Archive for June, 2015

Welding creates a serious challenge to visual acuity. The light emitted from gas and arc welding is intense, and contains high levels of both UV and IR light in wavelengths harmful to the human eye. For this reason, welders (myself included) wear helmets and goggles that utilize filters to reduce brightness, strip away the harmful wavelengths, and protect us physically from welding splatter, which is very nasty. Unfortunately this seriously compromises visibility of the welding task and its surrounding. While the arc itself illuminates the surrounding, the contrast between the arc itself and the area around it is so great that this affords little clarity. When smoke and splatter are included, most welding is done within a very poor visual field. In some case, it is done almost completely blind.

The concept is to use narrow spectrum green light, in this case 530nm Green, to more efficiently deliver visible light through welding glass filters. This increases intensity in the area of the task.

The concept is to use narrow spectrum green light, in this case 530nm Green, to more efficiently deliver visible light through welding glass filters. This increases intensity in the area of the task. It is very difficult to photograph exactly what one sees through the darkened welding glass, and impossible when an arc has been struck.

Most welding glass passes light in a narrow green centered bandwidth, which is why the view through them is green, to the point of being monochromatic. That means most of the light from any task light used that generates white light will be filtered out along with the welding arc emission. That seems inefficient and reduces the effectiveness of the lighting system to the point of being essentially useless.

To address this, over the last three years, I have been working on a task light that delivers a narrow spectrum green light, centered on the emission of the welding filter glass itself. This means that 100% of the light from the task light will come through the glass – a much more efficient approach. You can download a white paper WIP of my findings and concept at: http://www.lumenique.com/New_Lumenique/Files/Narrow Spectrum Welding Light KLW.pdf

An early test mule using three 5W green LEDs and medium narrow optics to create intensity.

An early test mule using three 5W green LEDs and medium narrow optics to create intensity.

This is a work in process. However, so far, with the same energy applied to an identical white light source, vs. a green light source, the amount of brightness visible through the welding glass is doubled.

There have been a few interesting discoveries in this process:

  • The early test mule (shown in the image above), utilized optical reflectors to intensify the beam pattern. I was hoping to amplify the effect of the focused task light into the visual welding field. This actually proved to be less useful than it might look, due to the creation of harsh shadows from the welding gun or torch, so later models have reverted to a more diffused, softer beam pattern, which reduces these effects.
  • LEDs act like low efficiency photo-voltaic sources when exposed to high intensity light. This creates voltage back into the driver during welding work. For the most part, this is not an issue. However, with a few drivers I have employed, this effect causes internal failures (not fully explained). I isolated the voltage from the welding area, electromagnetic effects, and all other factors, before testing the theory that some drivers cannot deal with this by applying a small external voltage to them in operation, which duplicated the failure mode. Now I test all drivers under a welding arc, on aluminum and steel substrate (each emit a different spectral power state), to insure this does not create undesirable results.
  • When gas welding under the green light, I find the appearance of the flame kernel (main heat source) more pronounced, which appears to be from the increased intensity of the surrounding field. This is a happy development, as it increases visibility of the location of that heat source to the weld zone. There is also an enhancement of the colors seen in the weld pool to a small degree I am working toward intensifying further.

I will be working on this more as time passes, so will update this entry as new discoveries are found. Ideally, working with a welding glass producer to create an idealized combination of glass filer and light source, coupled with a hood manufacturer to mount the light in the welding hood itself, activated by the arc itself would create an even more interesting result. The next phase for me is to prototype such an animal for my own use. Stay tuned.

 

While designing cool lighting products is fun and all that, there are other areas of lighting development I am involved with. Whether it is UV curing of resins and plastic parts, inspection lights, or special single spectrum light sources and task lighting, it all comes under the umbrella of lighting for me. In this case, it’s about light measurement, particularly in an easy to use, and simple to set up for gathering data for use during product development, as well as verifying and evaluating design changes in process.

This goniometer delivers a simple to use platform for in-house testing of a wide range of luminaire configurations

This goniometer delivers a simple to use platform for in-house testing of a wide range of luminaire configurations. The meter can be located anywhere from 24″ to 96″ from the optical center of the luminaire.

While large scale, accredited LM-79 photometry demands  the use of expensive and sophisticated test gear beyond the reach of most organizations smaller than a conglomerate, a great deal of accurate data can be gained from simpler platforms. In the past I created a simple desktop Type C goniometer for customers who were creating small light source scale products.

An earlier example of a bench-top system was designed for testing of small light engines and LED optics, shows the same basic configuration in smaller scale.

An earlier example of a bench-top system was designed for testing of small light engines and LED optics, shows the same basic configuration in smaller scale.

Since then, I’ve built others with similar purpose for manufacturers setting up in-house test facilities on tight budgets. Having access to a goniometer, where tests and experiments can be carried out as part of in-house design operations can be a very valuable tool. It is also an excellent tool for quality inspections, and establishing variations on test results obtained from accredited labs.

For this specific instance, the requirement was for a system for testing fixtures that might be as large as 24″ in height, and up to 48″ in length, with intensities ranging from small low power sources to high intensity optically focused products. The design is basically the same as for the desktop unit, but scaled up to accommodate the larger scale of the luminaires to be tested.

Note that this is a horizontal Type C, which rotates the fixture around a fixed vertical axis, as well as the horizontal axis. This is a common approach to general lighting products, and can produce Type B results as well. However, since every test fixture is mounted with the light source aimed horizontally, including downlights, the results need to be revolved in creating usable IES files to reflect the actual luminaire orientation in use. Further, with SSL products, care must be taken to avoid including errors in light output that might result from thermal effects of mounting a vertically oriented product in the horizontal position for testing. However, in the 9 years I have been testing fixtures in this type of lab setup, I have not found this to be of significant concern. I have also devised methods for revolving the output data to create the appropriate IES formatted file for end use lighting application studies.

The other aspect of making this type of lab setup affordable, is the use of inexpensive light meters. While those in the business of accredited lab testing will scoff at the idea of using footcandle meters or hand held spectrometers for this type of application, I have found, in back-to-back testing, the results of tests done in house are within a maximum range of between +2% to -10% of those attained by independent lab testing services. Meanwhile, tests accomplished back to back between accredited labs using the same luminiares, has returned variations of +5% to -8%, while the variations in actual installed applications have been far greater due to the variance in surrounding reflective surfaces, condition of fixtures, variations between fixtures manufactured, and other factors outside the confines of the fixture designs themselves. So, while I am not saying this simple lab gear will replace independent test lab results (it won’t), I am saying that, if the operator is careful about setting up the test, diligent in detailing the data, and verifying his/her results, tests completed in-house, during design and between designs, can be reliable and valuable, and a significant cost and time saving advantage. The single largest variable that independent and accredited test labs bring to the table is consistency in process, and independent non-biased reporting for end user application. This is not always necessary for every test completed during and after designs are completed.

Rotation of the vertical axis is accomodated using a CNC rotary table and ring bearing base.

Rotation of the vertical axis is accomodated using a CNC rotary table and ring bearing base.

 

Fixture mounting is the a second CNC rotary table with T-slots to attache adapter plates.

Fixture mounting is the a second CNC rotary table with T-slots to attache adapter plates.

The meter attachment post can accomodate any instrument the customer might want to use, from simple light meters for quick tests, to more involved spectroradiometric testing.

The meter attachment post can accomodate any instrument the customer might want to use, from simple light meters for quick tests, to more involved spectroradiometric testing.

I have applied a wide range of meters to these types of test rigs. This includes the $100 Probe Fc meters through the more sophisticated Orb Optronix Spectrometer. The more expensive meters do deliver greater fidelity, the ability to capture multiple reading samples for averaging to eliminate error, etc..  However, I have also found that instruments like those I covered in the meter review, all delivered very similar end results. The use of the UPRTek, or Asensetek meters deliver the layer of reading color over angle in addition to standard footcandle readings, which is very useful in LED fixture evaluation. To create a candela distribution table, I use MS Excel and some simple inverse square law calcs.

For this latest creation, I have includes a rail based meter mount, as well as a rail for the vertical fixture platform. This makes setup much easier, in that moving the meter and the luminaire mount along the rails maintains alignment of the two to one another. Rotation of the luminaire in the vertical and horizontal axis is accomplished using CNC mini-mill rotary tables, actuated by remote control. These can be rotated in increments as small as .006 degrees, with 2.5, 5 and 10 being the most commonly used. The vertical axis rotation table is mounted to a large diameter rotary bearing, which can support 600 pounds.

A laser alignment tool insures the fixture rotational center is aligned with the meter sensor

A laser alignment tool insures the fixture rotational center is aligned with the meter sensor

A laser line tool attachment alingns the fixture rail with the meter rail to assure squareness of the setup

A laser line tool attachment alingns the fixture rail with the meter rail to assure squareness of the setup

With the meter post/rail aligned and the fixture center aligned with the meter sensor, the rig is ready to mount and test the actual luminaire sample.

With the meter post/rail aligned and the fixture center aligned with the meter sensor, the rig is ready to mount and test the actual luminaire sample.

This latest rig I also includes alignment tools. One is mounted to the center of the fixture horizontal axis  (a modified rifle bore sight) aimed at the center of the meter’s receptor window. The other (contractors laser line tool) is located on the rail below – emitting a vertical line for checking the zero position of the vertical axis rotating table. With these two in alignment, the rig is set to go. Mount the fixture using adapter plates to the horizontal axis, set the optical source to the center of the vertical axis, light it up and the temperature to stabilize, and start testing.  A typical test for in-house use can take less than 20 minutes after the fixture has reached its operating temperature (2 to 24 hours to taste).

There are other small additional components involved. I personally like to connect the test products to a reliable power source. The easiest way to gain this is using a UPS generally used to connect computers to. They are affordable, and offer much more reliable and consistent voltage output than wall plugs do. I also add temperature measurement (a simple Amp two position meter works for most applications – one for ambient, one for fixture hot spot), and in some case room heaters or coolers to attain a stable ambient temperature where this is not inherent to the lab itself.

So that’s it. An affordable in-house Type C test rig. Not a light source, but related to development of them. I use a similar setup for my own product development, along with a cannon style integrating chamber, a small integrating sphere, and  some other cobbled together test rigs that have proven to be accurate for relative comparison of results to a known standard.