space telescope design using artificial guide stars

Hi,

I came across this article mentioning using cubesats or similar small satellites as a constellation of artificial guide stars for a next generation space telescope to have a laser feedback signal to maintain pointing accuracy on telescopes that don't necessarily have inherent pointing accuracy, ie cheaper telescopes:

formatting link

That reminded me of an idea I had before, in this thread from 2014:

formatting link

That is having a coded aperture between the telescope and the object(s) being imaged.

I think the two ideas of artificial guide stars and coded apertures could be done by the same small satellite constellation around a large next generation satellite.

cheers, Jamie

Reply to
Jamie M
Loading thread data ...

If the telescope is parked at Lagrange point 2 (L2), the cubesat would have to be even further away. It is hard to imagine, what earth or sun centered orbit it would have.

To handle telescope targets in the ecliptical plane only, a large number of satellites would be required. For telescope targets outside the ecliptical plane, a huge number of satellites at different inclinations would be required.

Reply to
upsidedown

Hi,

That could be done with a smart orbit of the constellation of cubesats (which could be useful for other purposes too maybe).

Or maybe a simpler idea is to use a next generation telescope that uses a large solar sail, not to travel, but just to stay parked at a certain distance from the sun, ie at 10 AU, a 10,000 kg telescope would require a 0.59 Newton force to maintain its distance from the sun with no orbital velocity. It could be powered by a solar panel feeding an ion engine or a solar sail, in either case it would be integrated as part of the sun shield.

If a solar panel is used, at 10 AU, a 20% efficiency panel giving

2.7watts per square meter, and the spacecraft requires 20kW of power that is a 85m x 85m solar panel. If a modern space rated solar panel weighs 2kg/m^2 of collecting area that would weight 14,450kg just for the solar panels, so to keep the solar panel weight under 4000kg it should weight at most 0.72kg/m^2 of collecting area.

Then the cubesat guidestars, say 100kg, would only require a 0.006 Newton force to stay in proximity to the telescope. I think the cubesats could be powered by microwave or laser energy transmission from the telescope, to avoid them requiring having solar panels or RTG nuclear reactors, since the cubesats would stay in the field of view of the telescope and within a specified range. That would increase the solar panel requirements on the telescope, so it would be best to use a 40kW nuclear reactor on the telescope, which would power the telescope and all cubesats via wireless power transmission.

(0.006N force of gravity on a 100kg mass at 10 AU from the Sun)

If an ion engine gives 0.25 N thrust for 7kW, and can be throttled down to 0.5kW ie:

formatting link

Then three of those ion thrusters could provide propulsion for the telescope and one ion thruster for the cube satellites.

It has a specific impulse over 4000s so at a force of 0.006N, a

100kg cubesat should use 0.12ug of propellant per second, or about 10grams of propellant per day, 3.78kg of propellant per year.

If the thruster is rated for 10 years operation, that would require less than 36kg of propellant to counter the force of gravity at 10 AU for 10 years. Which would give the cube satellites a 64kg dry mass.

For the telescope, using slightly more than two full throttle ion thrusters, providing 0.59 Newtons of force, about 100 times the fuel would be required than each cubesat, so 378kg of propellant per year for the telescope. Or 3780kg for 10years of propellant, which would give the telescope a 6220kg dry mass. The propellant used would decrease as well over time as the spacecraft mass decreased.

The extra complexity of ion engines providing non uniform thrust could defeat the purpose of high accuracy artificial guide stars, so I guess they could be operated only when a specific cube satellite isn't being used as a guide star.

This could be accomplished for the telescope and cubesats by running the ion thrusters at 50% duty cycle at twice the thrust, and only doing high sensitivity imaging during the time the telescope and artificial star cubesat are falling towards the sun. Batteries would be needed to keep the solar panel size the same in that case of 50% dutycycle at twice the thrust.

Instead of the 85m x 85m solar panel on the telescope, a 20kW nuclear reactor could be used too: (ie four Topaz 2 10kW reactors)

formatting link

cheers, Jamie

Reply to
Jamie M

Hi,

One simple modification, is to make the artificial guide stars totally passive and extremely lightweight hollow pyramids:

formatting link

with the apex pointed towards the telescope and illuminate them from the telescope for the dual purpose of solar sail positioning them by aiming the laser off center of the apex, as well as using them as an illuminated guide star. They could be stacked like cups compactly if they are hollow with no base, and deployed from the telescope to its field of view. Ideally the required average illumination for guide star use would equal the average illumination required for solar sail positioning to maintain the cup within the telescopes field of view, and not falling back towards the sun. If the average laser power is too high, more hollow pyramids could be deployed to lower the average laser power per cup (ie 100 cups would have 1% duty cycle from the laser on average).

To implement the idea of a coded aperature on the artificial guide star that the telescope views objects through, perhaps the cubesat could be a light sensitive transparent material and a laser image projected onto it from the telescope which temporarily would darken the cubesat to form a variable coded aperature pattern. If that didn't work then the artificial guide star should have an active 2d coded aperture, that could be powered and have data sent via the wireless laser power transmission from the telescope.

cheers, Jamie

Reply to
Jamie M

Hi,

To further simplify the telescope design, and allow for the artificial guide stars to maintain their position in the telescope field of view at an arbitrary distance, the telescope could be launched from earth, removing earth's orbital velocity of 30km/s, and having an apogee of approx 10 AU, so the life of the telescope would be: plugs into google: sqrt(pi^2 * (5 AU)^3 /(G(m_sun+m_saturn)))

5.58 years *2 = 10 years

The telescope would fall into the sun after about 10 years, and would require no propellant or thrusters, it would only require a relatively small solar panel to power control moment gyroscopes. The artificial guide stars could be passively released from a stack, or actively controlled if a larger power source is available.

The field of view of the telescope could approach 180 degrees of the sky (limited by the size of the sun shade).

cheers, Jamie

Reply to
Jamie M

Hi,

Just one more optimization:

The telescope field of view could approach 360 degrees (the whole sky) if the telescope can aim independently of the sun shade, ie they would decouple after launch, with the sun shade blocking the sun and providing electricty over small wires to the telescope.

cheers, Jamie

Reply to
Jamie M

Hi,

To allow for the artificial guide stars to be used over the whole 360 degree field of view, for artificial guidestars that are in the 180 degree field of view of the telescope looking away from the sun should have slightly larger mass to cross section area, so that they fall towards the sun slightly faster than the telescope, and have a miniscule laser correction applied from the telescope to keep them in the field of view.

For artificial guidestars that are in the 180 degree field of view of the telescope looking towards the sun, the artificial guidestars should have a slightly smaller mass to cross section area, so that they fall towards the sun slightly slower than the telescope, and have a miniscule laser correction applied from the telescope to keep them in the field of view.

cheers, Jamie

Reply to
Jamie M

Hi,

One more option that could increase the pointing resolution by effectively averaging many guide star measurements, would be to deploy 100,000+ very small retroreflective artificial guidestars, ie 1 gram each, dispersed around the telescope, and then use a high power laser pulse at a high repetition rate while using the telescope and in a 2 arc minute by 2 arc minute field of view, have at least 10 retroreflections from the passive artificial guide stars visible.

That could give more light and pointing accuracy than a single laser from one guide star perhaps, however the cloud of artificial guide stars would dissipate over time, so multiple batches might need to be released. If a known area of sky will be viewed, the artificial guide stars could be released in a smaller field of view and then far fewer would need to be deployed too.

cheers, Jamie

Reply to
Jamie M

I think what they are trying to do is a variant of the holographic correction of the shape of the segmented mirror after launch and that the press release has garbled it into "guide star".

The technique was pioneered at Jodrell Bank to fine tune the big dish in one of its refurbishments and code derived from that was used to quantify the abberations on the original Space Telescope and deduce the corrections needed for COSTAR.

formatting link

Can't find any references not behind a paywall.

There was a time when this was done for hard X-rays before narrow angle glancing incidence imaging was possible.

Parallax would almost certainly prevent using it as a guide star.

Most fields of view contain at least one star bright enough for an autoguider so outside the atmosphere I don't see what the advantage is.

--
Regards, 
Martin Brown
Reply to
Martin Brown

It looks like a variant of the usual terrestrial adaptive optics technique, which uses high-altitude scattering of a laser beam to figure out the dynamic aberrations due to the turbulent atmosphere. AO systems usually use Shack-Hartmann sensors to measure local wavefront tilt. SH sensors are crappy but fast, and easily good enough to take out the 'seeing'. I doubt they'd be good enough for exoplanet detectors though--you'd need something more like what's used in semiconductor litho tools, which is usually moire.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

[snip]
+1

Also outside the atmosphere and in microgravity you shouldn't need to actively tune the system up provided it is moderately rigid.

Interestingly you can do incredibly well on amateur sized instruments by getting the focus right and taking a video aka Lucky seeing. Basically by selecting only the sharpest images, register and stacking them.

formatting link

It has revolutionised amateur astronomy with cheap webcams and Jupiter now imaged at high resolution almost 24/7 when it is in the night sky.

Big professional instruments need a lot more terms than just tilt-tip.

I also thought that the intention for next generation planet hunting scopes was to aim to null out the central star interferometrically.

It sounded to me like a project derived from this earlier work but it is hard to tell from the popularised press release:

formatting link

--
Regards, 
Martin Brown
Reply to
Martin Brown

That idea has been around for a long time, e.g. speckle interferometry with silver halide plates, but it's sure easier nowadays.

That only works if it's unresolved, I think, so it's better suited to smaller space-borne telescopes.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

Hi,

That is a stationary mask pattern that was used which is integral to the X-ray telescope.

Imagine if you replace the stationary coded aperature mask pattern with a transparent LCD screen, ie like this one:

formatting link

If a space rated version is put in space, between an optical telescope and the object being viewed, it could have the effect of vastly improving the resolving power of the telescope by cycling the pixels from transparent to opaque in a repeatable pattern. Also optionally colour filtering could be done if using a monochrome CCD on the telescope.

The distance from the telescope to the LCD coded aperature depends on how wide the LCD is and the pixel size, and the angular resolution of the telescope, but it would be effective over a large range of distances.

The simplest way for it to be practical would be to have the telescope and LCD having no orbital velocity, and either falling towards, moving away, or stationary in relation to the sun.

If one or more LCD's are placed say 1000km to 10000km away from the telescope, they could each cover a portion of the telescopes field of view and move into position on the surface of a 1000km to 10000km diameter sphere around the telescope. As long as the LCD, once in position between the telescope and object being viewed, maintains radial position within half a pixel on the LCD (ie 0.1mm approx) it would work fine.

If the LCD is 1000x1000 pixels and at a given distance from the telescope, masks a viewed object that appears 500x500 pixels when viewed from the telescope, then even if the resolving power of the telescope only can see a 10 pixel blur of the object, by cycling the LCD pixels in a repeatable pattern (with full blanking for clock timing with the telescope), then over time the telescope can resolve a 500x500 pixel image of the object with some computer processing.

The limit to resolving power is in how many pixels are on the LCD.

cheers, Jamie

Reply to
Jamie M

you'd have to make it optically flat and deal with diffraction artifacts from the grid.

--
  When I tried casting out nines I made a hash of it.
Reply to
Jasen Betts

At least one was proposed (not sure if it was ever built) that consisted of a 1-D quadratic residue mask in front of a wider block of detectors. The spinning satellite would then image the object in radial slices.

They were all big blocks of scintillator with photomultiplier detectors. The initial design was complete rubbish designed at Saclay in France where they put their atomic bomb makers out to grass.

formatting link

It was actually for gamma-rays.

They tried to use maximum entropy to tart up the results but basically as published the thing lacked adequate discrimination and had ghosts.

It wouldn't make a blind bit of difference to the resolving power of the telescope which is determined fundamentally by the longest baseline that it correlates photons over (typically diameter of its optics).

The aperture would have to introduce a uniform phase delay if it was to work at all. Mainly coded apertures allow you to alter the depth of field after taking the image data but when everything you are looking at is effectively at infinity this is not an advantage in astronomy and merely degrades signal to noise with no benefits at all.

--
Regards, 
Martin Brown
Reply to
Martin Brown

Hi,

Imagine a simple case of a single pixel detector, and an LCD coded aperature with 4 pixels (square 4 quadrant detector)

pixel detector----(1000km)----(4 pixel LCD)---(10light years)----(star)

If the star when viewed with the hubble telescope, has a huge sunspot on the upper left of it's visible hemisphere, then the single pixel detector in conjunction with a 4 pixel LCD that has 3 pixels active at a time alternating the dark pixel, will produce a brightness signal at 1/4 the frequency of the pixel step frequency, as every 4th change in the LCD pixel configuration would darken the upper left pixel, which would mean the rest of the star is not darkened by the LCD and thus brighter when viewed by the single pixel detector.

It would be a lot more efficient to use an actual telescope and a higher pixel LCD but the concept is the same, except for the specific pattern of active/blanked pixels would depend on what part of the image is desired to be resolved more etc.

There is no phase delay issues it is just simple light masking, and would increase angular resolution of a telescope.

The brightness seen by the telescope would be reduced compared to not using the LCD, due to small transmission loss through the transparent screen, as well as the percentage of masked pixels. Noise floor considerations on the telescope would be a limit to resolving power, since even a single pixel telescope with sufficient light detection could resolve with an LCD coded aperture.

cheers, Jamie

>
Reply to
Jamie M

I think it would only need to be optically flat enough so that the light from the LCD hits somewhere on the telescope, ie +-0.0001 degree? (depends on the aperture of the telescope and distance to the LCD)

As since the LCD aperture is software controlled the error in optical flatness can be compensated for as long as some part of the telescope gets hit by the light from the LCD.

cheers, Jamie

Reply to
Jamie M

Congratulations. You have reinvented the pinhole camera.

You really are utterly clueless. The incident wavefronts must remain coherent through the optical chain to make a diffraction limited image.

The moon dark side occulting stars and radio sources has in the past been used as a means to determine their position. Back in the days when radio telescopes all had very poor resolution. These days star positions are known so well it is used by amateurs to determine the heights of mountains at the lunar limb in grazing occultations.

A very lucky occultation of a fairly bright star by Titan gave the first hints that it had an atmosphere in 1989. There is a video of it (link seems to have vanished). A diffraction peak at mid eclipse was predicted but it was way brighter than expectations from a pure geometrical disk. The atmosphere boosted the signal.

formatting link

Looking for microlensing by compact objects passing in front of distant stars is an active area of research.

--
Regards, 
Martin Brown
Reply to
Martin Brown

Hi,

A pinhole lens camera has a single small aperture and projects the image onto a CCD, not sure how you confuse that with the 4 pixel coded aperture I described.

Assuming the waveform isn't coherent is just ignoring the main idea of using a variable coded aperture to effectively increase angular resolution of a telescope, in favour of irrelevent details. Here is a simple fix if you are too lazy to imagine a sufficiently flat LCD masking screen to keep phase error low enough:

Replace the transparent LCD with a laser cut stainless steel sheet, with concentric cutouts in the same idea as a DVD optical disc. The cutouts will pass the light from the star with no impact on the light as it is travelling through free space. Rotate the stainless steel sheet on its axis, axially between the telescope and the object(s) being imaged to have the effect of creating a repeatable variable coded aperture.

cheers, Jamie

Reply to
Jamie M

Hi,

I think if using a spinning disc with cutouts for the variable coded aperture, to have an optimal placement of cutouts, I think

2D golomb rulers, or Costas array's would work:

formatting link

formatting link

cheers, Jamie

Reply to
Jamie M

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.