Forgot to mention an interesting development. A company is selling an item called "GPS Lightlock". It is an electronic astronomic time switch mounted within a twist lock photoelectric control  enclosure. It has a gps receiver that allows it to determine the time and location it is at. I purchased one and it seems to work very well. It turns on and off about 11 minutes before sunrise and after sunset. I think it would be very useful on lights where the operation of a photoelectric control is affected by ambient lighting.

That GPS LightLock is a brilliant idea, but I wonder if that 11 minutes after sunset & before sunrise can be adjusted closer to say 30 mins.

Thanks Ugly1 for the details in N.Y.'s transition! That civil twilight standard seems to indicate an average of just over 11 hours of use per night. I had calculated 11.5 hours somehow.

@Ugly1: With 200,000 street lights in the late fifties I would imagine there was a period of a few years when N.Y. converted all the incandescent lamps to MV. I've read somewhere that the conversion period started in 1961 (I think you mentioned it HERE). But that wouldn't include new streets or highways and streets that never had lights which accepted MV posts+lamps I'm sure well before 1961.  If this is correct, how long did it take to switch all 200,000 incandescent lights?

 Concerning the Lightlock control,I agree that 11 minutes is too early.
 The first large scale use of mercury lighting in NYC was in 1955. A private business group,The Sixth Avenue Association,wanted to improve the lighting on that avenue. The plan was to replace the existing cast iron poles at each intersection(that held 10,000 lumen incandescent bulbs) with new steel poles using either a 15,000 lumen incandescent or 20,000 lumen mercury bulb. An additional two poles were to be added between the intersections. They chose mercury. The Association paid for the material and the City paid the installation. OV-20 type fixtures were used with pole base ballasts. It was coincidental that the rise in crime in the late 50's came at the same time as the introduction of less expensive mercury fixtures with internal ballasts and longer lived bonus line mercury lamps. This made the installation of mercury lighting more cost effective. Starting in 1959, the City placed mercury lighting in high crime areas. The reduction in crime in the newly lighted areas was so pronounced that the City decided on a 64 million dollar program to upgrade all streetlights to mercury vapor. The program was basically finished by 1968.
 By the way, in 1966,the Sixth Avenue Association  raised  money to replace the 400 watt mercury fixtures they had installed with 400 watt  lucalox hps lamps,one of the largest installation of hps at the time.

Fascinating detail Ugly1! What was (or still is) your affiliation or association in the N.Y.C. street lighting area?

That 1966 lucalox installation is probably the earliest I'm aware. I believe niobium seals hadn't been invented yet (around 1970) so they must have beeen high maintenance 400w hps lamps.

It took 7 years to complete the MV transition from 1961-1968, but I'm sure some incandescent lighting was ignored..Are there ANY incandescent-based street lights anywhere to be found in N.Y.C.?

 I had no connections concerning streetlighting in NYC, just another enthusiast. I was always interested in electricity from the time I was a tot, and  wound up working as an electrician for the NYC Human Resources Administration( a fancy name for the Welfare Department) for 34 years.
 An even earlier lucalox installation was made in  Manhattan during 1965 on a street adjacent to GE headquarters. The lamp listings at the time gave the two available lucalox wattages(275 and 400 watt) an average life of 6000 hours. It was this test installation that led to the hps relighting of 6th avenue that I mentioned.
  There were quite a few cast iron poles with incandescent luminares that were overlooked during the mercury conversion. In 1973 most these poles were landmarked to be preserved. In 1984, almost all the incandescent luminaires were replaced by modern hps fixtures, which raised quite a howl among preservationists. In the 90's reproductions of the original luminaires ( made by Sentry Electric) were installed. Incandescent lamps remained on railroad and highway underpasses until the early 80's when most were replaced by 90 and 135 wat low pressure sodium.

 E-bay has listings for two of the astronomic time switches used by New York City. Kind of high priced. Item # 1119 9883 9604 and # 2319 4934 5566.

Super info Ugly1! The first HPS installation I ever saw was in 1978. I was really not aware they were testing lucalox lamps as early as 1965!

As Ugly1 uses it in his replies, the correct term then is "Astronomic Time Switch". I used the exact wording on ebay and came up with 14 listings, from modern electronic to vintage. Thanks!

I have read somewhere that the earliest automation of street lighting had mechanical spring wound time clocks.  Since these required weekly winding, they were adjusted at the same time as well so there was no immediate need for astronomic functions. This system most likely preceded the photocells and were probably mostly extinct by the time mercury vapour installation started.

Well, mercury vapor installations made the necessary service visits about at least 2years apart, what became way too long for the clock to remain in sync.
The incandescents require visits every 2 months (or more frequent), what means just sufficient for the mechanical clocks to hold the time reasonably well (assume the spring winding was made electrical).
Plus after the WW2, when a lot of electrical and electronic factories originally set up to make the war equipment became abundant, the CdS cost had dropped a lot, so the photocells became way cheaper to buy alternative. It then needed the usual 10..20 years till that new technology became adopted into wide use.

Just need to put this here since it's very ontopic..

There's a restriction of the use of photocells in the EU for over a decade because of a little known law. The RoHS restricts the use of cadmium among other toxic metals (like Pb & Hg). So CdS based PCs are not that common, plus of course the history of street-lighting in Europe maintaining separate 230-240 volt power distribution grids since the arc-lamp era and controlling the ligths via central control.

I'm surprised to read this I think every streetlite in our town has a photocell maybe we've managed to avoid this particular stupid regulation. Street lamps here are run from normal mains just tapped off at convenient joints so that if one phase goes off not all street lights are affected

Just need to put this here since it's very ontopic..

There's a restriction of the use of photocells in the EU for over a decade because of a little known law. The RoHS restricts the use of cadmium among other toxic metals (like Pb & Hg). So CdS based PCs are not that common, plus of course the history of street-lighting in Europe maintaining separate 230-240 volt power distribution grids since the arc-lamp era and controlling the ligths via central control.

The CdS is quite long time replaced by the silicon sensors in the photocells, even way before the RoHS took any effect.
One of the main problems with the CdS is it's degradation over time and consequent parametric shift, which causes the light threshold to shift up, till it becomes a dayburner. So since the electronic and silicon sensors became cheap enough (in the 80's or so), practically all new designs were without the CdS photoresistor sensor.
Other thing is, the CdS is quite expensive, if it is supposed to handle high power dissipation (more than 0.1W or so), which means the concept of relais dire4ctly controlled by the CdS photoresistor became way too expensive over the systems using electronic amplifiers. And because the amplifiers are inherently very sensitive, the weak signal of even cheaper silicon detector could be processed and so bring the cost even lower (compare to the CdS).
Another benefit of the semiconductor sensor: It could be well made sensitive only to the IR part of the spectrum. Together with the most light sources being some sort of high efficacy source (include MV), the lanterns do not emit that much of the IR compare to the thermal radiation of the sun (daylight). So the sensors could be made way more selective to respond only on daylight and not that much on light from other lanterns.


The individual vs group control:
That is mostly a legacy thing, coming already from the gas lantern age. The first ones were necessary to light one by one by the light man. So to save cost, they were just connected to the gas distribution piping. Of course, all this installation was very expensive, so it was build only in the richest districts and countries, one of them was England. And we have the first individually controlled systems, then spreading over the whole UK territory.

Then the cost went down a bit, plus the shock wave remote ignition systems were invented, which allows for the lighting to suffice with way less attention. That allowed even the not that rich cities to use the public lighting system and that were many cities in the continental Europe or so. Because of the ignition shockwaves, the piping must be dedicated for just the lighting, hence the first group controlled system.

The following, even cheaper to operate (so installed as well in even the way less rich cities) arc lamps need series circuit on a dedicated generator. That imposed separate wiring and group control as well, spreading across most of the continental Europe.

This then made the individual control a standard in the UK (and related countries), with all the automation introduced later (the clockworks, then the individual photocells)
The group control with separated wiring then became the standard for the continental Europe. The automation was then brought to the main control boxes.

And because once set, no standard is easy to turn around...

That's intriguing Medved. All of the photocontrols I've seen or own are clearly with a CdS photoresistor visible in their window. Have I missed something? Is that a silicon sensor I see in my Torks? Is it possible to direct me to a slicon PC here on LG?

That's intriguing Medved. All of the photocontrols I've seen or own are clearly with a CdS photoresistor visible in their window. Have I missed something? Is that a silicon sensor I see in my Torks? Is it possible to direct me to a slicon PC here on LG?

Practically all modern ones with the transparent case, loaded with all the part night features or so.
An example...
Another example...
And yet another example...

We have a `best of both worlds`, photocell/timer hybred set up on our yard lights at work.
The cell switches all the yard lights on as it starts to go dark, then the timer switches them off at 9pm, (an hour after the site has closed to public).

It comprises of 1 cell and 1 digital timer switching a bank of 10 x 3 phase contactors to split the load up all over the site.
It saves having to have the lights run until midnight with a standard `part night` cell set up.