Light Fitting Rating

I can't work out from your post whether it's the new fittings or the old fittings that are limited to 60 or 75 watts. However, the new LED bulbs will use less, so that's okay then.

Quite a lot of new LED fittings come with bulbs already fitted. These usually can't be replaced.

Maybe the simpler answer was just "yes"?

The limit is because of the heat generated, which is the actual wattage, not the "equivalent". No household LED bulb will approach the limit.

-- Richard

Indeed, the rating is to do with the amount of heat you can emit before melting / setting fire to the fitting.

Although LED bulbs in some fittings have been known to gradually cook themselves which shortens their life, well below the amount of heat needed to damage the fitting. It's worth thinking about ventilation of the fitting

- eg an upturned bowl will collect the heat, whereas a cylindrical lampshade open at the top and bottom will allow fresh air to circulate to cool the bulb.

Theo

I don't have any figures, but I'm doubtful about how significant this is. I suspect that the internal temperature of the bulb depends more on the design of the bulb itself than on the surrounding air flow. Cheap bulbs are likely to lack enough metal to conduct the heat away to the air.

-- Richard

Theo is right. I have some small ceiling lamps designed for G9 halogens. They point down, but can be angled slightly as well, are made of glass, and have no ventilation at the socket end. when I changed to G9 leds I was surprised at how quickly those bulbs died. I assumed they overheated, and tried a different make, but that also failed fairly quickly.

I eventually found that some Wilko 3.6W (33W halogen equivalent) lasted well without dying. In fact, I have a couple of spares which I haven't needed for 5+ years. Those bulbs are somewhat longer than the other G9s and perhaps aren't so susceptible to the heat.

With incandescent and halogens, they produce 90-95% of their rated power as infrared rather than visible light, which can escape somewhat.

LEDs tend to generate heat within the LED itself, so even if they are lower actual power to begin with, 40-60% of that power ends up as heat, with limited airflow for cooling, can tend to cook ...

essentially yes

You are correct

Also a tungsten filament is by nature hot up to about 3000K. Semiconductor junctions are good to about 125C, but higher temperature causes faster ageing, well below the tungsten temperature.

Persistently elevated temperatures can also damage other components like capacitors that are included as part of the lamp - if your capacitor dropper loses capacitance then it's not dropping so much voltage, you start to overvolt/overcurrent the LEDs, they run hotter, and conditions continue to deteriorate until something finally goes.

Theo

Nearer 180°C these days

Yes. I've had a lot of trouble with a room, used a lot, with a pair of bulbs hanging down. LEDS keep popping even though the bulbs have plenty of airflow.

other high use rooms have either 'candle' style bulbs or spotlight bulbs. These seem to last better

Your technical description of the cause of events is not entirely correct (a fall in the capacitance of the dropper will lead to less current at the same voltage drop, making the light dimmer) but I don't disagree about the end result!

Capacitance dropping will underrun the leds.

There are two capacitors.

Or maybe more. So - doesn't alter the statement that that reducing the voltage dropper capacitance will under run the LEDs.

If you can find capacitor dropper LED bulbs BigClive has info here on how to hack them, but as that page says LED bulb design has moved on. Many , probably most, LEDs bulbs no longer use Capacitor Droppers as its hard to achieve the required power factor. Instead they use multiple LEDs in series to allow them to be driven at close to mains voltage, with a constant current chip to keep the dissipation constant.

If you want to extend their like take a look at this "Big Clive" video which I found interesting on how these work, why some are cheaper than others and how to lower the dissipation...

And reducing the larger value capacitor in the dropper will *increase* the LED drive.

It works pretty much like a resistive voltage divider if the current drawn is small, remembering that the larger the capacitor, the smaller the reactance.

No it doesn't. The voltage across the LED is largely governed by the properties of the LED, and putting a large capacitor across it might vary the reactive current but doesn't act as a voltage divider at all.

The problem is that once you introduce a non linear element like an LED you can essentially throw all of the linear analyses out of the window.

A large part of circuit design consits in arranging things so you dont have to deal with that.

For instance by setting bounds within which linear approximations are valid.

Exactly... And using parts which have linear characteristics to compensate for those that do not