Ultraviolet ( UV ) light is used extensively to cure (polymerize) paints and other coatings, often in a conveyorized process.

The UV is usually produced by extremely bight "arc" lamps also rich in visible spectra (they make a lot of white light & heat).

Most UV sensors are laboratory-grade units too sensitive and too delicate for these processes.   They typically integrate light intensity in mW/cm² or uW/cm² (milli- or microWatts/square centimeter).

This robust, chrome sensor (shown with a Super M.O.L.E.©* from ECD) can accurately measure UV light intensities in Watts/cm² !

Successful products solve a problem, which in this case involves the arc lamps that make the UV light.   They typically contain mercury and rare gasses in a quartz envelope.   In time, the mercury "plates out" on the inside of the envelope, reflecting the UV light back into the bulb, which heats up.   The UV output decreases and the Infrared component ( IR ) rises.   The lamp still seems bright, but it's no longer doing its job.   UV curing cannot occur; the process makes "potato chips".   Industry needs a way to know when to relamp the conveyor.

The design goal was an accurate, affordable UV sensor that could withstand the heat of the curing process and interface to an existing datalogger through its thermocouple signal-conditioning circuit (extremely sensitive to low-level signals).   It had to be self-powered and flat, to fit the process.   And it was to feature a "cosine response", which basically means that it is equally sensitive to light arriving from any angle.

Ultraviolet light looks "fuzzy" to us because it begins at the blue end of the visible spectrum and extends beyond our range of vision.   The first UV research was accomplished by the medical* community , which categorized it by its ability to penetrate our skin, with names like UVA, UVB and UVC.   In time, other uses generated other names:

Electronic engineers usually discuss light by its wavelength (1 / frequency).

UV   light sources for curing have strong "spectral lines" at 256nM and 365nM (among others), and thus most UV-curable compounds are based on chemical reactions triggered at those wavelengths.   We needed to center the unit's overall response in that passband.

We chose a sensor with good "blue" response, designed an attenuator to scale the output, and a filter to reject the unwanted (visible) light. There are many filter technologies from which to choose, including thin-film depositions (which provide a tailored, "optical impedance-match") and various glass mixtures and chemistries.

Filters are fun, and the physics of light and electricity are similar (Maxwell's equations).   We needed to accept the desirable light and reject the rest, but this can be done in several ways: reflection, absorption and transmission.   One idea involves "hot mirrors" that reflect the red spectra but pass the rest.   Another technique is "cold mirrors", which work the opposite way.   Once the signal is in the medium (usually glass or quartz), it can either be transmitted to the sensor or absorbed by the medium.

The popular "Blue-blocker©" sunglasses use thin-film techniques; they appear blue because they reflect that end of the spectrum but admit the rest.   On the other hand, red glass is actually "everything but"; it looks red to us because it reflects red light and accepts the remainder.  

Absorption is perilous because the energy can become heat, which introduces inaccuracies and drift requiring compensation.   But transmission is not simple either, because the medium can introduce attenuation, phase shifts and group delays.   And the signal may require additional "optical impedance matching" when it exits the medium.

Of course, while meeting the specifications, we also had to balance the issues of schedule and budget.  

The exact details of the sensor's construction remain proprietary to our Client, but the sensor works extremely well with their line of M.O.L.E.©* dataloggers.

While you're surfing, be sure to visit ECD's website for an update on their broad line of datalogging hardware and Statistical-Process Control (SPC) software products.

* We are not Medical Doctors and the information on this page should not be construed as medical advice.   See your Doctor for medical advice.

* The terms M.O.L.E.©, Super M.O.L.E.© and Gold M.O.L.E.© are registered trademarks of ECD, Inc.

* This work was accomplished by our Principal Engineer as an employee of ECD, Inc. (a W-2 type contract) although he was also actively engaged with our company, with ECD's full knowledge.