Test patterns: QVC's beta team examines the Buhlite and Philips CDM technology


In the summer of 1997, I was hired by Chuck Lester, the lighting manager of the QVC shopping network, as a lighting engineer. I was eagerly looking forward to assisting in the construction of QVC's Studio Park complex in West Chester, PA, covered in the April 1998 issue of Lighting Dimensions ("Purchasing power," page 40). As I walked through the new facility, then just a warehouse, and reviewed drawings of dimmer layouts and preliminary wiring schematics, I began to realize how enormous this project would really be. The plans included more than 1,400 dimmers and 1,500 instruments, triple the lighting inventory of the previous studio space, for live broadcasts 24 hours a day, 364 days a year.

As we progressed, several things became evident. Manpower and maintenance, especially lamp replacement, became very big issues. The cost of replacing approximately 75 burnouts per week could not be overlooked. We had also noticed a frequent pop, or in some cases a large bang, as lamps would go out in the studio--especially troublesome during live broadcasts.

We began recording when lamps were placed in an instrument and when they were taken out. This data was sent to the lamp distributor for evaluation, then forwarded to the lamp manufacturer, Philips Lighting, for further study. The result was a partnership between QVC and Philips' Jeff Van Etten (East Coast sales OEM division) and Richard Scott (project application engineer), who were intrigued by the information they received. We began working together to solve the problems of premature and noisy burnouts, and discovered that lowering the gas pressure in the lamp envelope and changing the sensitivity of the fuse alleviated the popping noises and lengthened lamp life. We also began testing some of Philips' new 575W prototypes and assisting them with updates on the 1k EGT and 2k CYX lamps.

At the same time, we noticed reflectors beginning to fail on some of our Mole-Richardson cyc units. It was obvious that the 24/7 environment was taking its toll. But rather than complain to manufacturers about their products, we created a team to achieve solutions. We started a beta-test program with several of our product manufacturers (Mole-Richardson, Martin Professional, Philips Lighting, and others), to accurately track problems and failures, share data, and test new products.

Mike Parker from Mole-Richardson explains, "Mole-Richardson and QVC are working together to provide industry professionals with quality products able to withstand the demands of the 24/7 environment. For example, we have teamed up to initiate a minor design change in their Cyc-Strips by adding a narrow ventilating slot that added substantial life to the reflectors. Product design and problem-solving to meet the needs of lighting professionals is what this team is all about."

Says Will Bauer of Martin, "QVC has played an important role in the development of the Martin Lighting Director [MLD] product. Testing by QVC resulted in the finding and fixing of one very obscure bug related to Compaq computer power-up sequences. Additionally, Doug Rae of QVC has been very helpful with suggestions for improving the product. It was his original suggestion that resulted in the creation of the 'DMX Master' feature now released in version 2.4 of our software. We are looking forward to wo rking with QVC in the testing of other new MLD features, such as 40kHz operation, presently under development."

QVC is constantly searching for companies who have products that reduce operating costs, decrease manpower, address our unique needs, or challenge the mainstream in lighting. Our program has become invaluable for locating and defining products that can endure our 24/7 environment. Jeff Van Etten concurs. "The beta-test program that QVC offers has been of tremendous assistance to Philips Lighting. There is nothing like true field experience and immediate feedback to measure the results of your product. We have been able to make several changes to enhance the performance of our halogen line due to our findings through this program. In addition, the beta tests are a great way to introduce and evaluate new lamp systems like our CDM technology."

In one of our early meetings, Jeff had mentioned that Philips was developing a new product line, MasterColor CDM lamps, which combine the color stability and long life of high-pressure sodium lamps with the white light and high output of metal-halide lamps (CDM stands for Ceramic Discharge Metal-Halide). Up to this point, my only experience with metal-halides had been in automated and HMI PAR fixtures, which have their limitations. They can only be used in situations where a color temperature over 5000K and a lamp life of approximately 750 hours is acceptable.

CDM technology provides better color control (a color temperature of 3000K or 4200K +/-200 throughout life) and a CRI of 80-85 with a long life of approximately 10,000 hours and growing as tests progress. The MasterColor line includes 35W, 70W, and 150W versions at both 3000K and 4200K. With a high LPW (lumens per watt), the CDM-150W is close to the output of the standard quartz 1,000W lamp. It does have its drawbacks: As of now, if a power interruption occurs, the lamp will go out and needs time to cool before restarting, which can take up to 10 minutes. Also, due to the technology, dimming is not recommended.

Jeff gave us the names of several companies experimenting with this new lamp technology. One stood out--Buhl Industries, a well-established manufacturer of presentation equipment including LCD data video and overhead projectors, presentation accessories, and video-teleconferencing equipment. The Fair Lawn, NJ-based company has also purchased General Electric's Photo Lamp Socket product line and remains a dominant supplier of halogen and metal-halide sockets for the OEM lighting and projection industry in North America.

Buhl also manufactures electronic power supplies, which are used in most metal-halide OHP and LCD video projectors. The company has spent 12 years working with metal-halides in the projection and imaging industry and has combined its years of experience with CDM technology to produce a new line of professional lighting fixtures, called Buhlite. The line is comprised of fresnels, softlights, and profile spots available in various wattages up to 150W.

Our initial contacts from Buhl Industries were Matthew Kyhl (vice president of sales and marketing) and David Kyhl (vice president of product design and development), who were very eager to supply us with some samples of their early prototypes. I asked them to bring the 150W version of each style of fixture. Not totally understanding the capabilities of the units, I couldn't see how any fixture of less wattage could produce enough light.

From our first meeting, I could tell that Buhl was not just another lighting manufacturer. The company had done its homework and presented us with a product with very creative features. On first examination, the instrument reminded me of an early version of a Times Square Lighting fresnel, but that's where the similarities ended.

The fixture contained a 150W Philips CDM MasterColor lamp. The ballast supplied by Reliable was neatly tucked within the unit. I wanted to control the dousing of the unit remotely, so Buhl supplied me with the DMX-controlled dousing fixture, Model S150-D. The fixture, finished with a black powder coating, came complete with a lamp, gel frame, barndoors, and Edison connector--basically plug in and go. We plugged it in and waited approximately five minutes for the lamp to come up to temperature. We did some initial tests checking the unit to specs, comparing it to our other conventional fresnels and checking colors as well as black and white on camera, all with surprising results. I was no longer a skeptic; I wanted to see more.

I discussed with Chuck Lester the possibility of installing several of the units in our Show Kitchen studio. This studio is outside our main area but is heavily used, by guests like Emeril Lagasse, Sophia Loren, Richard Simmons, and George Foreman. Its area is approximately 20'x20' (6x6m), with a ceiling height of 12' (4m). Chuck agreed, and suggested that we install a complete Buhlite system.

As we surveyed the Show Kitchen, Buhl agreed to supply us with the necessary lighting. I was a little concerned--we had seen what the light could do in a small testing area, but were we ready to move it to a live studio? Matthew Kyhl mentioned that Buhl had just hired former CBS LD Lincoln John Stulik, and discussed the possibility of him designing the lighting. Stulik had been working with the lights for a month before our request and was quite confident they would fulfill our expectations. Within a week he was at QVC surveying the studio and plotting a design.

With particulars still to be worked out, we proceeded with the project. First, the power. We had forty-eight 2.4kW dimmer circuits in the grid but no conventional power. I decided to pull various dimmer cards and insert breaker cards to power most of the units. Our biggest fear was the lack of hot restrike--in a live studio situation, we cannot afford 10 seconds, let alone 7-10 minutes, without lights.

To solve this we ran a dedicated UPS circuit to power the remaining key lights (Buhl does supply a complete UPS back-up system to alleviate this problem). For DMX to operate the dousers, we chose to use the existing standard DMX controller mounted in the Show Kitchen studio. Our plans were to break up the light into four groups: keys, backlights, wall wash, and effects. Stulik's plot contained 21 of the Buhlite fixtures: four S150-D softlights, seven F150-D fresnels, five S70-D softlights, four EPS-70-D ellipsoidals, and one EPS-35-D ellipsoidal.

With just some quick unpacking and fastening a C-clamp, the fixtures were ready to hang. All cable connections and dip switches are conveniently located in the same area at the rear of the fixture. I made some quick DMX assignments with our ever-handy little screwdriver (the units are factory-set to 0) and up they went.

The Buhlite fixtures weigh slightly more than conventional lights, ranging from 5lb (EPS-35-D) to 20lb (S150-D)--still light enough to handle up a ladder. The yoke on the early unit was constructed of thin bent steel held by a single screw on each side. The new yokes have been upgraded to a heavier gauge with a locking knob. The unit is well balanced and required little tightening to keep in place.

(Side note: During the beta-testing of the F70-D fresnel, with the lamp on, the fixture fell from a stand approximately 8' [2m] high, bounced off a case, and landed on a cement floor. To our surprise the unit never blinked--its lamp, optics, and DMX controls worked as good as new. With only a minor dent and scratch we quickly placed it back on the stand and tightened it down.... Sorry, Buhl.)

Within a couple of hours, all of the lights were properly placed. We had several discussions on the best way to run power to each one individually. All units came with an AC outlet to daisy chain, but our concern, going back to the restrike issue, was that if the breaker failed, it would turn off all lights in the chain. We decided to do two things. Instead of a daisy chain, we took all lights back to a common power source, which made it easier to replace or add a single unit without breaking the loop, and limited the number of lights to two or three per circuit.

As for DMX, we used the in and out receptacles supplied on the unit. Buhl recommends a termination plug in the last unit just to be safe. Upon firing up the fixtures, all the DMX dousers closed to their default position. As I raised the faders to open them, I noticed that familiar blue-green glow (with a slow transition to a smooth white color) of the CDM-150 lamp heating up. As I continued to raise and lower various faders, I found the best I could achieve was a slow fade; it was impossible to do any type of quick blackout let alone an on/off chase. I did find, however, that the fader-to-light output ratios were precise.

Several other small things concerned me about the DMX control. Changes in the DMX dip switch assignments would only take effect after the unit was reset. Unfortunately, this extinguished the lamp as well. I did notice that when the DMX was interrupted, the dousers held their position (a nice feature if your console should go down). The douser can be operated manually via a lever on the top front of the unit without using DMX. But in this mode if AC power is turned off all dousers will return to their default closed position when power resumes. David Kyhl says, "These are adjustments that can easily be made depending on your application."

Stulik started focusing with the key and backlights, the S150-D and the S100-D softlights. The softlight diffuser kept the light nicely on position and created illumination that was soft and smooth to the edge. The reflective barndoors added back extra light where needed, a nice combination. With just a couple of tweaks, we walked across with the light meter and read an even 160fc. Next, we added in the wall and cabinet washes using the F70-D and F150-D fresnels. Again, the units provided a very smooth and easily controlled light, with the barndoors keeping the light contained to the focus area. With the F70-D, a little double imaging was noticed with the doors very tight. Washes were enhanced by the control of spill from the S150 key lights.

Gels can either be installed in front via the dual clips or through a slot provided in the barrel, depending on your application. The beam and lamp adjustment knobs made it easy to keep the varied beam widths even. I did have a little problem getting the lamp's hot spot to center, however: The adjustment knob, located in the rear of the fixture, gave me a nice throw in and out of the reflector, but gave me no adjustment side to side. Having a fixture adjustment handle in the way wasn't helping much either, though it came in quite handy otherwise. Buhl reports, however, that the whole lamp adjustment assembly is being redesigned.

Changing the CDM lamp is quite easy, though with a 12,000-hour life you won't be doing it too often. A simple loosening of two screws and out slides the complete lamp assembly. The lamp is held tightly into its socket with two spring clips.

Using the EPS-70 ellipsoidal, we finished up by adding various gobos using standard metal patterns. The zoom two-knob focus/beam-width control was a little shaky but adequate to find the size and focus desired. The optics were strong and easily cut through the existing wash. With the unit's reduced heat, the pattern was easy to move and change, and with only a 70W lamp, these won't need to be changed for a long time.

With the studio completed, both hosts and guests noticed an immediate difference. We had started off with curiosity about CDM lamp technology and ended up finding a new, well-rounded fixture. Our goal at QVC is to maintain a consistent light source for broadcast and keep costs for operating and maintaining the equipment at a minimum, and the Buhlite falls well within our plan.

We had previously used a 200A service to power the kitchen lighting system; we are now using less than 30A, running the complete studio on less than one-fourth the power. With the reduced heat from the lights, we also have huge savings in air conditioning costs, not to mention the added comfort to our hosts and guests. Our standard 1k lamps average 750 hours--the CDM lamp averages 12,000 hours, resulting in savings in manpower and replacement lamps.

The Buhlite, burning continuously in the QVC Show Kitchen since March 17, will also be widely used in commercial markets. Buhl plans to expand the product line into lighter, portable fixtures, bringing the Buhlite to location applications with plug-and-play convenience.

Our beta-testing program has now expanded to include products from other companies, with data confidentially released to them. We are developing a formal information packet to give to manufacturers at the completion of each evaluation, and we enjoy working with companies whose goal is to improve the lighting industry and its products.

The author can be reached at [email protected] or phone 610/701-6756.