Photo Gallery

Running a rare 15 W metal halide lamp

In 2011 Osram released the smallest commercial ceramic metal halide lamp on the market, the HCI Powerball 15W, which was also distributed in North America by OSI-Sylvania under the Metalarc Powerball brand. The intention was to provide an energy-efficient and long-life alternative to 50 W tungsten halogen lamps in the retail sector, but LEDs were beginning to make inroads into that market already. Because of the latter's success, Osram/OSI abandoned its 15 W metal halide system the same year of its introduction, thus making it an exceedingly rare lamp. Since its electronic ballast is practically impossible to find nowadays, the problem of properly operating the lamp is a rather difficult one.

I solved this challenge by modifying a Philips PrimaVision ballast intended for 35W CDM lamps, which actually runs at 39 W.  Changing the system's current-sensing resistor from 1.5 to 4.4 Ohm lowers the regulated power to 15-16 W only. The picture above shows the end result with a Sylvania MC15TF running optimally. I came to that solution when I sought to modify this type of electronic ballast for running a 35 W SDW white sodium lamp properly, and those normally work at 33 W, so I had to find a way to lower the ballast output by 6 W, which is achieved by increasing that resistance value to 1.8 Ohm. I ended up testing various resistances, which led me to such low power level compatible with the rare 15 W MH lamp shown here (still amazed at the circuit's flexibility!). Additionally, varying that resistor value effectively enables dimming from an electronic ballast originally designed for a fixed power output, which is very convenient since to the best of my knowledge dimmable 35 W HID ballasts do not exist (and I needed one :). Although I had the MC15TF shown here for a couple years, I never ran it for fear of damaging it. So it was quite a treat seeing such unique and rare discharge lamp finally come to life.


Keywords: Miscellaneous

Running a rare 15 W metal halide lamp


In 2011 Osram released the smallest commercial ceramic metal halide lamp on the market, the HCI Powerball 15W, which was also distributed in North America by OSI-Sylvania under the Metalarc Powerball brand. The intention was to provide an energy-efficient and long-life alternative to 50 W tungsten halogen lamps in the retail sector, but LEDs were beginning to make inroads into that market already. Because of the latter's success, Osram/OSI abandoned its 15 W metal halide system the same year of its introduction, thus making it an exceedingly rare lamp. Since its electronic ballast is practically impossible to find nowadays, the problem of properly operating the lamp is a rather difficult one.

I solved this challenge by modifying a Philips PrimaVision ballast intended for 35W CDM lamps, which actually runs at 39 W. Changing the system's current-sensing resistor from 1.5 to 4.4 Ohm lowers the regulated power to 15-16 W only. The picture above shows the end result with a Sylvania MC15TF running optimally. I came to that solution when I sought to modify this type of electronic ballast for running a 35 W SDW white sodium lamp properly, and those normally work at 33 W, so I had to find a way to lower the ballast output by 6 W, which is achieved by increasing that resistance value to 1.8 Ohm. I ended up testing various resistances, which led me to such low power level compatible with the rare 15 W MH lamp shown here (still amazed at the circuit's flexibility!). Additionally, varying that resistor value effectively enables dimming from an electronic ballast originally designed for a fixed power output, which is very convenient since to the best of my knowledge dimmable 35 W HID ballasts do not exist (and I needed one :). Although I had the MC15TF shown here for a couple years, I never ran it for fear of damaging it. So it was quite a treat seeing such unique and rare discharge lamp finally come to life.

Philips_PLE23-CPro.jpg _IMG0221.jpg MC15TF_run.jpg SOX_jacket_m.jpg DSCF0621.JPG
Lamp/Fixture Information
Manufacturer:Sylvania (Osram)
Model Reference:MC15TF/U/GU6.5/830
Lamp
Lamp Type:Ceramic metal halide
Filament/Radiator Type:Thermal discharge in argon, mercury and metal iodide vapor (Na, Tl, Dy, Ho, Tm)
Base:GU5.6
Shape/Finish:Tubular clear
Service Life:15 kh
Burning Position:Universal
Fixture
Ballast Type:Modified Philips PrimaVision 35W CDM
Electrical
Wattage:15 W
Optical
Lumen Output:1200 lm
Lumen Efficacy:80 lm/W
Colour Temperature:3000 K
Colour Rendering Index:82 Ra8
Physical/Production
Factory Location:Berling Spandau, Germany
Fabrication Date:December 2011
Application/Use:Shop and retail accent lighting
File information
Filename:MC15TF_run.jpg
Album name:Max / Misc lamps and lighting
Keywords:Miscellaneous
Filesize:791 KiB
Date added:Jan 23, 2025
Dimensions:1200 x 1069 pixels
Displayed:16 times
Exposure Time:1/550 sec
FNumber:f/11
File Source:Digital Still Camera
Focal length:40.7 mm
Software:Digital Camera X-E4 Ver1.12
URL:https://trad-lighting.net/gallery/displayimage.php?pid=638
Favorites:Add to Favorites

Comment 1 to 5 of 5
Page: 1

Tuopeek   [Jan 24, 2025 at 10:03 AM]
That's an impressive, and good to know, modification to the ballast Max. I guess you had the circuit diagram to help, or that's even more impressive. I used a switchable ballast to drive a 250W SON, which was the lowest rating, and destroyed the lamp on warm up. I suspected the run-up characteristics didn't change for any of the 250w to 600w options, so initially a bit hard on the lamp. The lamp was close to end of life so may just have been unlucky. The lamp would spend a few second of start up rectifying on 50Hz, so I though I would see if the HF ballast would be more stable, It didn't noticeable flicker but broke the arc tube in two during the run-up.

Lovely rare lamp. I liked the idea of small halides replacing tungsten halogens. I used a 70w as an uplighter for a bit, but run-up time can be an issue in a domestic situation.
Max   [Jan 24, 2025 at 10:56 AM]
Thanks Mark. No, I didn't have the circuit diagram at hand, but I know the general design principles of such ballasts and I just looked for the current-sensing resistor. I found a good candidate and I took a gamble, which paid off in a big way! I'm now testing a fully dimmable 35 W SDW lamp system that I will probably integrate in my living-room HID uplighter (changing the lamp current during operation works as well, even for MH lamps!!!). I was not expecting the solution of changing that resistance value to work so well, especially given that all aspects of the driver's operation are managed by a micro-controller, which for me is a black box given my limited knowledge about this specific ballast. It turns out that current adjustment is controlled just by comparing a voltage drop across a resistor against a set value. Change the relation between actual current and sensed voltage and you can vary the output power to any desired level within a certain range, which is luckily quite broad here.

What you describe is a cautionary tale about certain deleterious lamp-ballast interactions that can happen with certain electronic drivers. In your case, there could be several causes to your problem: either the current fed to the lamp was just too high, causing a too quick temperature rise, or there was an onset of acoustic resonance at one point during run up, causing pressure modulation. Both issues can result in the rupture of the arctube.

As for HID lamps used in domestic applications, the run up and hot restart delays are certainly major drawbacks from a consumer's perspective. I use many homemade HID lamp systems at home and personally I don't mind the hot restart delay, and I quite like the colorful show during run up, but that's just me Smile
Sammi   [Jan 24, 2025 at 10:01 PM]
You sir, have done it again.! Love Mind Blown
Ria   [Jan 25, 2025 at 01:07 AM]
He has indeed, another wonderful addition to the site Bulb Man Love
Max   [Jan 25, 2025 at 10:38 AM]
Thanks ladies! I was so pleased to have finally managed to run that 15 W CMH lamp that I had to share it here Smile

Comment 1 to 5 of 5
Page: 1