SOX Lamps, couple of queries re lamp design and common problems

[Zelandeth]

Just been writing up the page for a couple of my SOX lamps (two of which are a bit of a mystery thanks to having been used in leaky street lighting fixtures which have washed all the text off!).? I'll worry about those two later though!? Firstly, My Osram SuperSOX 18W lamp.

Data I need: Lamp voltage & Current.? Lumen output.? Rated Lifetime.? Ideally I'd like to know what it cost when new (1988) and what it's worth now, but that's really not important!? Incidentally I bagged this lamp complete with an immaculate GEC-Osram NightWatch NW18 fixture off eBay earlier in the year for just under ?20.

I've got a general query about problems which can happen with SOX lamps though.? I know that if the getter in the lamp comes into contact with one of the electrically charged components that this is bad news - especially if this also comes into contact with the semiconductor film, as this can lead to electrolysis causing sodium to be drawn into the glass, eventually causing cracking.? (Can anyone tell that I've read through the Sodium vapour lamp tech page on Lamptech just a few times?).? I believe that this has happened with this lamp, having discovered that there appears to be a "short" circuit between one of the electrode connections and the indium oxide film.? This I discovered when playing around with a 12V CCFL driver.? The net result of this was me getting a zapped finger from the base of the lamp and looking really puzzled.? I discovered that it was possible to get a discharge from the lamp with a connection to one terminal and the other wire held up against the glass.? Due to the CCFL driver operating at some stupidly high frequency it'll conduct enough to the film inside to get a glow (which is all you'll get anyway from this), it wouldn't work to the other terminal though, just acting as though it was shorted out.
? Enough rambling though (and yes, I know not to play around with high voltage things like that), What I was wondering was whether this connection between one of the leadwires and the semiconductor coating would be why the coating of this lamp appears far more opaque than on other SOX lamps I've seen?? Would this be due to some sodium having been drawn through the discharge vessel?? The lamp's got very little use on it (previous owner used it for viewing holographic materials I believe), so I would not expect any signs of failure by this means to be very obvious yet.

And just because it's an excuse to post pics of my collection...here's the lamp in question.



Also...would someone like to explain to me what the different cap colours mean on these lamps?? I'm assuming it's a quick way to identify the main "models" of SOX lamp?? White being plain SOX, grey SOX-E, red SOX-Plus, blue the newer SOX-PSG.? Those are guesses based around both lamps I've seen, and the examples shown on Lamptech.? Confirmation would be nice there!? Let me know precisely what my other two are too!? Though there will be another post about them in a minute.? Going to attempt to keep this one on topic for a change!

EDIT: Just remembered one of the original questions (hence, design in the topic).  I've got a couple of other Osram lamps here (the two I'm going to be looking for details on), and note that there doesn't appear to be any means of sodium retention on the inner tube.  Other than a wider band just before the U-bend (as with the 18W lamp shown here).  Is this what Osram-GEC decided to use as their means of keeping the liquid sodium in these lamps in place, or did they go for the approach of using a graded thickness of IR coating on the outer bulb instead?

[mr_big]

The 35 watt Osram-GEC lamp if you can get more pictures of this id be happy to tell you maybe what happened to it

[Zelandeth]

The 35W one which has failed, I'm pretty sure simply expired of old age, the huge amount of sodium staining on the discharge vessel and amount of electrode sputtering in evidence tending to suggest that.

I've just got to actually get around to writing up the page for that lamp!  Though it'll be a little scarce like the Osram GEC 55W one's page, as all the markings have been washed off, so a few details are having to be guessed!

[mr_big]

It looks like the discharge vessel cracked around the electrode  could you post a picture of the electrode area of the lamp

[Zelandeth]

Yep, it has done.  I'm tending to think that the crack there was actually the main mechenism for the failure of the lamp.  Can't prove it though as the lamp was in a fixture which had been rolling around the back of a van all morning before I bagged it!

Will grab a photo tomorrow evening.  Camera's needing new batteries just now!

[James]

In your 18W lamp it is intentional that the indium film is electrically connected to one of the lead wires.  In this manner the IR film acts as a third auxiliary external electrode, and assists in starting the lamp.  However you will note, that to prevent the aforementioned problems of sodium loss from the arc tube, that the metal top support does not make contact with the discharge tube.  There is a glass tube passing between the centre of the U-bend, and the top support contacts this only, and not the discharge tube.  Consequently no sodium loss is possible.

The slightly opaque film you refer to is normal.  This can happen sometimes during normal production, in which a jet of indium trichloride particles in alcohol is sprayed into the glass tube while heated to about 500 celsius.  Some of the spray hits the hot glass and is oxidised to form the IR coating.  Much of the spray just sits in the tube centre and is not reacted, and if it is allowed to stay there and not enough oxygen is drawn into the tube, you get a milky coating.  Its important to extract the vapours from the tube very quickly and draw in fresh oxygen, but sometimes the process is not so efficient as it should be, and you get this more opaque coating.  Look at a GE SOX lamp where the coating process is not optimised well at all, and you will see that the coating is very much more opaque than on competitive lamps.

On your second question, Osram's method of sodium retention was in the wider diameter rings just after the U-bend that you refer to, used in conjunction with a graded thickness IR coating.  Today Osram's SOX lamps are made by Philips, and the same U-bend mould and graded thickness coating is employed.  Philips lamps are however of course made with the regular dimples along the tube length, plus a non-graded IR coating.

Best regards,

James.

[Zelandeth]

Thanks for that James!  Exactly the information that I was looking for.

[Max]

Hello Zel,

to add to James' comments on the connection of the ITO film to one electrode, I would like to point out that this connection is not direct, but done via a small ceramic capacitor. This way, only transient currents, not DC, can flow through - thus preventing the continuous pullout of sodium ions from happening. This is a reason why early Philips SOX18 made with all-metal supports did not require any electrical insulation with the discharge vessel.

Max

[James]

While this is true for Philips 18W lamps, in the GE and Osram versions a direct connection from one lead wire to the IR coating is employed.  This is possible because both those lamps have insulating top supports.  In the case of the Osram lamp, a small glass tube is disposed between the metal top support and the discharge tube.  With the GE lamp a special mica disc is used as the top support.  This simplifies lamp construction somewhat and obviates the requirement for the capacitor found in the Philips lamps.

James.

[Lightbulb Collector]

I've got an 18W low pressure sodium lamp. But where can I buy a ballast for them? Even Ebay (which seems to have EVERYTHING) doesn't have a ballast for these.

[Max]

Hi,

The electrical specs of a 18W SOX lamp (57V / 0.35A) are nearly the same as those of a 18W fluorescent tube (57V / 0.37A); you can therefore use a 18W fluorescent ballast to drive your SOX lamp.

Max

[Lightbulb Collector]

The running voltage is 50 some volts. The striking voltage is MUCH higher. And therefore it will need a striker like in a fluorescent tube fixture. But it can't be wired like a fluorescent tube ballast (a series inductor), because that NEEDS to have 4 pins on the fluorescent tube to work. but an SOX bulb only has 2 pins. So I have NO clue how to strike the arc. As the starting voltage is well OVER 120 volts.


And HOW do I find a socket to put it in? It uses a SPECIAL type of beyonete socket. But I have NEVER seen a socket like it would need at ANY hardware, or electrical supply store. I haven't even seen it on ebay.

[Max]

The striking voltage of a SOX lamp is around 200V or a bit less, it will therefore start fine when connected in series with a 18W fluorescent tube choke on 220V mains. For 120V network you indeed need a starter to ignite the lamp and again, the SOX can be wired like a fluorescent tube, the trick is to short-circuit each electrode's filament, which is already done in the SOX lamp.
The circuit therefore has the choke that is series connected to the lamp, and the starter which is in parallel with the lamp's leads. After switch-on, the starter will energize, its bimetal will touch each-other which will enable a current to flow in the choke. When the bimetals re-open, the sharp drop in current leads to a voltage spike across the lamp's lead, enough to strike a discharge in it, et voil?!

Max

[Lightbulb Collector]

And HOW do I find a socket to put the 18W SOX bulb in? It uses a SPECIAL type of bayonet socket. But I have NEVER seen a socket like it would need at ANY hardware, or electrical supply store. I haven't even seen it on Ebay.

[Max]

Go to ebay.co.uk and you'll find a few like this one. however, mind that these 230V base are not suited for pulse starting of the lamp, so use them at your own risks.