Dummy Tube Fluorescent Ballast Tester.

The common form of fluorescent light used in Australia (240V supply) is a 4 ft tube and it is associated with a glow bottle or electronic starter and a simple iron ballast inductor that limits the current drawn by the tube. Tradesmen who change a lot of fluorescent tubes tell me that shorted ballasts are a common occurrence and when confronted by a non working fitting  they suspect a shorted ballast and first try an old tube. When a ballast is shorted there is nothing to limit the current and thus a new tube will be instantly destroyed together with  the starter. This destruction can be dramatic and may cause the circuit breaker to trip and smoke to be produced.

Described below is a tester that may enable shorted ballasts to be detected. The description is only of the CONCEPT. A WORKING prototype HAS NOT been made and  HAS NOT been used practically. The construction of mains equipment should only be undertaken by qualified persons.  

Circuit description.
The arrangement of 6 x 3.3K in series with the 12V 50ma globe produces a bright light at 240V and a much dimmer light at lower voltages.
The OKO6 electronic starter in common with similar starters provides a short circuit for the first 2 seconds after power is applied then a virtual open circuit  from then on until the power is removed. During this 2 second period a significant load is drawn through the arrangement of 6 x 330R  resistors the ballast and the external (dummy shorted ) starter. If the ballast is shorted then the full mains voltage will be available to light the globe. If the ballast is not shorted it will reduce the voltage to about 170V and thus the light will be dim.  The 6 x 330R resistors have to dissipate about 100W with a shorted ballast but the time is only about 2 seconds so the temperature rise is only about 3 degrees. The thermal fuse protects the device if there is some fault condition. The resistors in series with the light globe dissipate about 11W. It is thought that this would not lead to a excess temperature rise but thus has not been confirmed. The addition of a second thermal fuse positioned near these resistors is recommended.  The resistors have to dissipate high power for short periods and a wire wound type that is specified at "five times the power for 5 seconds" should be used. When the OKO6 turns off the external ballast being an inductor will attempt to keep the current flowing and thus produce a very high voltage spike. This spike is limited by the MOV.

Concept testing.
A circuit board (see pdf) was designed to hold the components and to fit inside a polycarbonate tube of 28.6mm diameter. The external diameter of such a tube was 31.8mm and it was hoped that this diameter would be small enough to allow the dummy tube to be inserted in common light fittings although this has not been confirmed.

Below is the prototype set up on a breadboard. The dummy starter (short circuit) has a tag on it so it so it will not be forgotten after a test. 

 

The resistors were mounted on both sides of the board as shown below before soldering.


Below shows the OKO6 starter circuit board mounting  and the MOV.


Below left shows the relative brightness of the light for 2 seconds on a good ballast. A shorted ballast shows bright only.

If some company could produce a properly engineered safely housed version of this circuit or something similar it would be useful to those involved in the bulk replacement of fluorescent tubes.