Transcript Document
PHILIPS TU/e Experiments on CFL Ignition Ignition Studies of Low-Pressure Discharge Lamps M. Gendre - M. Haverlag - H. van den Nieuwenhuizen - J. Gielen - G. Kroesen Friday, March 31st 2006 PHILIPS Outlines TU/e Goals of the study Set-up DC breakdown AC resonant ignition Summary 1/22 PHILIPS Goals of the Study TU/e Understanding Plasma Ignition Motivations Physics: better comprehension of dielectric-plasma phase transitions in general Technology: understand how compact fluorescent lamps ignite under various conditions 2/22 PHILIPS Goals of the Study TU/e Interest of a better understanding 1900s Courtesy of R. Richter, private communication 3/22 PHILIPS Goals of the Study TU/e Interest of a better understanding 2000s 4/22 PHILIPS Goals of the Study TU/e Understanding Lamp Ignition Q: How does low-pressure breakdown work ? Townsend model: electron avalanche between electrodes - homogeneous E field E - infinite electrode extension Neglected by Townsend - inhomogeneous field anode cathode ion atom electron - diffusion losses of charges toward the walls - wall surface charges 5/22 PHILIPS Goals of the Study TU/e Goals and Approach A: Thorough study of ignition in a ‘standard’ linear lamp studies: - different experiments on same lamp design - different lamp configurations (gas, pressures…) - control of experiments (repeatability, accuracy…) - cross-comparisons between results Global Overview of the Phenomenon 6/22 PHILIPS Outlines TU/e Goals of the study Set-up DC breakdown Back to AC resonant ignition Summary 7/22 PHILIPS TU/e Set-Up Global Circuitry power amplifier function generator HV probe RL HV probe 321electrode pulse Faraday current potential time optical generation reference recording imaging heating cage probing low-voltage current active same fast custom-built critical iCCD time electrode for measured camera the base waveform floating accuracy at forat for all ch digital 1 oscilloscope -200-1200V amplified three 1000K instruments 50-500ns capacitive of the critical experiments forto time eprobe points emission resolution IA 2 ch 4 3 double subelectrode one lamp full ms voltage potential pulse lamp and impedance mA scheme ignition taken mapping timeas for and constant the in time charge each main current and image over reference clean-up space resolutions time taken Faraday cage IF load current electrode simultaneous 20 7mm toresistor and 60waveforms images voltage 1ms for triggering space lamp are kept lamp current recorded at of added and constant scope time for regulation and by resolutions each value oscilloscope camera time step IG TTL trigger lens computer camera controller ICCD camera capacitive probe driver voltage regulator data no drift processed of electrode to give performances space-time diagrams during the experiments 8/22 PHILIPS Set-Up TU/e Experimental Frame of Reference electrostatic probe lamp ITO Faraday’s window rackcage probe - transparent - senses - stable housethe 145mm and electrostatic and long, lamp electrically locate 10mm surface the environment diameter conducting probe potential 9/22 PHILIPS Outlines TU/e Goals of the study Set-up DC Breakdown AC resonant ignition Summary 10/22 PHILIPS DC Breakdown TU/e time Ignition Mechanism Overview light K A K potential A Argon 3 torr global –600Vevolution identical in both cases Pre-breakdown wave: apparent lag of light - starts at the cathode emission (max 1ms) - propagates toward the anode smooth evolution of - speed and intensity decreases lamp potential Return strike: potential gradient in the ofwave first wave - starts at endwake of first propagation - propagates toward the cathode x (cm) 14 -500 - speed and intensity decreases 0 +5000 -400 -300 -200 -100 0V 11/22 PHILIPS TU/e DC Breakdown Cathode-Initiated Breakdown 3torr Ar-Hg lamp : -500V dt=100ns K A (0) 12/22 PHILIPS DC Breakdown TU/e Argon Argon 3torr 3torr –400V –200V –300V : Successful : Failed Ignition Ignition resolution potential: 10ns optical: 500ns 100ns 13/22 PHILIPS DC Breakdown TU/e Pre-Breakdown Wave Position vs. Voltage 51- 44 km/s 10 - 3 km/s - wave speed directly proportional to voltage - ignition condition: first wave has to reach the anode 14/22 PHILIPS DC Breakdown TU/e Lamp Net Charges vs. Voltage - first wave charging effect increases with voltage - decrease of net charge only for successful breakdown - Ignition condition: charging threshold to be reached 15/22 PHILIPS TU/e DC Breakdown Global Overview of the Phenomenon _ Qualitative model e + Wall and field displacement High charging Local radial/axial Townsend-like E-field breakdown component Further Current ionization stabilized in by front ballast of cathode Wave Field rotation propagation and toward anode Electrodes Exponential current circuit increase closed - ionization wave drivenbridged, by front enhancement field, rate ofthe wallcathode charge - wave speed dependent on E/p value Global lampelectron charge decrease Decreasing current flux Steep current increase - gradual decrease of field and wave speed during propagation - ignition condition : E/p high enough for 1st wave to reach anode 16/22 PHILIPS Outlines TU/e Goals of the study Set-up DC breakdown AC resonant ignition Summary 17/22 PHILIPS AC Resonant Ignition TU/e Optical Recording 18/22 PHILIPS AC Resonant Ignition TU/e Electrostatic Recordings 19/22 PHILIPS AC Resonant Ignition TU/e Correlation with DC Breakdown - synchronous propagation of K and A waves - importance of surface charge memory effect - easier ignition in alternating potentials as a result 20/22 PHILIPS Outlines TU/e Goals of the study Set-up DC breakdown AC resonant ignition Summary 21/22 PHILIPS Summary TU/e Global Overview - multiple diagnostic tools running simultaneously - cross comparisons between optical/electrical data - various experimental conditions investigated - correlation between wave propagation and lamp charging - minimum lamp charging required for successful ignition - new information inferred from data analysis 22/22 PHILIPS ANY QUESTIONS? TU/e PHILIPS TU/e Set-Up Global RC-Probe Circuit Z - Faraday’s cage + R r x Provides lamp surface potential vs. time/space - limited field disturbance around the lamp - Z chosen so total system transfer function = pure real - little need for post-experiment data treatment U PHILIPS TU/e Set-Up Calculated Data from RC-Probe Output Measured: F=f(t,x) F Calculated: axial E field: EX (t , x) x F radial E field: ER (t , x) a ln(b / a) ( g / a ) ln(c / b) |E| field: E (t , x) E X2 ER2 1 EX <E field: E tan ER disp. current: I D CL F total disp. current t linear charge: QL F CL total charge