I think that the LED lighting already caused the population vision to degenerate, because it causing people to prefer brighter lighting than usual and to not seeing good with lighting that was once enough.
In the past, street lighting with <125W MV lamps was enough for safety and now it is too weak.
I can see this phenomenon at the entrance of Carmel hospital, where when I turning off part of the LED panels, most people says that the lighting is insufficient and turning them back on.
Also: Most LED streetlights at Kiryat Ata, are brighter than the former HPS streetlights.
Your opinion?

For something that appeared in seconds but then does not continue growing, i would guess some foreign substance on the electrode, which then either is flash-evaporated, or (if it is a dust of some form) deposited from it electrostatically when power is applied. I have no explanation what it might be and from where it got there

What i do wonder is - In FL tube manufacturing, the cathodes are heated quite brightly for emitter activation, and tubes are tested for having vacuum by lighting up for a second - and i guess there is nothing more than applying high voltage in instant-start way. Whatever the blackening is, why it does not appear already there and then ?

It looks like heat marks, more than wear marks. Maybe that ballast you have overdrives the tubes severely? Or the cathode heating is too severely intense, causing severe heat in the cathodes, which can cause blackening on the phosphor itself within minutes.

I've done experiments with severe overdrive on larger T8 tubes and severe end blackening can happen within 10 minutes at twice the rated power. It looks like the tubes I've tested with have been running for years and years even though they haven't!

@RRK
Yes, they really didn't catch on here, very unfortunate.

Note that general LPS lighting was absolutely absent in USSR too, just except some experimental lamps.
The reasons probably were the same as in USA, cheap electricity, quirky lamps hard to make, quirky use - position dependent and high ignition voltage, zero CRI.

Not all the things that happen in the lighting industry are completely logical


@Ash - forcing a leak transformer to have a higher OCV at the same current increases losses / weight / costs proportionally due do the necessity of having proportionally higher reactive power, so more turns at secondary, heavier core and thicker winding at primary.

There are saturation tricks in leak transformers allowing to get peaky output voltage, but these are likely taxed with higher losses too. 

How does this work?

More so, some lamps can be ignited by vigorously wiping the outer bulb with a synthetic cloth! If you do this in the dark, you'll see buffer gas in the burner flashing due to the strong electric field induced, and even small blue corona at metal parts in the outer bulb (nitrogen filled, I assume).

It might have made sense for fields that didn't receive much use at night or for residential tennis courts where the payback for MH would be much longer.

Overdriving 1000w MV lamps at 1500w was a practice mentioned in the Sylvania Engineering Bulletin for mercury lamps. Lamp life would be reduced to 2000 hours from 24000 hours. The ballasting details was unclear if dedicated 1500w ballasts were available or if two 700w ballasts were connected in parallel. 

I have recently installed a staircase lighting timer (actually not in a staircase, but still a lighting setup, something for my own use). While discussing this setup with a person and later the timers themselves, i went online to show some of them, such as the Theben Elpa 8 (my favorite, although the one i installed isn't it) :

https://www.theben.de/de/elpa-8-0080002

While looking through the descriptions i found out that the 3 wire circuit is not permitted in new installations in the EU. (No such limits in Israel, here both 3 and 4 wire circuits are widely used)

Any ideas what is the reasoning ?

The 3 wire circuit switches the Neutral not to the luminaires, but to the timer module somewhere in a DB, which is permanently live anyway. So the normal "circuit is open but still live" considerations dont really apply in this case

Our electrical code does require switching of L (or switching of N only together with L), but specialised control circuits are generally viewed as exempt from this requirement. In this case, the button circuit is indeed a control circuit, while the lamp circuit is not



To anyone not familiar with the staircase timers :

Those are commonly used in apartment blocks with stairs. There is a push button (momentary on) at each floor. Pushing it switches on the lighting, after a preset time (few minutes) it goes off

The function is implemented by a timing module installed in a DB somewhere in the building

Back in the day those timers were based on a slow-cooling thermal relay (Israeli "Thermion") or on hydraulic mechanisms (something that was described here by Medved in past discussions). Nowadays most are electronic, with few exceptions like the Elpa 8 which is electromechanical

The timer wiring can be one of two circuits (see attachments)

The 3 wire circuit had been used exclusively in the past with Thermions because they can only be wired this way. In the 90s the first electronic timers appeared, some of which required the new 4 wire circuit. Nowadays virtually all timers support both, sometimes requiring setting a DIP switch or similar during installation to choose the correct circuit type

The timer i used (something almost 20 years old, saved during renovations in an air raid shelter 2 years ago) is an electronic unit requiring the 3 wire circuit, and was wired accordingly

Here overcompensated lighting circuits were quite common. Since our street lights were equipped with dual wattage ballasts most of them were running on the lower wattage but the built in capacitor had the value for the higher lamp wattage. Same goes with the dimming ballasts where the cap had its right value for the full wattage and not for the reduced one.

I saw that video, very very cool. That MH lamp is almost certainly on a ballast though. I think it would very rapidly self-destruct if that wasn't the case.