Very nice to see.!

Awesome shot! It's the first time I see the run-up sequence of the NA-1, thanks for sharing that. I acquired this lamps nearly 25 years ago but never got around to lighting it due to the lack of proper gear.

@Sammi
Thanks!

@Max
Just need two filament transformers with 1.5V @ 2.5A and a ~4A current limited source. If I were in your situation I would just wrap two 1.5V secondaries on a random transformer and use any 400W ballast on a variac for the main supply. Though of course any non-OEM setup has risks and you seem to have a lot more self-control than I do lol . Breaking one of those lamps would be nauseatingly unfortunate! Cool that you found one of these

Thanks for the tip, I'll definitely try that. I have to find a way to insulate the lamp properly so it can run up to its nominal temperature.

Ahh right, no dewar jacket. I have seen others operate these without dewar jackets and they seem to work decently (probably not as bright of course), I don't think it will damage the lamp but I don't know and haven't tried it myself. Maybe a simple glass jar or globe placed over it would do, just to limit the airflow a little bit. If you do give it a try I would love to hear the results!

Cool to see it lit. I'd guess running it without a Dewar will work it hard and keep the voltage higher than for full sodium vapour. A simple jar will help but not be the best. As a kid I broke a SLI/H lamp I was trying to cover when a cold draft caught it and cracking it when running.

@Tuopeek
That is a good point, I haven't taken measurements to compare its specs while covered and uncovered. That must have been devastating to break an SLI

Tuopeek - good point about the higher running voltage. Even if the lamp runs with its typical sodium yellow color, too little of this metal in the gas phase will impact the voltage drops across the discharge and the electrode sheaths. Although the external heating of the electrodes should limit the risk of damaging them, I think it is still a good idea to get the lamp to run at its optimum temperature. I see two options there: use one of those double-walled tea/coffee glasses that are widely available in shops (although I don't think they are evacuated like a proper Dewar flask), or use a cut top section of a SOX jacket. The latter should work well given the high IR reflectivity of its inner ITO coating. I guess it's a matter of trial and error, and the lamp voltage should be a good indicator of the thermal insulation effectiveness.

Next time I am home I will measure the voltage during striking and warmup to see what it does. I know that most HID lamps exhibit a lower voltage drop while they are cold, but I am unsure if this same phenomena occurs with LPS lamps. What better way to find out then to measure. If the voltage is indeed elevated at lower temperatures then yeah that probably isn't great for the lamp.

Yeah I agree I doubt those double-walled glasses are under any sort of appropriate vacuum. Good idea with the SOX jacket, I didn't think of that. It would be a shame to cut open a lamp for this, but if you already have a broken one then no harm done. If not I also just found a cheap Ebay listing for a 30 ohm/square ITO coated PET film (roughly SOX thickness I believe), I don't know how heat resistant PET is but that could probably be used for this purpose as well. Just roll it into a tube and give it a good cover to seal it up, or maybe sticking it to the inside walls of a jar would work.

I like the idea of using a double walled drinking glass. I can confirm they are not evacuated and are at atmospheric pressure but the double glazing will be quite effective at reducing some of the heat loss.
It would be interesting to know if the electrode external heating lowers the voltage for the colder lamp vapour. The voltage range does indeed fall for LPS as opposed to increase for HID but by a much smaller margin. I checked for a 35W SOX and the run-up voltage on the starter gas is about 125V dropping to about 80V in my test with an older lamp (Spec says 70V). Interestingly, as Max pointed out, the voltage drop occurs once the lamp appears to be in full sodium vapour emission and not during the colour change phase.

Whar about using a dewar jacket from an SO lamp.?

If you can find a spare one Sammi. I think it may also need to have the ceramic base removed as the access profile is for the narrow shape of the U-tube rather than this round style.

We might have a spare one, so if Drew could send that NA-1 to us we'll give it a try...

@Sammi
That could work if the lamp would fit, I think they have a major diameter of 1.5" but next time I am home I will measure it to be sure. I don't know the opening diameter of one of those dewar jackets. It would be a real shame to have to permanently deface one of those dewar jackets, if that is the case I personally think using the envelope from a SOX lamp would be a better idea since you can still buy those fairly reasonably, and since you can cut them down to whatever convenient length you want unlike a dewar jacket.

Edit: I think 1.5" is not including the ridge at the bottom of the upper envelope, so probably like 1.75" or something around there

Edit:
@Sammi
Just measured now, major diameter is around 1.66". I have no idea if that will fit in an SO dewar jacket without modification or not

@Tuopeek @Max
I measured the voltage, at striking it is 10V, which then slowly drops to a minimum of around 8.8V before later rising up to around 10.5V and settling there. After removing the dewar jacket with it running, the voltage dropped to around 9V and stayed there (with a brightness decrease of course). So I believe that running it without a dewar jacket will actually reduce lamp wattage and therefore not cause a lifespan problem if I am interpreting this correctly. I suppose the voltage drops as sodium gets into the arc, and then rises a little as the pressure increases with higher temperatures. This was measured with an RMS meter

Of course none of those numbers align with the specified 6.5V that this lamp is supposed to run at, but that is probably due to cathode wear.

Thanks for the update, Drew. Your electrical measurement is very interesting, I expected a much greater difference in voltage... I guess this is the result of a very short electrode gap length and an operation at high current. In any case the voltage change following the removal of the Dewar jacket is too small to enable an accurate estimation of the lamp's thermal insulation, so I won't use that approach when I'll run the lamp, especially since I have no idea what lamp-to-lamp variability in electrical characteristics I should expect with the NA-1.

Also, I think you're right that operating the lamp without its Dewar flask won't impact its life expectancy - not because of its reduced wattage, but because of its voltage which remains low (thus there is little risk of an enhanced electrode sputtering). The oxide-coated electrodes are doing an effective job here (likely because they are externally heated!) and the lamp is clearly designed for its discharge to generate more than enough heat to ensure that there are enough sodium in the vapor phase, even without the jacket. That's not the case of SOX lamps whose thermal balance is much more optimized to ensure the highest efficacy possible.

@Max
Good to know it's the voltage that is the main concern with sputtering, and also good to know that running without a jacket is unlikely to cause a problem. I was also expecting a much greater change more similar to SOX lamps but like you said these things operate pretty differently.