Ok, heres an interesting one. I found the lamp pictured below in a recessed can in an entryway of a clinic I am currently remodeling. At first I thought it was just another LED par-30...until I switched on the power. It started and ran up just like a metal halide! Actually just like one of those HID/Xenon automotive headlights. It took about 10-15 seconds to reach full intensity, which appeared about the same as a typical 90w halogen par-38, color temp was close too. So of course I took this thing home (the clinic got an LED lamp in its place). There is no way this is a self-ballasted metal halide as there simply is no way a typical 39w MH lamp capsule could fit in this lamp. Also, the only info stamped on this lamp is "40w 65,000h". A kill-a-watt reads 42w, and 99 for power factor! If "65,000h" means 65,000 hours then this has to be some type of electrodeless plasma lamp. Any idea guys??

More pics... And theres an opaque diffuser over the light emitting component as you can see. I dont have a small enough screwdriver to take that lens off, but ill pick one up tomorrow as im dying to see whats under that lens!

Maybe electron simulated luminescent (ESL) lamp?

It could be some form of a single ended MH, but the 65000hour rating would be too long.

For anything microwave I strongly doubt, if it is some form of an induction, it would be more "classic" RF feed (no more than few MHz).

But I have to agree with Max, the best chance is, you will find some form of LED there.
But on that case the "warmup" is really strange. I don't think it may influence the life at all, the chip thermal time constant is in ms range, so a 100ms rampup would be sufficient... What it could be: Some electrolytic is on it's end of life, so it's ESR is too high (and fool the controller so it reduces the power) until it warms up. It does not have to be any of the power components, but just some small signal filter, so not exposed to any power causing it to overheat from the internal dissipations or so.

This is most definitely a gas discharge source of some type, the warmup is just so exactly like an HID warmup. starts a dim xenon-white color, shifts to green as it slowly gets brighter, a bit of flashing different colors, then eventually becoming warm white with a slight pink tinge. As it warms up there is a noticeable high frequency buzz coming from the lamp, when its at full power the HF noise is nearly silent. When switched off and back on it instantly restrikes at about 75% intensity and warms back up from there. I really thought it was an LED too, but we will find out tonight

Interesting.
Deff looking forward to seeing pic's when you get it taken apart!

I too am really interested to see what's inside this thing!

Ok, bad news, that white frosted glass lens is not only screwed to the solid metal lamp surface, but also epoxied I could only get 2 of the three screws out, the third is clearly epoxied in and the 2/3rds of the unscrewed lens wouldnt even budge and is seemingly completely sealed to the lamp "chassis". That glass lens also must be rather thick as it wont break even when striking it with a flathead screwdriver. I could take my metal disc grinder to it, but I then risk completely damaging that mysterious light-emitting component. Ill see if I can upload a warmup vid of it first as maybe somebody will know what it is just by watching that. It really looks exactly like those 35w HID car headlights turning on, exactly the same in fact. Also that un-removable lens gets far hotter than any LED could withstand, I nearly burned myself touching it to see how hot it gets.

 Its gotta be some type of this technology... http://www.luxim.com/ maybe a chinese knockoff?

"Light emitting plasma" are nearly all discharge lamps for already 150 years...

If you hear it, it is for sure not high frequency, with that you would hear nothing (maybe some 100/120Hz buzz when it demodulates somewhere).
So I'm more and more thinking of a single ended MH on rather "standard" LFAC ballast (same as the selfballasted MH's)...

But some squeak during ignition is common with any ballast, that tells nothing about the ballast operation.

But here it could be fed by way higher power densities than the lab sources, for general lighting application the spectrum broadening is welcome than unwanted effect, I would not exclude some either medium or even high pressure operation...
The hot restrike does not have to be of any problem: With capacitively coupled lamp (the small size would suggest that mode) you need high voltage anyway (to pass sufficient current through the rather high coupling impedance) and that will ignite the plasma even in the hot, higher density gas.

Thanks for the explanation! Actually I am quite familiar with indium MH lamps, as I have several blue lamps, as well as daylight lamps that only contain indium. If this mystery bulb is an indium discharge, then obtaining that 3000K temp would mean a very high pressure discharge right? Otherwise, it would burn blue at a low pressure. I could try the CD spectroscope, but im not too familiar with that. Though, with the dozens of different indium lamps ive seen warm up, I can venture to guess there is no indium in this lamp. Most indium daylight lamps still go through a saturated blue phase early in warmup, this lamp does not. After switch on its a dim xenon-white color, then mercury green, then the yellow/pinkish final 3000K color.

Based on Bluehalide description, and the fact that the lamp color is 3000K, this is probably simply a self ballasted ceramic MH lamp. If the lamp indeed operates directly on the 120V mains, use a DVD to capture an image of the spectrum. DVD gives better results with the camera than a CD.

I could try the CD spectroscope, but im not too familiar with that.

You look at the lamp (light source,...) as a reflection in the disc. Then you tilt the CD/DVD further away, till you spot the "rainbowy" reflection. And that "rainbow" is the spectrum.
It's resolution is given by the distances, but as well by the size of the light source. Every wavelength is displayed as reflection image of the complete lighting object, but the image position shifts with the wavelength, so larger objects will smear the spectrum.
When capturing the image, theoretically shouldbe possible to do some post processing to extract back the sharp spectrum, but usually distance of about 1..2m for object of 1..2cm size gives already good results.

The DVD is better when evaluating something in hands, because the lines are denser, so the same wavelength diference causes larger image displacement. But the CD could be better for capturing the image, as the complete spectrum usually does not "fit" onto the DVD at once.
In any case, best to use the "stamped" CD/DVD (just plain aluminum sputtered on a transparent plastic, so no clor filtering), the "R" or "RW" have extra layer on top (the one bearing the recorded data), which may distort the spectrum.

Did you ever find out what this is?