Back Ground Information:
One of the important tools in the modern dental clinic is a fairly strong and concentrated light source which is used to cure (harden) light-activated resins. After placing the malleable plastic resin where it is needed the dentist puts the tip of the light gun near the resin and turns it on for an appropriate period of time. It is important that the correct level of light be applied and for the correct period of time. If the level of light is too low the resin will not harden properly. If the light is too intense or applied too quickly the resin may shrink or change shape and not make the proper seal with the tooth surface. As a way of monitoring the intensity of the light output of the gun most of the more expensive light cure units incorporate a light sensor that is used before each use to check that the light intensity is correct.
If you would like more information on the light-activated resin curing procedure here is a link:
https://en.wikipedia.org/wiki/Light-activated_resin
Here is where I come into the picture. As I was salvaging a circuit board for one of these cure lights I came across the TSL230 sensor IC being used as the light quality sensor.
Here is a data sheet for the TSL230 Light to Frequency converter:
http://www.ti.com/lit/ds/symlink/tsl230a.pdf
When something interesting like this TSL230 comes from a salvaged board I like to hook it up on a bread board and run a few experiments to get a physical feel for how the circuit works. The data sheet tells us that the TSL230 takes a light level on its photo diode and converts it to a pulsed output. Further the frequency of the pulsed output is dependent on the intensity of the light. The greater the intensity of the light level the higher the frequency of the output. The TSL230 also has programming options that allow the sensitivity of the input to be chosen between OFF, 1X, 10X, and 100X. The output is similarly programmable with the frequency output scaled with dividing by 1, 2, 10, and 100. It is noted in the data sheet that the unity selection produces an output that are pulses. The divide by 2, 10, and 100 selections are output as 50% duty cycle square waves.
For my experiment I have selected the 10X sensitivity level and a divide by 10 level for the frequency output. My light source will be a 6 volt miniature incandescent bulb. By controlling the voltage to the bulb we will be able to control the light level falling on the sensor. It will be important to shield the sensor from ambient light of the shop lights. I have found a black boot that will cover the TSL230 and adapt to the 6 volt bulb so that practically the only light getting to the sensor will come from the 6 volt bulb.
Here is a picture the bread board wired up and also a picture with the boot and bulb in place over the sensor
The circuit will be powered with the recommended 5 volts. We will adjust the voltage to the bulb with a variable power supply and the voltage will be monitored with a Fluke Multimeter. The output of the TSL230, itself, will be measured with an oscilloscope and a small frequency counter. You will probably recognize the frequency counter from one of my previous blogs.
This is the baseline with only 15 mV on the bulb and the output showing 13 Hz.
To properly illustrate here is a short video of the output of the experiment displayed on the oscilloscope and the frequency counter as the voltage, displayed on the Fluke, is increased to the 6 volt bulb. The Oscilloscope will initially show no movement as we are out of range on the low frequency end and, eventually, as we approach full brightness, the resolution will again be lost on the high frequency end. This is because I lack the hands necessary to change the horizontal time on the scope as needed. I have chosen a middle of the road setting for the time base. The frequency counter will better reflect the entire range.
This has been a very simple experiment but it gives a good illustration of the capabilities of the TSL230 chip. Further experimentation will allow us to explore the other sensitivity settings and the other output frequencies. In a separate experiment using the 100X input sensitivity it was noted that even with 6 layers of hand towel over the sensor it could still detect the presence or absence of my hand above the sensor shading the stacked up towels.
John
Top Comments