bigclive
patreon
Perplexing capacitors
🕑 Added 2025-03-01 05:28:10 +0000 UTCSince I recently found some old 8W fluorescent fittings I thought I'd compare the efficiency and circuitry to modern electronic ones. In the process I also discovered why the old ballasts get so hot. I've never actually calculated the effect of the winding resistance in a ballast before. The copper loss (resistance of the windings) is high enough to dissipate 5W resistively for an 8W tube.
There's also a slightly perplexing capacitor in the base circuit of one of the transistors to allow simple start-up of oscillation. But their choice of polarity is odd (but common to other designs).
Comments
Big Clive
The load will determine the output voltage. If left open circuit, the circuitry would probably just not start.
Big Clive
Fluorescent tubes are not just electrically quite noisy, but also have a high voltage gradient across them. I recall the design department of a lighting company wrapping a data cable around one to simulate a harsh electrical environment.
Mike Page
Thank-you Clive I enjoyed that little experiment. As I often do. 5W is a fair bit and that's just copper losses. My guess is there are some iron losses, too. To reduce this would need a larger and heavier choke, running cooler. After years of nagging my employers have finally fitted LEDs to replace the 6x 6 foot tubes which cranked out 400W of heat in Summer. Lovely, bright, everywhere. Funny story. I was mucking about with a photo-multiplier tube in the late 90s for a "secret project" and found fluorescent tubes have a quite high frequency bandwidth, unlike incandescent bulbs of the time which emit mostly DC with a bit of 100Hz ripple on top. Fluoros give you the full poached egg on toast effect. That reminds me of another thing. The lecture hall where I did my degree drove their flouros from DC to avoid camera flicker. I've only just made the connection.
Russell Levine
Did you mention what voltage the transistorized ballast puts out? Do the 400V components set the limit?
