IRAC Instrument Development Model

First Look at Stray Light Suppression

(Last edited: February 3, 1998)

For all the following images, the Z value is the focal length of the AutoCAD 'camera' generating the image. The shorter the focal length, the wider the field-of-view. The Z=2.5mm focal length images cover a field-of-view of about 150 to 160 degrees.


Views from the center of Detector 3

Views from the center of Detector 3 looking toward the center of the Channel 3 Pupil stop. Each small (4-10K)image shown on this page is a hyperlink to a larger version (27-105K) of the same image with notation added to the image. As I reviewed the images, I was impressed by how much of Detector Assembly 1 could be seen from the center of Detector 3. Baffles might be added around each detector assembly to prevent their view of each other.

Also, a single reflection path off either wall of the Housing-Optics-Bench can get past the entire optics assembly without passing through any lenses or filters. Any light traversing this path will have perhaps two orders-of-magnitude less attenuation than light going through the lens/filter system. It is suggested that baffles be added which can eliminate this path entirely.

[Z=20] [Z=10]

[Z= 5 ] [Z=2.5]


In order to assess the potential light leak past the Housing-Optics-Channel (Green, above) which does not pass through the Lenses, Beam-Splitter, or Filters, views were created from the front of the Aspheric Lens looking toward Detector 3.

[AV] [SH]

[AV] was rendered with AutoVision. A lot of edge detail was lost because all the Housing-Optical-Bench is the same color. The base of Detector Assembly 1 can be seen to the left.

[SH] was rendered with AutoCAD Shade and shows all the edges. The edge detail shows the clearance notch in the Housing-Optical-Bench at the lower right.

What is very apparent from this view is that the Lens Assembly is a circularly symmetric object and it does not block the corners of the "square" aperture in the Housing... Detector 1 could not be seen directly, so the view direction was aimed directly at Detector 1 and the view point was moved closer to the gap between the Lens Holder and the Housing...

[D1a] [D1b]

[D1a] is looking directly at Detector 1 from the same position as the viewpoint in [SH].

[D1b] is looking along the identical path as [D1a], but moved 1/2-inch closer to the gap.

The tiny gap, the close proximity of the Lens Holder and the Housing, is what is seen in all the orthogonal top and side views usually presented and supports the first impression that this gap is not very big. Even in [D1b] it is very difficult to tell what part of the Detector 1 Assembly is visible.

Refering back to [SH], the corners are potentially very wide open. In the next views, the view direction is still toward the center of Detector 1, but the viewpoint is moved 1/2-inch vertical, toward the lower left corner in these views.

[Z=5mm] [Z=20mm]

[Z=5mm] shows the detector can be seen directly through the gap from the corners. It is not blocked by the Housing-Optics-Channel (green).

[Z-20mm] is a narrower angle image showing the details of the Detector Assembly 1 with more than half of Detector 1 directly in view. The support arms are also seen. If a cover is added over the support arms, it will still have an opening to this path unless the cover is in virtual contact with the Filter/Pupil Stop Assembly.

The best place to close this path is at the Lens Assembly/Housing-Optical-Bench gap. This could be done with a labyrinth around the Lens Assembly so there is no physical contact with the lens itself. Since this is a non-moving opening, it can be sealed better than the opening between the Housing-Optical-Bench and the Calibration Mirror/Shutter.


In order to assess the mutual view of the detectors for each other, I set-up a series of views from a closest corner of Detector 3 looking to the center of Detector 1. It is possible to look backward from the corner of the detector and see the Detector Carrier Plate and the attached electronics. For this model I had deleted all items aft of the carrier plate in order to reduce the size of the drawing from 20 MBy down to 4 MBy. That is why the Support Arm can be seen behind the plate. The detectors can't quite see each other, but they can see the majority of each other's mounting assemblies, including parts of the Detector Carrier Plates. I think the modeling of the edge of the active detector is not correct. There should be come sort of physical buffer around the pixels at the edge of the detector. Also, are the pixels physically on the surface of the detector or is there a transmission cover element between the detector and the active pixel area?

[Z=2.5] [Z=5]

[Z=10 ] [Z=20]

There are light leak paths past the base of the Filter Mount which will allow light to get to the detector without going through the Pupil Stop, identified as A (around the flood lamp opening) and B (through a cutout in the Housing-Optics-Channel web on which the Filter Mount is bolted). The leak at B can be eliminated. There may be similar leaks on this and the Channel 1 Filter Mount. It should be possible to eliminate the path around the Flood Lamp opening by adding a flange to the Filter Mount or the Flood Lamp housing. Sealing this opening may be difficult.


In order to determine the objects seen throuth the A and B leaks mentioned above, the Beamsplitter Assembly was modified to different colors and installed in the drawing. The views below are the same as [Z=20] above, without the Filter Holder [WOFH] and with the Filter Holder [WFH].

[WOFH] [WFH]

The object viewed through B was the Filter for Channel 1, not shown in [WOFH].
The objects viewed through leak A are (from left to right):


[FC3-7.5mm] [FC3-20mm]

The [FC3-xx] views are directed from the tunnel in the wall of the Flood Lamp that lets light reflect from the ball bearing to the Detector. These views are of Detector 3 from Flood-Cal 3. This is a very complicated opening path and there are gaps past the Flood Lamp to the interior of the Beam Splitter volume. Modeling this aperture in APART for light leaks is virtually impossible. The seal around this opening needs to be reviewed with the hope of making the seal more complete.


Views from the center of the Aspheric Lens toward the Pick-off Mirror

[Z=40] [Z=20]

[Z=10] [Z=5]

[Z=2.5] [Isometric View]

The Pick-off Mirror is golden and centered in each view. The Calibration/Shutter Mirror is dark green (at the right of each view). The Pick-off Mirror Support Arm is blue. An odd looking channel cut in the wall of the Optical Bench web in front of the lens can be seen in the lower left of Z=5 and Z=2.5. I used the Isometric View to confirm the presence of the channel. I don't know what the purpose of this channel is. Nothing seems to be near the channel. There are several other similar channels elsewhere.

The Calibration/Shutter Mirror will reflect whatever comes from the left of the Pick-off Mirror toward the lens. That is the direction of the MIPS 50-100 and two 70 micron array entrances. When the calibration mirror is closed for calibration, the Optics-Bench wall surrounding the mirror can be illuminated by the diffuse source. This diffuse light can then get directly past the lens and optics and reach the detectors either directly or in a single bounce.


[Hidden Line] [Top View]

The Hidden Line view is a perspective view with a clipping plane through the centerline of the Lens. This shows the nearly 180 field-of-view seen from the center of the lens, looking toward the Pick-off Mirror.

The Top View is an isometric projection with the potential stray light paths labeled. From near Detector 3 at "A" there is a potential path ABCD or ABCDEF and Abcd which involves only a single scatter to get past the lens and filter optics. From Detector 1, the path is bcd.


Back to EE - Home Page