Hello everyone,

I've been wondering this for ages now:  How on earth do retrofit HPS lamps work?  I know some have that coiled wire that wraps around the arc tube, but even in the absence of high voltage pulses, how does that function to actually ignite the gas?  Does it have to do with somehow passing electrons through the gas, or is there more to it than that?

Likewise, I've also seen ones that use the neon glow starters, similar to older preheat fluorescents.  What does that glow starter do in the case of the models that rely on these?  Beyond this, some others, like ones made by EYE lighting, actually have a small halogen capsule inside - which makes NO sense to me!

Finally, I've noticed that many retrofit HPS bulbs start with a red glow, indicating a high presence of neon gas.  I imagine the neon has something to do with reduced current in ignition as well?

Any insight would be greatly appreciated!

The standard high voltage start HPS lamps are optimized to efficacy, and this includes the gas choice (Xenon) and its pressure. Those lamps require a pulse of about 3.5...5 kV to start. (some low power lamps may suffice with ~2kV)

In the lamp designs you mentioned, some trade off had been done to lower the required starting voltage :

The 70W SON I lamp (with a starter) is a common lamp in 230V countries. It is not a retrofit lamp for Mercury, but "native" HPS lamp in its own right. (it runs on the same ballast as the pulse start HPS, but without ignitor). This lamp is made to bring down luminaire costs (by eliminating the ignitor), and to allow the manufacturing of HPS versions of former Mercury lanterns, in which there is no place for an ignitor. The ignition voltage of this lamp is brought down to the 1kV ballpark, so it suffice with just the 1kV or so pulse made by switching across the ballast output

The Mercury retrofits (with no moving parts) ignition voltage is brought down to <230V, so they start just from line voltage (at the expense of even more reduction in efficacy, by changing the starting gas composition)

Some oddball lamps that had been made include the thermal switch lamps (some of the GEC Solarcolour lamps) that are probably similar in their starting requirements to the 70W I lamps, and Iwasaki lamp with nonlinear capacitor, which capacitance collapse at a specific voltage, so this lamp gets its starting voltage essentially from switching across the ballast output as well. (Older Iwasaki lamps used the starter)

Am I right in thinking that some lamps such as those with glowbottle starters, and possibly the Solarcolour lamps, used these to induce an inductive 'kick' from the ballast..?

Thats right. The switching across the ballast output (specifically, shorting it out to create a magnetic field, and then releasing the short, hopefully during the peak current, to collapse the field generates this kick. The voltage can really vary, but is generally assumed as 1kV

High voltage ignitors work by charging a capacitor, and then dumping its charge across the primary of a pulse transformer. The transformer then steps up this pulse to the kV range according to the turns ratio like any transformer. In superimposed ignitors it's a ferrite transformer included inside. In ballast dependant ignitors (as their name suggests), the ballast itself acts as the transformer - where the short winding part between the tap and nearest end is the primary, and the entire winding length is the secondary

Am I right in thinking that some lamps such as those with glowbottle starters, and possibly the Solarcolour lamps, used these to induce an inductive 'kick' from the ballast..?

This group also includes those Iwasakis with FEC ignition device.
And instead if the glowbottle, some designs used thermal contact starter.