Halogen / Incandescent bulb internal resistance

[made2hack]

Hello all,

I've got a question relating to Halogen / Incandescent bulbs. Do they have a certain internal resistance that needs to be overcome before they turn on? For example, take a Philips H4 car light bulb (60/55W @ 12V). I assume 55W is the low beam power. So it will pull a bit over 4.5A.

When I try to give it 12V @ 100mA, nothing happens. Not even the smallest hint of light. If I give it a lower voltage (say 5V) at whatever current it wants, it turns on. It just seems to not like small current.

How could I conceivably predict the lowest current at which it will turn on without guessing endlessly? I don't have a supply where I can just dial in the current.

[amyk]

The filament resistance varies with temperature, when it's cold it will be much lower than when it's hot. The rated power and voltage suggests it will be around 2.6 ohms when hot, and a fraction of that when cold.
 
Thus that "12V at 100mA" is really going to be far less than 12V if the current is being limited to 100mA, and so the power output will be far below ~1W. There is no real "turn on" point for incandescent bulbs - the filament just emits mostly infrared before it gets hot enough to be visible.

[Ian.M]

Tungsten has a high positive coefficent of resistance between room temperature and the filament's running temperature.  Typically the cold resistance will only be 10% of the hot resistance (which can be calculated from the rated voltage and wattage).   Therefore if you attempt to drive it from a limited current supply that is a small fraction of its usual operating current, it will dissipate far far less power than you expect. 

As an experiment, take a PSU with an adjustable current limit, and a bulb of a voltage and wattage it can supply.  Set the output voltage to the bulb's nominal voltage, and the current to 0.  Slowly turn up the current while monitoring the voltage across and the current through the bulb.  You will find that there is a point at which the voltage starts rising far faster than you would expect from the rate you are increasing the current and it may even rise on its own due to the resistance increasing and the resulting proportionate filament power increase resulting in a runaway temperature increase.  The filament wont visibly glow below that point.

[Delta]

You can't "give it 12v at 100mA".  It don't work like that!

[made2hack]

You can't "give it 12v at 100mA".  It don't work like that!

What I meant I was supplying it with voltage through an LM317 set to current limit @ ~ 170mA.

[Delta]

You can't "give it 12v at 100mA".  It don't work like that!

What I meant I was supplying it with voltage through an LM317 set to current limit @ ~ 170mA.

Right, so you were supplying it with 170mA.  With a constant current supply, the voltage will be _whatever it needs to be to drive the set current through the load_

[Zero999]

If you measure the actual voltage across the bulb with only 170mA flowing though it, you'll find it's something on the order of 50mV.

[alsetalokin4017]

Sounds like time to do a meditation on Ohm's Law.....   "ohmmmmmmmmmmmm......."   

[alsetalokin4017]

You can't "give it 12v at 100mA".  It don't work like that!

Well.... you sort of could, using PWM, and considering the 100mA as "average" current.  You might see a very dim glow on the filament.

For example, the simple 555 PWM circuit that has been discussed recently in other threads might be an interesting one for the OP to experiment with.

[Seekonk]

I hate any circuit with a 555, there are just so many better chips.  Here is one that will give you independent frequency and duty cycle with a SG3525. 
http://www.ebay.com/itm/262136732576?_trksid=p2055119.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT

[Delta]

I hate any circuit with a 555

BAN HIM FOR LIFE!   

[Zero999]

You can't "give it 12v at 100mA".  It don't work like that!

Well.... you sort of could, using PWM, and considering the 100mA as "average" current.  You might see a very dim glow on the filament.

For example, the simple 555 PWM circuit that has been discussed recently in other threads might be an interesting one for the OP to experiment with.

Then the average voltage would be below 12V so you could hardly say it's running off 12V.

PWM is a perfectly valid way to dim incandescent lamps but the peak current will be much higher, especially on the dimmest setting,  to the cold filament having a lower resistance.