In one 347V light fixture I found two different bulbs, one 100W MH and another small halogen with DC/BA15d base. The burned ballast was connected with relay inside the fixture. Does someone know how it is connected(schema) and what is the purpose of the small halogen bulb?
The small halogen lamp is a meant to operate for the period while the Metal Halide lamp is warming up. That way you will instantly have some light when the fixture is switched on. It is called a "restrike" bulb. The timer relay which operates that bulb should drop out when the main (MH)lamp reaches full brilliance.
Generally the relay is tapped off of the 120V lead and the common of a multitap ballast.
The coil of the relay will connect to the 120V connection on the ballast, and the common (neutral) connection. The quartz lamp will connect to the switched leg of the relay, and the neutral connection. The quartz lamp is 120V,not 347V. The 120V tap on the ballast will provide the 120 that the lamp requires.
**Keep in mind that some of the newer restrike methods are current sensitive, and will have different connections. That's why I said generally in my post above**
To my mind the better term is CWA -- constant wattage autoformers.
Ballast -- as a term -- reminds me of old, old classic fluorescent fixtures in which the ballast performed more like a 'choke.'
BTW, these lamps need healthy field wiring, too.
While the damage certainly shows up in the fixture -- it's often the result of failing connections elsewhere up the line.
Rather like an induction motor -- CWA's draw heavy until the lamp fires off. If there is too much impedance ( particularly on the neutral ) one or more lamps won't 'strike.' Then, after powering down, a re-strike attempt will cause the pattern to entirely reverse: the dead lamps fire up -- and the previously lit fixtures won't re-strike!
This situation is not that rare -- because by the nature of the devices they are used across long distances from the panel. Even bumping up to #10 is not enough, sometimes, to get these babies to fire.
Sometimes you'll have to split the run in half ( double the #10's ) joining them together in a gutter near the panel. This up-sizing is purely for impedance reduction.
So your job is not done until you've inspected the condition of the field wiring.
Actually, the CWA version draws lower Line Amperes at Startup, then tapers off at a steady operating level - provided the Voltage at the Reactor Input stays within +/- 10%. Constant Wattage Autotransformer setups typically draw apx. 80% of the stable operating Line Amperes at Startup.
There are Two other setups, which draw lower Starting Line Amperes -vs- Operating Line Amperes. These are:
Constant Wattage Isolated Transformer setup (AKA: "CW" and "CWI")
Regulated Lag Isolated Transformer setup (AKA: "REG-LAG" and "MAG-REG")
Both of these designs use a Tapped Primary Winding for multiple Input Voltage choices.
The CWI incorporates a single isolated secondary winding for Lamp operation. Lamp regulation is performed via a Capacitor in series with the Lamp.
The REG-LAG setup has (2) Isolated Secondary Windings - One is connected to the Lamp, the other is a Tertiary Winding, connected to a Capacitor. This is used to regulate the Lamp Performance by regulating the Lamp's Current.
The Current Regulating device which draws higher Line Current at Startup, is the "High Reactance Autotransformer" (known as "HX-NPF" and "HX-HPF"). This setup is a CWA without a Series Capacitor connecting the Autotransformer (input) side, to the Reactor Coil (Choke). As with the CWA, the Lamp is connected in series with the Reactor section, and the Input side's "COMMON".
Standard P.F. is nominal (50%). Improved P.F. is achieved by connecting a Capacitor across the Line Input at the Autotransformer section.
The standard basic Reactor version also draws higher Starting Current than the Operating Current level. This type of Current Limiting "Auto-Regulator" device is simply a Reactor Coil in Series with the Lamp (no input Autotransformer section). Without a Shunt Capacitor across the 2 Wire Line input, the Power Factor is 50% (200 VA drawn for a 100 Watt Lamp - 173 VARs bounced between the Reactor and the System Transformer) Improved P.F. via Shunt Capacitor brings the P.F. to "High" (80%-90%).
It appears that the reason for the higher Starting Current, which tapers down to a lower Operating Current, would be the increasing Capacitive Reactance of the Operating Lamp / Arctube.
Looking at the complete Circuit, when the Assembly is first connected to an AC Source, the Generating Source sees the Reactor Coil only - as the Lamp is a Negative Resistance device. The Reactor "uses" very little True Power (Wattage), and requires Magnetizing / Charging Power (Reactive Power), so the Reactor Coil draws a large Apparent Power (Voltamps), with an extremely low Power Factor.
As the Lamp's Arctube becomes more active - and the Lamp progresses to Operating level, the Lamp takes on more and more Capacitive Reactance - thereby storing more and more Reactive Power. As the XC increases, it "counteracts" against the XL, and in doing so the Reactive Power stablizes. At stable Operating level, the connected Impedance (Z) reflected back to the Generating Device, appears to be comprised of nearly equal levels of Resistance and Reactance, and the VA Package impressed to the Load carries mostly True Power.
Scott " 35 " Thompson Just Say NO To Green Eggs And Ham!