Experience from experimenting with Philips S10E: The "rapid start like" behavior happened, when the lamp was not preheated enough (too low ballast current for that lamp,...) or was of higher arc voltage than the starter trigger voltage. So I would guess thhe newer starters were of the lower voltage "series" type. As different lamp wattages have different arc voltages, the starter could indeed behave differently.
What the starter actually does: When the voltage become above the trigger level, turn ON the thyristor with full gate current, so it conduct all the complete current wave till the real zero cross. After the "preheat" time (~2 seconds on the S10e), two things happen: - The thyristor is activated only when the voltage is above the threshold - After activating the thyristor the gate current reverse, so the thyristor holding current rise, so switch OFF at significant current (~100mA). As it interrupt the current behind the inductor, it cause an inductive kick to ignite the lamp. Normally, when the lamp ignite, it's voltage is supposed to stay below the trigger threshold, so the thyristor is not activated anymore. But when the voltage remain higher, another timer expire (it is in fact steered by further charging the timer capacitor; the time is about 0.2..0.5s), the starter stop activating the thyristor completely (as "failure to strike" fault mode).
Now if you use tube with higher arc voltage than the starter trigger threshold, the lamp start the discharge during the first ignition attempt, but next halfwave it is again shorted out by the thyristor, so the lamp appear to glow with only low brightness, but glow. As the second time section progresses, the switch-off current of the thyristor is gradually increased, so the lamp arc is fed by growing phase angle, so the brightness gradually increase. After the starter went into the "failure to strike" mode, the thyristor is not activated at all, giving the full wave for the lamp, so it turn to full brightness.
So it is not, what happen in the RS (there the cathodes gradually build up the emission, so the discharge current rises), but it create similar effect. Compare to real RS, the cathodes are already hot when the discharge first appear, so it cause no wear at all, only a form of "soft start"...
|