One thing that was always intrigued me is the use of uncoated mercury lamps. Seems to be mainly Amercian thing. Never seen such a lamp in use in Europe or ex-Ussr, until it is not a case of broken outer bulb of course. May be except a rare case when tubular sleeve-less mercury lamps were used in bulky square ground luminaries in a park to light up the trees.

What is the reason to use clear mercury lamp, other than a tradition? Light output is about the same as coated, but the CRI is definitely much worse.

Hi RRK, i wonder this too!

May be better optical control, since bare mercury arc is much more concentrated?

I think someone here once mentioned that the reason clear mercs stayed in use in USA was slightly cheaper price compared to coated lamp. Based on what I have heard, coated lamps replaced clear lamps pretty quickly in europe.

Cost is indeed the main reason.  Many phosphors require exotic rare metals.  For instance in a typical modern triphosphor fluorescent lamp, the cost of the Europium metal in the phosphor represents as much as half the total raw material cost of all the other lamp components together.  Due to higher energy costs in Europe and the approx. 10% improved efficacy of coated lamps, they quickly became popular.  In the Americas electricity used to be so cheap that it was more cost-effective to make a saving on the lamp purchase price and live with higher energy consumption for a given light level.

The only main exception is for when mercury lamps are used in certain older luminaires that were originally developed for clear lamps.  The clear versions remained in production so as to deliver the same optical performance.

Another exception exists in case of floodlighting of trees and city parks, especially in Asian countries, where clear mercury lamps are preferred because of their superior rendering of green colours.

James, thanks for the explanation.

But why if the cost is main reason, American manufacturers just did not settle to using cheaper old generation phosphors, like USSR lamp plants did, using halophosphate?

I've noticed on my mercury cobraheads that there are more light distribution options available using a clear mercury lamp compared to a coated one. As a example I've attached the chart showing the different socket positions and the light distribution patterns available on my Sylvania B2227 (250w MV, Mfg 1982). Clear has a lot more distribution patterns available compared to coated.

I suppose that unlike Europe were clear lamps pretty much died out for general lighting, the street luminaires here were never fully optimized for coated, instead they were optimized for clear. I guess do to most luminaires here using prismatic glass optics.

Speaking of cost and using a older style of phosphor, there were some low cost "coated" mercury lamps available here from manufacturers such as Feit that used a simple diffuse coating instead of a phosphor. They were mostly sold in hardware stores as replacements for homeowners. As expected those pretty much have the same colour rendering as a clear lamp but with the optical performance of a coated lamp.

Wasn’t lumen output another reason for it? I know that clear MV has more lumens than any of the MV phosphors and is thus a much brighter light source.

Isn't it opposite just as James said,

approx. 10% improved efficacy of coated lamps,

That’s in terms of color rendering I’m pretty sure.

There's a bit more nuance to this, whether or not a phosphor results in reduced or increased lumen output over a clear lamp depends largely on the type of phosphor used. I compared the specs for some Westinghouse 400w lamps and here's what I found:

Lumen output relative to a 400w clear lamp:

/C - 5% decrease
/W - 12% increase
/Y - 30% decrease
/X - 33% decrease
/DX - 7% increase
/R - 7% increase
/N - 9% decrease

The large cut in lumen output and efficacy was probably why the /X and /Y lamps were rather short lived. /X got replaced by the /DX lamps and the introduction of HPS killed off the /Y. I also didn't know the /R was on par in terms of lumens as the /DX as well didn't know the /N lamps actually outputted less than a /C.

Well, /C type seems to be fluorogermanate, nice color at slight loss of the efficacy
/W is orthophosphate
/Y is yellow filtered, killed by the availabilty of yellow sodium lamps
/X - don't know, some warm filtered version?
/DX - europium phosphor
/R, /N - don't know what it is.

/W gives maximum light gain, CRI is still poor, but the phosphor is dirt cheap too.

The American manufacturers indeed chose to offer the old cheap phosphor /C and /W types for many years after the superior /DX was introduced, and evidently many customers bought them.  But the catalogues indicate that the clear types were still considerably cheaper.

Eventually once mercury sales volumes began declining it was costing more to manufacture and stock so many variants, and one by one they discontinued all except the /DX and clear.  The logistical savings justified to sell the /DX types only at more competitive prices. Recently the selling prices of some clear types surpassed the technically more expensive /DX only because of the low volumes and the hassle of changing over production systems to make them.  Ontop of this, continual improvements of /DX phosphors allowed their luminous flux to eventually exceed 110% of the flux of a clear lamp.

Yeah clear lamps have become increasingly harder to find in more recent years compared to /DX and as you mentioned usually at a higher price point too in most cases. On some more common types such as 100w and 175w, sometimes clears and /DX were the same price.

I've also noticed around the same time manufacturers were shrinking their mercury lamp lineups, they also gradually eliminated some clear wattages, 50w and 75w clear lamps seemed to be the first to go but even 1000w isn't that easy to find now in clear.

Based on what I have heard, coated lamps replaced clear lamps pretty quickly in europe.

Narva continued to produce clear lamps well into the eighties, perhaps up to 1990, years after they phased out the orthophosphate and fluorogermanate versions. No idea why they did this, these clear lamps appeared at random in streetlighting installations (I posted a few photos over the years here).