The timer and its power supply are on all the time there is unswitched steady live present, and the Zener diode makes sure that the voltage on the timer chip is more or less constant by absorbing the surplus current when the triac is not firing, - so in turn the voltage on the dropper resistor is also more or less constant too, at about 220 of the 230V... (i think before it cooked it was probably red red orange, so 22K ohms, about 10mA constant load, a couple of watts, but the colours are burnt so maybe not)

Personally I prefer capacitor dropper designs, as they  run cooler but they are mains frequency dependant,  and still need some series resistance to keep the current from  becoming near infinite during supply switching transients, and on non-sinusoidal waveforms like some inverters they fry completely. A pure resistor is more or less universal.

I suspect this one worked really well on the lab bench and was then boxed into the fan casing without thinking that the airflow had changed.  Much  like cable ratings a 2 watt resistor is only really able to sweat off  the full 2 watts for a given set of ambient conditions,  and must be de-rated when the heat is not free to escape. Two resistors of the same physical size and half the value, in series would have an easier life.

Sometimes you see ceramic or glass spacers on the legs to lift the resistors up a bit, or slots in the PCB, in an attempt to get a few % extra airflow under the device. These are the classic signs that the designer realised they were sailing close to the wind and is trying to improve reliability.

This one didn't realise..

Once the board is charred it conducts a bit, and depending how charred and where, that can start to affect the timer  by leaking charge into or out of the timing capacitor.  In the limit it becomes unusable.



There is a lower power version that uses the spare gates to turn the steady state triac firing pulse into short bursts - unlike a transistor, a Triac stays on once triggered until the next zero crossing so you do not need a continuous turn on signal. By using the low voltage DC cap to provide the short turn on burst you can have a lower average current, and run cooler. This is much like the tank on the loo allows you to use a a half inch filling pipe, and get a short duration flush of full flow in 2 inches diameter - as opposed to installing a fat pipe all the way back to the street main and needing no tank..

Mike.

The timer and its power supply are on all the time there is unswitched steady live present, and the Zener diode makes sure that the voltage on the timer chip is more or less constant by absorbing the surplus current when the triac is not firing, - so in turn the voltage on the dropper resistor is also more or less constant too, at about 220 of the 230V... (i think before it cooked it was probably red red orange, so 22K ohms, about 10mA constant load, a couple of watts, but the colours are burnt so maybe not)

Personally I prefer capacitor dropper designs, as they  run cooler but they are mains frequency dependant,  and still need some series resistance to keep the current from  becoming near infinite during supply switching transients, and on non-sinusoidal waveforms like some inverters they fry completely. A pure resistor is more or less universal.

I suspect this one worked really well on the lab bench and was then boxed into the fan casing without thinking that the airflow had changed.  Much  like cable ratings a 2 watt resistor is only really able to sweat off  the full 2 watts for a given set of ambient conditions,  and must be de-rated when the heat is not free to escape. Two resistors of the same physical size and half the value, in series would have an easier life.

Sometimes you see ceramic or glass spacers on the legs to lift the resistors up a bit, or slots in the PCB, in an attempt to get a few % extra airflow under the device. These are the classic signs that the designer realised they were sailing close to the wind and is trying to improve reliability.

This one didn't realise..

Once the board is charred it conducts a bit, and depending how charred and where, that can start to affect the timer  by leaking charge into or out of the timing capacitor.  In the limit it becomes unusable.



There is a lower power version that uses the spare gates to turn the steady state triac firing pulse into short bursts - unlike a transistor, a Triac stays on once triggered until the next zero crossing so you do not need a continuous turn on signal. By using the low voltage DC cap to provide the short turn on burst you can have a lower average current, and run cooler. This is much like the tank on the loo allows you to use a a half inch filling pipe, and get a short duration flush of full flow in 2 inches diameter - as opposed to installing a fat pipe all the way back to the street main and needing no tank..

Mike.