Halogen Lamp Limit Resistor?

[i200yrs]

Hello All again...we have a halogen lamp rated 12V, 100W (with fixed power supply input). Let say we want only to power up 50% (like 50W), is ok to add resistor in series? From ohms law that halogen lamp draw 8.3A. So if i want limit to 4.15A, how to calculate the resistor value? Any expert please help...thanks in advance.

 

[betwixt]

Use Ohms Law but beware of the ratings you need!

The resistor value depends on what you mean by 50%, half electrical power will not equal half brightness. The resistor value in Ohms is simply R=V/I so if you ignore the changes in lamp characteristics as the current drops and assume you want it to run from 6V instead of 12V, the value is (12-6)/4.15 = 1.44 Ohms. However, it will dissipate in heat Watts=V*I which makes 6*4.15=25W. In other words the half power you are trying to lose as light is lost as heat instead.

The biggest problem is that the filament resistance is highly dependent on its temperature, the resistance increases as it gets hotter so guessing what it might be at reduced power is difficult.

Usually, to control the brightness of high power, low voltage lamps we use a technique called PWM (Pulse Width Modulation) because it can accurately and easily control power fed into a load (your lamp) without getting anywhere near as hot as a resistor. The trick is not to reduce the current but to reduce how long it is available to the lamp. Imagine you wanted 50% of the light and you switched the lamp on for one second then off for one second repeatedly, the average light output would be 50% but the flicker would be irritating to say the least. Now imagine you speeded up the on/off cycling many times, the average would still be 50% but the thermal inertia of the filament would stop the flicker being visible. We often use speeds in the thousands to hundreds of thousands times per second to do the switching so there is no visible flicker at all. The beauty of PWM is that the switch itself is either fully on or fully off, in both states it loses no power. To vary the brightness the speed is kept the same but within each time period, the proportion of on to off time is altered. 100% off time means no light at all, 100% on time means full brightness.

For your application, and assuming you can use a DC supply, a simple timer IC and a MOSFET switch will suffice and give you variable brightness by turning a small control.

If your lamp runs from an AC supply, there is an alternative but less efficient method call 'phase control' using a device called a Triac. Basically, a Triac is a switch that you turn on electrically and it stays on until you remove the voltage across it. As AC alternates positive and negative it must pass through zero twice per cycle. The phase method is delay the switch on (trigger) to a time after zero. The available conduction through the Triac will continue until the next zero at which point the lack of voltage will turn it off again. If triggered early, more of the cycle passes through the switch, if triggered late, less of it passes through the switch so by adjusting the delay you control the output power. This is the method used in domestic and commercial light dimming controls.

Brian.

 

[i200yrs]

Hello betwixt, could you please share me link on how it looks like to cut in between the simple timer IC and a MOSFET switch? thanks

 

[betwixt]

Look at this link: https://www.electroschematics.com/9730/high-power-led-dimmer/

It is for high power LEDs but it will work just as well with a Halogen lamp. Note that it runs on DC, you have not told us what your power source is, if you are working directly from AC it has to be done differently.
The important characteristic of the MOSFET is its "Rds", the lower it is the better.

Brian.

 

[i200yrs]

Look at this link: https://www.electroschematics.com/9730/high-power-led-dimmer/

It is for high power LEDs but it will work just as well with a Halogen lamp. Note that it runs on DC, you have not told us what your power source is, if you are working directly from AC it has to be done differently.
The important characteristic of the MOSFET is its "Rds", the lower it is the better.

Brian.

My apology for not being clear on first post. Actually it runs with 12Vdc. Thanks

 

[BradtheRad]

As Brian says, it's more efficient to pulse the headlight.
Nevertheless it's interesting to figure out how to achieve the resistive drop.

To send 50W through your headlight, apply 8.5 volts average. (This converts to 71 percent duty cycle at 12V.)

Calculations per Ohm's law:

As your post states, since it draws 100W at 12V, current=8.3 A. ( W divided by V )

Lamp's running resistance=1.44 ohms. ( V / A )

Therefore 50W in 1.44 ohms results in 8.5V. ( SQRT(W/R))
Current draw =5.9 A. (V/R)

Inline resistor shall drop 3.5V (12-8.5)

It shall carry 5.9 A (same as lamp). Therefore value=0.6 ohms (V/A)

Watts dissipated (or watts wasted)=20.6W (V*A)
(Notice the inline resistor does not carry 50W.)

The same can be done with a transistor or mosfet. It has an advantage since you can adjust it to change the lamp brightness. However it needs heat-sinking to prevent it from frying.

So it's a better option to pulse the transistor/mosfet. You might get by without a heatsink, if you get it to turn fully Off and fully On.

 

[i200yrs]

Thanks both of you....can you recommend a no-heat_sink needed mosfet that can handle 12Vdc/100W halogen lamp? thanks

 

[BradtheRad]

Thanks both of you....can you recommend a no-heat_sink needed mosfet that can handle 12Vdc/100W halogen lamp? thanks

This article on halogen bulbs points out ways they are different from normal incandescents. (Halogen type has a cycle of filament metal being vaporized and redeposited.)

ww1.microchip.com/downloads/en/AppNotes/Driving Halogen Lamps.pdf

As the filament must generate the heat necessary to maintain the wall temperature of 250 ° C, it is important not to operate the lamp at any more than 10% (continuously) below its rated design voltage.

It suggests your desired brightness might be achieved at a duty cycle of 80 or 90 percent. As to whether this shortens the bulb's useful life, more research is advisable.

The article also refers to inrush current at start-up. This can be several times the normal running Amperes. A reasonable rating for your mosfet is 2 or 3 times expected maximum Amperes.

With no heat sink, reasonable heat dissipation is probably 3 Watts for the mosfet (even though it burns our fingers).
The lower its spec On-resistance, the better.
To stay under 3 W, you need to drive it to a resistance of 1/20 ohm.
By putting 8A through it, Ohm's law results in 0.4 V.
Then V*A = 3.2 W. Less than that running at 80 or 90% duty cycle.

There is a formula for calculating temperature rise of a device based on Watts dissipated.

If you cannot drive it to a low enough On-resistance, you may need to consider paralleling two identical mosfets.

 

[schmitt trigger]

What Brad says is correct!

I once made the mistake of running a halogen lamp thru a dimmer, and shortly afterward the bulb started to darken.
Don’t know the exact failure mechanism, but believe that the glass envelope does not get hot enough and thus filament material condensed there.

 

[c_mitra]

Any expert please help...thanks in advance.

Halogen lamps are not designed to be run at lower power. Lower power means lower filament temp and lower filament temp means poor iodine recycling (WI3) and that will cause the glass (quartz) cover to darken and the filament will fail.

As others have correctly mentioned, 50% power does not mean the light output will be 1/2. You should get a 12V 50W lamp!

You can run halogen lamps are 90% and that will extend the life significantly. There will be only a small drop in the light output but the filament life will be increased (approx doubled).

Anyway halogen lamps have short lives; but the key is the halogen that recycles the deposited W from the walls back to the filament and reduces chances of filament hotspots.

At 40-50% power, the lamp will be rather dim but the whole purpose of using a W-I lamp will be lost.

 

[i200yrs]

Halogen lamps are not designed to be run at lower power. Lower power means lower filament temp and lower filament temp means poor iodine recycling (WI3) and that will cause the glass (quartz) cover to darken and the filament will fail.

As others have correctly mentioned, 50% power does not mean the light output will be 1/2. You should get a 12V 50W lamp!

You can run halogen lamps are 90% and that will extend the life significantly. There will be only a small drop in the light output but the filament life will be increased (approx doubled).

Anyway halogen lamps have short lives; but the key is the halogen that recycles the deposited W from the walls back to the filament and reduces chances of filament hotspots.

At 40-50% power, the lamp will be rather dim but the whole purpose of using a W-I lamp will be lost.

Thanks, so it is advisable to use 90%...like my halogen lamp is 100W, I should powered it up by 90% (or 90W)? Then can I add resistor or is there any halogen driver circuit you can recommend? thanks

- - - Updated - - -

OK guys, as the calculation method shared by Brad. I will try to reduce the power to halogen lamp (12Vdc, 100W) by 80%. To make it fixed value, i will use resistor method. Based from the calculation, I will series a 0.17 ohms resistor with 10W rating. Please share your opinions before I finalize the modification....By the way, we are using this halogen lamp for microscope...thanks

 

[betwixt]

That's very bright (and hot) for a microscope. I use a low power LED. Why do you need it to be so bright?

Brian.

 

[i200yrs]

That's very bright (and hot) for a microscope. I use a low power LED. Why do you need it to be so bright?

Brian.

Hello Brian...this actually the original specs from the machine.

 

[FvM]

Halogen lamps are not designed to be run at lower power. Lower power means lower filament temp and lower filament temp means poor iodine recycling (WI3) and that will cause the glass (quartz) cover to darken and the filament will fail.

That's what I learnt as a student, but we see many halogen lamps with dimmer, in domestic as well as stage lighting. They seem to achieve a fair lifetime though.

I would use a PWM dimmer with soft start for the application.

 

[BradtheRad]

series a 0.17 ohms resistor with 10W rating.

Although 10W is calculated dissipation, nevertheless it's wise to make your resistor rated for higher than that. Consider putting several one ohm resistors in parallel. It's easy to add or subtract, until the bulb is the desired brightness.
5 yield 0.2 ohm.
6 yield 0.17 ohm.
7 yield 0.14 ohm. Etc.

Or, sufficient length of wire makes a resistor.
8 Amperes causes 0.17 ohms to dissipate 10W so it gets hotter than a 7W bulb.
Example, 25 feet of 18 gauge AWG copper.
40 feet of 16 gauge.
70 feet of 14 gauge. Easy to add or subtract.

To do the PWM method, a 555 timer IC can pulse the mosfet. Adjust duty cycle by changing volt level at pin 5.



Pulses become longer as the pot is moved from low to high.

 

[c_mitra]

Or, sufficient length of wire makes a resistor.
8 Amperes causes 0.17 ohms to dissipate 10W so it gets hotter than a 7W bulb.
Example, 25 feet of 18 gauge AWG copper.
40 feet of 16 gauge.
70 feet of 14 gauge. Easy to add or subtract.

Nichrome or Kanthal wires can dissipate greater heat for smaller length. Also, they have higher resistivity and hence will need shorter length. Plus, they have a smaller temp coefficient.

Also, they can be run bare. But using 1W resistors in series parallel combinations is also very simple. Both Nichrome and Kanthal wires are difficult (if not impossible) to solder- they must be clamped with screws.

For a microscope you will need adjustable brightness so that image brightness can be changed with magnification. Higher magnifications will need higher light intensity at the sample plane (slide).

Ideally, the light intensity should be controlled using ND filters (have similar color /balance or distortion).

 

[i200yrs]

Thanks for all the knowledge sharing guys, ok i will try using resistor method first

 

[c_mitra]

Thanks for all the knowledge sharing guys, ok i will try using resistor method first

If you are using the resistors in parallel option, try to match them by hand (so that their resistance values are as close as possible).

That will ensure best possible current sharing...:-D

 

[i200yrs]

If you are using the resistors in parallel option, try to match them by hand (so that their resistance values are as close as possible).

That will ensure best possible current sharing...:-D

Will connect the resistor in series with halogen

 

[c_mitra]

Will connect the resistor in series with halogen ...

No, I did not mean that!

You mentioned a 0.17 Ohm resistor in series with the lamp;

But you many not get that chosen value easily in the market;

You may have to improvise by using series - parallel combination of resistors- say 1W 1R- they are easily available.

If you put five of them in parallel, you can get 0.2 easily. But match them if possible.

If you put 10 of them in parallel, you get 0.1R 10W resistor. But if they are not matched, there will be thermal runaway.