Hi all, i know induction lamps works with mercury.
What about making an induction LPS?
I know this will never made, but i wonder if it would work.

Why not. But question is, would the lamp last longer this way (before it fails from something other then electrode depletion, for example, sodium corroding the glass)

One of the largest challenges with LPS design is the heat retention, so to keep even the coldest spot of the arctube hot enough to have reasonable sodium pressure for the lamp operation. That means everything should be either at some distance from the arctube, or operate hot as well.

With inductions the main problems are the tight coupling between the coupler coils and the discharge, while still keeping the couplers reasonably cold. So it should be close to the discgarge and at the same time not allowed to heat up.


With mercury you want rather low temperature for an efficient discharge, so both discharge and coupler requirements are the same, which allows the inductions based on mercury to become a real thing.

But with the Sodium one requirement quite excludes the other, so I can not imagine how to combine them...

Make the coupler able to stand high temperature. Hey, its just a coil and core. Sure better isolation materials can be used

... and the core should operate with low losses at HF at high flux densities and it's Currie temperature should be above the operation temperature, so due to the coupler losses sufficiently above the 210degC.
The later puts out practically all ferrites, the first practically all metal alloys.
And I do not see much materials left, in fact I do not see any...

Plus the material for the coil should have low resistivity at the operating temperature as well, the copper would be tripple the normal resistivity...

Arund 200C not sufficient ? That would be still reachable with some higher temp isolation materials and ferrite, the rest can go into the design to limit the temperature there :

 - Prismatic sleeve with IR reflector film, so that it reflects the IR not into the same section of the arc tube and heat it more, but to the sides away from the section that is under the coupler

 - Varying arc tube cross section in the spot where the coupler is attached, to lower the relative power dissipation there (and dissipate the power more elsewhere in the tube)

 - Air gap between the coupler and tube. There is the few mm distance anyway - the width of the vacuum space in the outer sleeve, what change would couple more mm in the air make....

 - Actively cooled coupler by heatsinking fins, heat pipes etc (shaped in the way that they dont form a shorted secondary turn)



The copper losses will be larger.. But maybe they will be still acceptable. If no, maybe the copper cross section can be enlarged and there still is a point of tradeoff between size and losses

That would be great for safety & security lighting - Too bad it didnt make it

With the Borate lined glass, how is it fused without making the inner coating inconsistent ? - in places like electrode presses (in ordinary lamps), fused glass "elbows" for turns, tip off tube ?

With the hot operation I didn't mean it is impossible or difficult to run the coupler cold by itself (indeed, the QL uses it at very high power densities), but I assumed for a reasonable discharge burner design, it would require to run the coupler hot, so it will not cool down the burner.

Actually I had in mind the Icetron/Genura concept as more favorable, just because of the simpler shape of the burner (no cavity with rather sharp edges for the coupler insertion, just a rather simple tube), so easier to manage the Borate protection layer.

With the double glass around the coupler I may pretty well understand, why it never catch up: Very complex shape, so the quality of the protection layer would be questionable, making the assembly expensive, the insulation not as perfect as with classic LPS concept (with induction you can not use the IR reflector - it would form a short circuit for the coupler, or require even more expensive to make isolation gap in the coat) making the lamp less efficient and the space occupied by the insulation making the coupler rather lossy, so further efficacy loss.
With all that, I doubt it would be reliable any close to the simple QL burners, mainly on account of the manufacturing defects.

so, it was an insuccess...
Ok, thank you all!

And now, another idea:
Medved, what about an mini-magnetron tuned between 400-700THz?
I'm sure it's quite impossible (this would be very small and capacitance would be an problem), but this would be an nearly perfect monochromatic source.

At first assume you would be able to accurately manufacture parts that whatever small.

Then first problem to face would be the output power: Normal magnetrons are of similar size as the wave of the wave they need to radiate. With cm range microwaves it means few cm, soi with half micrometer visible it would mean micrometer size. That means the volume would be 10^12x smaller, so with the same average power density it would mean power levels of just nW. That is not that much.

Second problem coming from the size: With such linear scaling to work it means the speed of the electrons stays the same. But the same speed means the same acceleration voltage. Well, I really have no idea, how to handle the 10kV+ at a distances of a micrometer...

Third problem: The accelerated electron trajectory would have to be bend to the same 90deg angle, but at distances 10000x smaller. As the bending is what the "magnets" ae for in the "magnetron", these magnets would have to be 10000x stronger, so in the 10000T range. Well, except for the core of a nuclear explosion, the strongest magnetic field ever created by mans was just few 100's Tesla. And it was just for very short time and it's generation involved high speed explosives (used to physically shring a conductor loop, so constrict the magnetic flux into smaller area). And even within the nuclear explosions, the magnetic fields were brief in time. With magnetron you need the field static all the time the device operates.


I think the free charge quantum annihilation method works with lesser problems...

At first assume you would be able to accurately manufacture parts that whatever small.

Then first problem to face would be the output power: Normal magnetrons are of similar size as the wave of the wave they need to radiate. With cm range microwaves it means few cm, soi with half micrometer visible it would mean micrometer size. That means the volume would be 10^12x smaller, so with the same average power density it would mean power levels of just nW. That is not that much.

Second problem coming from the size: With such linear scaling to work it means the speed of the electrons stays the same. But the same speed means the same acceleration voltage. Well, I really have no idea, how to handle the 10kV+ at a distances of a micrometer...

Third problem: The accelerated electron trajectory would have to be bend to the same 90deg angle, but at distances 10000x smaller. As the bending is what the "magnets" ae for in the "magnetron", these magnets would have to be 10000x stronger, so in the 10000T range. Well, except for the core of a nuclear explosion, the strongest magnetic field ever created by mans was just few 100's Tesla. And it was just for very short time and it's generation involved high speed explosives (used to physically shring a conductor loop, so constrict the magnetic flux into smaller area). And even within the nuclear explosions, the magnetic fields were brief in time. With magnetron you need the field static all the time the device operates.


I think the free charge quantum annihilation method works with lesser problems...

Ok, thank you 
So, i have to create an nm-sized magnetron powered by an atomic bomb!