Replaced an old, worn-out Universal Watt-Reducer 446 ballast that I had inside one of my garage lights, which was becoming hard to start. I replaced it with the Advance REL-2S40-SC electronic ballast because I was able to get one cheaply on eBay, and I have had good luck with that model so far in one of my other fixtures. It actually has a true rapid-start startup, which is uncommon for electronic ballasts.

Setting safety aside (assuming it's the 1940s), would it be more efficient?

When quartz pinch seals are made, the seal is typically made with molybdenum foils going through the quartz. I have the following questions about this process:

---

1) Why Molybdenum?
Quartz has a CTE of around .55e-6 which is very very low. So you would want to use the metal with the lowest CTE that you can find, right? Well (according to Google AI overview) molybdenum has a CTE of 5.2e-6 while tungsten has a CTE of 4.8e-6. Why wouldn't they choose tungsten for this application?

2) Definition of "Foil"...
So they use molybdenum foil for these seals, but what exactly do they mean by foil? Are we talking about .1mm or something like .01mm? I would assume the thinner the better, but obviously there is a practical limit to the thinness of foil.

3) "Feathered Edges"?
I always hear descriptions of Mo foil seals mentioning that the edges of the foil are "feathered". What is that supposed to mean? When I look at a halogen capsule all I see is foil, I don't see any visible edge treatment done to the foil.

4) Vycor Glass?
I just learned that there is something called Vycor, which is 96% quartz, but technically a type of borosilicate glass. It has a CTE of ~.75e-6. Was this ever used in lamps? I feel like straight up quartz might be overkill for something like a small halogen capsule.

---

Thanks!

I have read that some GE linear halogen floodlight fixtures would take 208V linear halogen lamps.

Now that I have a step up transformer that converts 120V 60Hz to 208V 60Hz, I am now able to comfortably run 208V 60Hz only HID ballasts. If anybody has been wanting to get rid of 208V 60Hz only HID ballasts, I am more than happy to take those ballasts off your hands!!!

The SOX lamp needs to be hot (i.e. be at high temperature) in order to work. The temperature itself does not represent power, but power in the form of heating must be continuously provided in order to compensate for heat lost out of the lamp due to radiation, conduction and convection

Lets say we don't have any other limitations on the efficacy of the discharge itself, and we can make it as efficient as we want to. But if we make it too efficient, we will reach the point where the lamp generates less heat than the heat lost to the environment, so the lamp simply won't work

How can we minimize heat loss to the environment :

 - Radiation : The IR coating of SOX lamps does attempt to reflect as much as the radiated heat back into the lamp. I imagine that there is a compromise to be done between reflecting the IR more effectively vs. not blocking too much of the actual light output. Since blocking the actual light has the most direct impact on efficacy, the balance is towards blocking as little light as possible, even if this would mean letting some of the IR out

 - Conduction : The conduction of heat from the arctube to the outer environment through lamp components in contact is fairly minimal, in the form of 2 mica or small metallic holders near the arctube ends, and through the connecting wires. I imagine that the heat lost there is minimal as is. The exact shape of the holders can be optimized to make it touch the arctube only in a few small points and not along a complete line or so, but in the end it will hit practical limits of how fragile it can be made before it breaks in normal use

 - Convection : The vacuum in the outer envelope effectively eliminates convection

Now that the heat loss is reduced to minimum, we can try to make the discharge as efficient as possible



Discharge is more efficient when the electrode loss is as low as possible. It can be zero in an induction lamp (but induction lamps have other mechanisms of losses), or otherwise it can be minimized by making the discharge as high voltage & low current as possible

With magnetic ballasts, making the discharge voltage more than approx 60% of the ballast Voc would make it hard to maintain a discharge

As long as we suffice with 230 Voc (for 230V line voltage), the ballast can be a simple series choke, which can be made very efficient (over 90% easily achievable with fairly standard size ballast, higher % achievable but with bigger and heavier ballast)

If we want to go higher, the ballast will have to be a transformer type, which makes it significantly less efficient for the same size. As long as we stay with magnetic ballasts, it makes sense to compromise in the lamp design to avoid a transformer

With electronic ballasts the limits are a bit less strict. It is possible to raise the voltage limits a bit before we get significant efficiency impact

The discharge itself is a little more efficient with HF driving. The efficiency of the ballast with the best electronic ballasts is approx. on par with the best magnetic ballasts - The efficiency advantage of HF comes from the discharge. (And some additional claims for efficiency advantage come when average or above-average HF ballasts are compared to the worst magnetic ones)

The discharge may be more efficient with different buffer gases. I am no expert on the matter



Some of the light emitted by the discharge is blocked by the vapors in the arctube in which the discharge takes place. This have been addressed in SLI/H lamps :

 - The linear construction means that there are no 2 tubes side by side, which block some light from each other

 - There have been arctubes made with a cross section shape that makes the arctube design flat - a shape which resembles an asterisk (Thorn) and a PG FL lamp (GEC)

However, linear lamp means larger outer lamp surface area for the same discharge power density, which may increase the heat loss from radiation, so is a compromise. The surface may be minimized also by increasing the power density (higher power in a shorter lamp) but this then means compromise on other things like fill pressure and electrode losses



With the as high efficiency as possible in the discharge and in the avoidance of light blocking, the heat generation is reduced. Ideally it must be reduced to the point where it just barely overcomes the heat loss from the lamp to the environment

This "Barely" implies that :

 - The lamp will not work at reduced power. So the ballast must provide precise power level. In case of magnetic ballast, this will also mean a compromise vs. the required tolerances for line voltage

 - The lamp will take ages to warm up, or the ballast must be designed to increase the power during warm up and switch to normal power level when the lamp reached working temperature

Both issues can be overcome with an electronic ballast. The 2nd can also be done with a magnetic ballast with a tap and a control module

Just the fact the lamp needs to operate large chunk of material pretty hot (the whole arctube, at around 200 degC) means it takes ages to get it to that operating temperature (normally 15..20 minutes, more with higher efficiency designs). And that by itself would make it rather impractical for any application requiring more frequent starting.
I think that is also a reason, why the commercial designs end up with cold started lamps: If it is not usable for more frequent starts and when the steady 10h/start burns mean even the hard cold started electrodes are not the weakest point anymore, it does not make sense to make the lamps more complex just to reduce the starting wear, it won't have any practical value.

Each hole makes it weaker and an unfilled hole would be rather problematic (you won't be able to use foam fill as a secondary arcing protection).
Plus you won't fit that many extra holes while still keeping the necessary isolation distances, maybe one and that it is.
Plus to make the base contacts so a "wrong lamp" won't make any bad connection would make the socket into a precission machined product. Everything but cost effective.

And you will solve just part of the tooling problem, each pin combination would still require its own jigs to guidethe inserts to make the contacts.

Don't forget that today the only difference between 60..100W lamps is the filament and the etch stamp. The rest stays exactly the same. And when extending the range downto 25W, you just get different fill pressure settings (down to full vacuum for the 25W).

I think you're really close to it if you check out the specs of the final generation of Philips SOX-E lamps.

Thallium is dangrously toxic, much more than Mercury. A broken lamp would very likely be a situation on par with a hazardous chemical spill, requiring evacuation, professional cleanup etc

(I don't know what quantity of Thallium would be needed to get the correct vapor pressure in such lamp, but it is way more than the tiny quantity used in some MH lamps. If it is anywhere near the quantity of Sodium in a SOX, it is immediately dangerous to life)

So do we need a lamp that is basically FL, but more complex to make, needs warm up, and is dangerous ?

Well I would imagine that would be a problem. And that also probably explains why such a lamp was never created.