I FINALLY got my Dura FS-2NA starters. These are automatic reset. I have yet to see any FS-4NA starters.  They start the lamps  F A S T ! ! !

I got a some Hubbell manual reset starters. In the box of  5 ONE worked. The others didn't do anything. What was wrong with the one I opened the very thin wire that heats and kicks the breaker was broken. These were made in Costa Rica and not made very well. If I was a bit more mechanically inclined I could fix them.

I just did a buy it now on a box of 40 watt GE COP starters.  I meant to bid earlier, but it expired and he relisted them. So then, when I saw it.... BOOOM.... I did a buy it now.


Powell

Usually electronic starters have automatic reset after lamp change.
Glow starters with a thermal cutoff, have a button that released, when the starter sense a rectification or fails to start the lamp, and you need to manually press this button after lamp change to reset the starter.
I don't know of mechanisms of automatic reseting for glow starters.
Is your Dura FS-2NA starters are electronic? Or glow bottle?

These automatic reset starters are glow bottle. There may be rudimentary electronics, but these date to the 1950s to the 1960s design. 

There is no ways to create a glow starter with an electronic cutoff.
If the cutoff is electronic, why that the starter itself wouldn't be as well. This is much cheaper to do.
I think that this is the same mechanism that disconnects magnetic ballasts, when they are overheated, and it don't require a reset at all.

I think it is a form of thermal circuit breaker in series with the glow bottle. The preheat current is sufficient to trip it, but te short preheat time in a good lamp is not. The restarting of a rectifying EOL lamp gives the breaker long time to heat up and trip

The starter cut out of manual reset starters is usually simple latched-close contact with bi-metal release, the bi-metal being heated by the glowbottle itself. When the bi-metal heat up, it release the latch, so keep it disconnected till someone arm it back.
The requirement for the latched shut-down is there on purpose, as it is expected the tube is already faulty, so need replacement - so someone have to climb to the fixture anyway, so he could easily arm (reset) the starter back.

When an "automatic" reset is required, the thermal cut-out could be then simple bi-metal spdt contact. When heated by the glowbottle, it disconnect the glowbottle and connect a resistor heater, what keep the bi-metal hot till you break the power.

In the simpler version a PTC is connected in series with the glowbottle. After it heat up, the PTC increase it's resistance, so reduce the current through the starter to the level the starter could withstand for long time of activity (~10..20mA).

I've never seen a glow starter with any sort of cutout, but what an excellent idea.

I use electronic starters but even these aren't 100% reliable as they occasionally fail to start or restrike a lamp in my twin 4ft fittings. This was the case at least with ageing GE F40/33 (T12 Cool White 4000°K) tubes (which are now in my collection for possible re-use on  my vintage TranStar twin 40W SRS ballasts in the workshop), it seems to start my German made Osram L36W/865 Lumilux Daylight (T8) tubes with no problem at roughly the same ambient temperatures, although these take a good few minutes to warm up when it's cold.


BG

I got the unopened box, and some of them don't work. I wiggled the reset button on one and it worked. I guess there is oxidation somewhere...

Powell

In my experience electronic are quite sensitive for any capacitance - if there is a RF suppression cap in the fixture (some old ones had that), it will kill it after few starts. Same when there is longer wiring between the ballast and lamp or to the starter, like remote aquariom ballasts. Again "excellent" electronic starter killers.

The "Lamp strike fault timeout" (the time, how long the starter generate starting pulses before it shut down) is indeed usually set too short. In the basic design the timeout is linked to the preheat time, so for 2 second preheat the ignition attempt is limited to 0.2 seconds, what is sometimes not sufficient.
The problem may get worse, when the starters have many hours on high temperature environment - it's electrolytic capacitors (used for the timing) dry out, loose it's capacitance, so the both the preheat, as well as failure detection time shorten.

The glow starters with thermal cut-out do not have this problem, as the cut-out need few minutes of the starter attempting to strike the lamp before it heat up to the tripping point of the cutout device. So the cut-out activate only when the lamp is really dead. But they contain mechanical contact(s) with their problems (oxidize, weld,...)

Electronic starters tend to behave differently depending on which tube they are used with. In my 4ft fittings they behave like rapid start, in the MLP13s and the 8-footer on the top floor they behave like normal preheat albeit with approx 1 second electrode heating and a single strike on the 13W fittings and approx 2 seconds on the 8ft 100W with maybe a couple of pulses to attempt to strike. Unfortunately in the latter case the starter sometimes gives up before the tube strikes (especially if the room hasn't been heated) so I may change it back to a glowbottle.


BG

BG101: A behavior of the lamp like a rapid start, with a glow/preheating electronic starter in a preheat/switchstart magnetic ballast (Except if the electronic starter, designed to intentionally rapid start the lamp with a HV), usually occurs if the lamp (T12 usually) have sufficient low starting voltage and it than cold starts directly from the mains with cold cathodes (Tungsram). The lamp will than initially glow dim and than suddenly brightens to full brightness without preheating, similar to rapid start.
Also, remember that like glow starters, the electronic starters can be 4-20W, 4-65W and 4-80W and probably 4-125W.

My experience with S10E (Philips) is, then they stop any cold start "attempt" of the lamp (even of the F4T5), then preheat for ~2 seconds and only then ignite the lamp. But the cold starting may appear, when the starter's timing capacitors are weak (based on how the circuit work).

Dor, I'm using 4-125W starters. The electrodes of the tubes in the 4ft fittings do glow as the tubes strike accompanied with a loud buzz. This happens with T12 40W tubes and is the same with 36W T8s.

I've never known T8s to cold-start although it's common with T12s, my SRS 5ft sometimes starts instantly.


BG

Cold starting also depend on whether the voltage across the lamp is close to 0 (as the starter is shorting it - programmed start starters in preheating mode), moderate (if the starter is timing as to avoid strong inductive kicks - "rapid start" starters), or 700V+ (full power indutive kicks from the ballast). This behavior depends on the starter design and the stage of starting

Loud buzz will happen is the starter is causing weird waveforms in the ballas (by switching the current at specific timings through the AC cycle), or if its rectifying so overcurrent in the ballast

I have seen T5 8W to instant start with the starter removed, and a batch of Tungsram T12 40W manufactured around 2000-2002 that would instant start (and get to fll brightness in seconds) on a normal switch start circuit with the starter removed, or sometimes even with the starter in there (lamp instant starts without the starter even reacting)