2012/07/08/watch-out-10-eerie-abandoned-observatories/

Comments on this entry are closed.