1987 Philips HPI/T 400W

Philips released its second-generation medium-wattage HPI in 1969. With the new lamp design, the Dutch addressed critical color stability and consistency issues that plagued the company’s first metal halide lamp introduced a few years earlier. To that end a three-piece barrel-shaped burner with a well-defined cold spot was devised by Cees Jacobs in 1967 and the evacuated bulb was replaced by one filled with a nitrogen-based gaseous atmosphere. This atmosphere impedes the flow of photoelectrons emitted by the metallic frame, thereby reducing the rate of electrolytic sodium loss from the arctube. In order to mitigate heat conduction and convection losses, the electrode gap length was shortened by 8 mm to 40 mm, which raised the arc power load to 85 W/cm and enabled the light output to remain at around 30 klm (initial level). The commercial life of the second-generation HPI spanned 21 years, the longest of any generations in that family of discharge lamps. During this time the lamp design was revised three times, in 1972, 1979 and 1983.

The HPI/T 400W shown here is the last version of the platform, i.e., generation 2.4, introduced in 1983 and featuring an upgraded barrel burner made with a purer quartz material free of dissolved gases. The tubing drawn from this glass material thus lacks the airline bubbles that were present in high density in earlier HPI burners. This improvement not only reduced light scattering and increased the source brightness, it extended the service life of the lamp as well. A common end-of-life failure mechanism that impacted previous HPI generations was burner puncture at the point of highest temperature, which always occurred on the top side of the arctube, at the location of a airline bubble due to the smaller effective quartz thickness there. Removing the bubbles thus prevented this particular mode of failure, resulting in the service life becoming limited by other failure mechanisms such as iodine and hydrogen build up and moly seal rupture.

HPI lamps were initially manufactured by a team of skilled glassblowers at the company's Emmasingel complex in Eindhoven, the Netherlands. This resulted in a rather costly production process which was justified by limited sales volumes in the early days of the technology. As demand increased during the 1970s, development work on a fully mechanized production line began around the middle of the decade, with the intention of reassigning the skilled labor to the also growing production of 2 kW HPIs, especially that of the 220 V version, whose burner manufacturing relied heavily on manual labor up until the 2000s. The processing of the burners was also improved further with the introduction of a cassette oven process which no longer required the temporary vacuum sealing of the arctube extremities. Coincidentally, during that time a new mold-blown single-piece arc tube design was successfully developed and was released in a commercial lamp in 1979 (full production capacity reached in 1981), resulting in the 3rd-gen HPI. Interestingly, this development ran parallel to that of the mechanized production of the 2nd-gen HPI, which eventually resulted in two distinct 400 W HPI/T production lines running at the company's then new HID mother factory in Turnhout, Belgium. The production of gen-2.4 and -3.1 lamp ran simultaneously until late 1990 when the second-generation platform was phased out and its automated production line was upgraded to the gen-3.1 platform.

The introduction of the gen-3.1 400 W HPI in 1979 saw also an improvement in the overall lamp design, applied also to the then existing 2nd generation platform. One of the burner's mounting straps was removed, which reduced the flow of photoelectrons towards the burner, thereby improving its chemical stability (i.e., due to a reduced rate of sodium electrolytic loss), which had a beneficial impact on lamp life and color stability. The use of a purer quartz material for the burner, combined with the introduction of a new getter (two holding bands with four zirconium alloy pellets per band), further limited the rate of hydrogen buildup over time, which also had a positive impact on lamp life. All combined, those improvements pushed the mean service life of medium-wattage HPI lamps to 8 kh.

Typical of 400 W HPI lamps produced before 1999, the model shown here is designed to operate exclusively at 3.4 A on a 400 W mercury lamp ballast. Discharge ignition is ensured by a soft ignitor which relies on the ballast for the production of 750-1000 V pulses, combined with a burner characterized by a low starting voltage. The latter is achieved via the use of a neon-argon Penning mixture as the fill gas, which also makes the plasma ignition process less affected by hydrogen and free iodine buildup over time, thus guaranteeing a long lamp life. In order to maintain consistent light technical properties between the two 400 W HPI/T lamp generations than co-existed from 1979 to 1988, the burner in gen-2.4 lamps had its mercury dosage increased a bit in order to raise the arc voltage to 125 V, resulting in a dissipated power raised by 10 W to 385 W. The higher arc power load combined with a burner wall free of scattering defects (i.e., airline bubbles) results in a higher operating temperature, so the lamp's Na-Tl-In iodide mix had to be re-tuned in order to maintain the cool-neutral white light emission at 4500 K that is typical of standard HPI lamps.