Screen Setup

The screen was set up by clamping it between two rectangular posts on each side. First, two posts were set up with 22 in. between them (thus allowing the screen to span a total distance of up to 24 in. when clamped down). To best stabilize the screen and to allow for it to be "pulled" taut by exploiting the give in the L-bracket, the L-bracket was bolted on the outside of the post, along the same axis as the screen itself.

On first attempt, the screen was too thin to be fully clamped between the posts. In order to have it fit snugly, sections of heat shrink tubing was used as a shim at the points where the posts were clamped together. The tubing was slid into the track of one optical post at the desired points. In order to accommodate the shim, the two posts had to be held together, ideally clamped, with the screen and shim in place. Then, the post clamps could be slid into the tracks of the post, moved to the optimal location, and tightened down. This required at least three people: two to tighten one side while the third holds up the screen on the other side. The screen was placed ~1/8" from the edge of the posts and flush with the top.

Camera Stage Setup

Once the screen was in place, the camera stage was set up by placing the XY-Translational with Rotation stage on four 3" optical posts. Then, the z-axis stage was placed in the center of that, with another 3" optical post on top, which was then topped with the camera. This was set and clamped down ~22.5" from the screen. This was about 1" closer than expected based on our theoretical models.

Fine Alignment

We used the visible camera to fine align the screen and to test the setup. Notably, the visible camera is placed below the infrared and thus requires a calibration in order to ensure the two are aligned on the computer image. This can be set by hand using the FLIR proprietary software (FLIR CamWeb) and adjusting the "MSX alignment". The image mode "Thermal MSX" allows both the visible and IR camera to be displayed at once and the difference in their positioning can be seen. We found an offset of ~0.5m to be nearly accurate (note: using this method, although you can get more accurate than this, the displayed value only has one significant figure).

In order to align the camera, we first used the exposed top edge to judge whether the camera was appropriately centered on the screen. We set the rotation as close as possible to being in line by eye, then adjusted the y-axis until the gap on both corners was a similar size, thus indicating that the rotation and y-position were correctly set. Rotationally, the camera required only a refinement of -1/2 degree. The y-axis is set at 1.25. Then, the camera was pulled as far from the screen as possible using the x-axis to allow the screen to be easily centered using the z-axis. Once the outlined test mass was centered, the x-axis was used to bring the camera close until the screen just barely filled the field of view. The x-axis is at 2.25. The z axis is set to it full dynamic range at 10. Unfortunately, the camera is still slightly too tall for the screen, likely requiring the purchase of a new optical post about 0.5in shorter the current one. This interchange will likely require a new fine alignment after.

Basic Imaging Tests

The camera was also focused on the screen based on the manufacturer's printed distance on the camera itself (using 22.5", or 0.572 m). Using the FLIR proprietary software, the camera appears to be in focus in IR (a hand was used as a good focusing tool for this). Additionally, the camera does pick up heat on the other side of the screen. A hand can be lightly seen warming the screen, as can a soldering iron tip. This was a very imprecise visual tool, but does indicate that the camera and screen are working roughly as expected.

Next Steps

A new optical post that is ~2.5" tall should be ordered to replace the one under the camera currently. The heating system also needs to be ordered and set up. Currently we are debating between a parabolic reflector with a hole in the back, and one without, as each would require a different mounting mechanism for the cartridge heater.