From my experience for many decades at OETM/OERM/SCRM, I remember we were buying "rough service" 120 volt light bulbs for use in the street cars that used five-in-string wiring.

In regard to the 30 volt street railway bulbs in long strings that would short out when they fail, we were admonished that any bulbs that failed and shorted out should be replaced as soon as possible; other wise the voltage to the remaining bulbs in the string would increase and the failure rate of these would increase with every failure.

I did get some of those Edison style bulbs from Home Depot for our trolley, but the president removed them and put in street railway light bulbs which IMHO don't look as good in a 1890's car as The squirrel cage Edison style ones. Also wondering if anybody has any contact with the solar industry they have arc fault breaker that are designed for HVDC. The only question is would it work just downstream or would It trip anytime during the interruption from the trolley wire and harp connection.

Jim, Brian - I wasn’t aware of that possibility but makes sense to me. Thanks for the education. Ya learn something’ new every day.

Aamsco was selling that kind of bulb and claiming it was not gas filled but vacuum. Many kinds of gasses will help sustain a 600 Volt Direct Current arc that is not able to jump the same distance in a vacuum. The danger with an electrical arc is that it is very HOT!

I suspect the problem is the vaporized metal filament mixes with the gas and continues to provide a conductive path. Anyone who has seen the arc rise up out of the arc chute of an MS switch with a defective blowout coil will appreciate the length a 600VDC arc can attain.

After about 1912-1914, tungsten became widely used for electric lamp filaments because its lumens/watt ratio is higher than carbon filaments. Like the earlier carbon filament lamps, these were evacuated. However, the tungsten would evaporate rather rapidly and deposit on the inner surface of the glass envelope, making the lamp blackish in appearance and markedly reducing light output long before the filament reached the end of its expected service life. Introducing a non-reactive gas solved the filament evaporation/deposition problem. However, this introduced a new issue: when a filament burns out, the resulting arc ionizes the gas, and ionized gas is conductive. On AC, this isn't a problem because the current goes through the zero point twice in each cycle, and this extinguishes the arc. On DC, however, the current never drops to zero, so once started, an arc can become self-sustaining. Electric arcs have a negative resistance, in other words, the not only can an arc sustain itself, its current will keep increasing. Eventually, the failed lamp will explode or the arc can burn down the filament leads into the light socket, where it can start a larger fire.

Evacuated lamps used to be designated "type A", and gas filled lamps "type B", but these designations seem to have become obsolete. It's also important that street railway lamps use a C-9 filament.

I wonder if something like these would work: https://www.amazon.com.au/Battery-Operated-Non-Hardwired-Dimmable-Wireless/dp/B0BKVYP7BT

If I'm reading the description right, the edison screw is only for holding the puck, not for powering the LEDs. Power is from 3xAAA batteries and they can be remote controlled which might make switching off and on more convenient without needing a step-ladder.

I'd want something that used AA batteries for longer life. They're probably out there.

Those are pretty low in lumen output, akin to a night light. That’s about all you can get on battery power. Consider that a 60W incandescent bulb outputs about 800 to 850 lumens.

There are 600vdc to 115vac converters out there. I have no clue if they’re robust enough for a trolley environment. You’d still have to rewire the sockets.

https://absopulse.com/high-input-voltage-dc-ac-pure-sine-wave-inverters-600vdc-csh-400-f6/

S.E.P.T.A. (Philadelphia, Pennsylvania)'s rebuilt P.C.C.s must have similar 600 Volt Direct Current to 120 Volt Alternating Current devices, as they were using ordinary Compact Florescent Lights for interior fixtures.

Since regular fluorescent lights do not use 120 volts I suspect that they didn't buy a converter but just used a ballast that could take the 600 volts DC input. Also Aamsco is selling off all their incandescent bulbs at a discount and will no longer be selling them after the ban takes place.

A possible solution occurred to me. They do sell multi voltage 480vac compatible LED screw in lamps, the ones I have found so far are high lumen, but it's a jumping off point.

I would suspect that any multi-voltage 480vac lamp feeds a rectifier internally that changes 480vac into nearly 700vdc internally, and then into the power supply that supports the LED needs. This kind of lamp would inherently be compatible with AC or DC at many voltages, DC in either polarity, and should be safe when fed 600vdc. Trolleys would need one that will function as low (or ideally lower then) 120vdc, have an appropriate lumen output, and have appropriate vintage styling. I think the sacrifice would be that the power supply would be hard to tuck into the base/stem of the lamp, so you might have more going on in the stem of the lamp then old.

These would be a start: https://www.bulbs.com/200V+/LED_A--Style/Medium_(E26)-Base/results.aspx?Ns=P_Volts+Sort%7C0&Ntt=480+volt&Ntx=mode+matchany

They should have an internal DC power supply tolerant of at least 400vdc. Starting to get in the neighborhood....

CPLs are generally 120V.

You still have the series vs. parallel wiring problem.

A multi voltage bulb would work in series and also withstand the full voltage from a potential fault or differences in voltage drops. The issue is getting one that sufficiently wide voltage tolerance.

Usually on a 480/277Y system, the lighting is 277VAC. 480 is the phase-to-phase voltage.

Getting back to the original topic, surely there must be some manufacturers that will still make incandescent bulbs for specialty purposes?