🚀 go-pugleaf

siegman@EE.Stanford.EDU (Anthony E. Siegman) writes:

>But can anyone say authoritatively, if you slowly ("adiabatically")
>vary either the current or the voltage applied to a light bulb
>(letting it come to temperature equilibrium at each setting), will
>there be a *true hysteresis*, or a true bistable behavior, in the
>light versus current, or the light versus voltage curves?  

>Or is the hysteresis only in the dimmer-light bulb combination, and
>really more in the dimmer's response to a nonlinear load than to any
>hysteresis in the light bulb itself?  

Yes - the hysteresis is in the combination.  The bulb by itself would
not exhibit a hysteresis behaviour under normal circumstances.
Normal circumstances being constant voltage.  When you first turn
on a bulb the immediate current is very high because the resistance
is very low which means a lot of energy is being dissipated in the
filament.  The filament has a very low thermal capacity and it heats
up rapdily raising the resistance, lowering the current and lowering
the power.  At some point the energy dissipated will equal the energy
radiated and it stabilizes at that point.

With the dimmer in the circuit you have an entirely different situation.
It is closer to being a constant current.  It isn't actually constant 
current but for the sake of argument I am going to call it that.

The energy dissipated in the filament is I**2*R.  So for a given 
current the power is proportional to the resistance.  At low 
temperatures the resistance is low and there isn't enough power
to overcome the heat losses and the temperature stays low.  If
somehow you can heat the filament (momentarily increase the current
or approach this current from the hot side) then the resistance 
increases and power dissipation increases - enough to keep it hot.

I don't have actual numbers for the resistance of a bulb, the
current and the temperature dependence but one could measure them.
You would find that this is indeed a true bistable state.

	Jim