The prior page describes the apartus and circuitry for testing infrared detectors. While preparing to test, a couple of interesting issues occurred that you are also likely to encounter. Specifically, I wanted to prove that my emitters were bouncing light off the target and into the detector, but I couldn’t see the infrared beams.
Human beings cannot see infrared; electronics can. One way of viewing infrared is to use a digital camera. Infrared usually shows up as a light pink or purple on the screen. Unfortunately, it is too faint to see on the test target when the target is positioned more than 10 cm or so away. I had to use another approach.
Below is a view of the detector surrounded by an aluminum baffle to prevent the light from the emitters from reaching the detector directly. The infrared light should bounce off of the target and back into the baffle, in a photoreflector configuration. Here is a good article that explains more about infrared experimentation.
Detector with aluminum baffle surrounded by emitters.
Instead of using invisible infrared emitters, I began with ultrabright red LEDs (Lumex SSL-LX5093XRC/4 from Digi-Key) with a similar output angle (25 degrees). Notice there is one LED on the top, one on the left, and one on the right.
Top emitter is angled
Oops. Even though the top LED appears to be flat against the black plate, it must be angled slightly downward. You can see that the red pattern on the target shows the LED light below the other two. Furthermore, it is leaning towards one side. This demonstrates the difficulty in being sure that the infrared emitters are going to be pointed in the expected position on the target.
The red LEDs also show another issue.
Emitter light barely visible at 100 cm
The emitter light is significantly more spread out on the test target when the target is placed furthest away. The light is only slightly visible, particularly in comparison to how bright it is when the target is nearer. This is expected of any light source; the brightness decreases as the square of the distance.
Another problem is that the center of each light is dim. As technology improves, newer LEDs will have a more even distribution of light.
Worst of all, though, the top emitter with the bad angle has now reached the bottom of the target. This means less illumination on the target and some highly undesirable illumination on the ruler. This could result in “false” detection signals where the infrared detector is sporadically seeing reflections off the ruler nearer than the white target.
Interestingly, the red LED’s wavelength provides sufficient infrared overlap with the infrared detectors that both the PNA4602M and the TSOP4038 operate with red LEDs. At 20 mA (ten times the IR test current), the Panasonic sensor detected at 100% up to 35 cm, and the Vishay sensor detected up to 83 cm. This means that if necessary, you can use red LEDs in projects where seeing the illumination on the target is beneficial.
The purpose of the aluminum baffles is to prevent modulated light from reaching the detector directly. The baffles are 5 mm thick and are carefully machined. The screw holes provide enough slack to affirmatively push the blocks together before tightening. Therefore, I was highly confident that no light would leak between them.
The baffles are separate to allow the top portion to be removed to test the effects of overhead lights. In hindsight, I should have made the entire baffle out of a single solid piece.
During initial testing with currents higher than 2 mA, the detectors tripped when no target was present. Due to all my years of working with reflective detectors, this immediately suggested leakage of illumination or electrical signals (from the 74AC14 oscillator). I swapped in blue LEDs to determine if the circuitry was the source, since the electrical noise would be just as strong but the no infrared light would be released. The false detections disappeared. That meant I had an illumination leak.
So, I took out a green laser and pointed at the edges and cracks of the aluminum baffle.
Green laser shows light leakage in aluminum baffle
Oops. Light is obviously able to pass through pieces of the baffle.
Aluminum tape seals light leakage on baffle
Aluminum tape is placed everywhere the baffle pieces touch each other or the black plate. This type of tape consists of real aluminum metal (like thick aluminum foil) backed by strong adhesive. Unlike masking tape or plastic electrical tape, the aluminum isn’t going to be partially transparent to light. This solved the leakage problem.
To end the article, we'll next look at how I measured the duty cycle of the infrared detector outputs when they were somewhere between 100% and 0% detection.