Reduction of DC Voltage To Lite LED

Unfortunately, you're stuck with high power losses if you try to use resistors or a linear voltage regulator. Power = volts X amps, so power = 118 X 0.02 = 2.36 watts.

The only altenative you have is to use pulse width modulation and a fairly hefty capacitor and resistor, and it wouldn't hurt to put a zener diode in the circuit, for minimum components. Your power losses will be reduced, but the bulk of the circuit may preclude use of it.

Dear Bill,

Thank you for your valuable suggestion for linear voltage regulator.

There is a space restriction to keep the regulator inside the Gauge which further to be fitted on Electrical Locomotive. But further I will try out for placing the regulator at appropriate place since 110 Volts DC required to be tapped from the main bus bar of the Loco and place restriction is big concern while placing Gauge.

But still I will try on your suggestion. Hopes I will also able to suggest something which you may required and I will be happy to serve you.

Thank you once more.

Warm regards,

WORKFLOW

I would purchase a small DC-DC converter (> 80% efficient) to convert your 110V to either 5V or 3.3V, and go from there. With a 3.3V supply, use about 100Ω, 1/8W (1.2V @ 20mA) for each LED. Why burn up all that power in resistors?

Dear bhankii,

Thank you for your valuable suggestion for DC-DC Convertor.

There is a space restriction to keep the Convertor inside the Gauge which further to be fitted on Electrical Locomotive. But further I will try out for placing the regulator at appropriate place since 110 Volts DC required to be tapped from the main bus bar of the Loco and place restriction is big concern while placing Gauge.

But still I will try on your suggestion. Hopes I will also able to suggest something which you may required and I will be happy to serve you.

Thank you once more.

Warm regards,

WORKFLOW

You should use individual resistors for each LED. There are slight variations in junction threshold voltage that will cause unequal distribution of current if all LED's are in parallel. Also, you distribute the power to be dissipated among many resistors instead of one and that will also help with reliability should one resistor fail.

Dear rcapper,

Thank you for your idea.

I am using individual resistor for each LED and nearly tried out all ideas you suggested but it did not work. But still I should appreciate for concerned suggestios. Hopes we should extend such suggestion to get the platform to become more stronger among the community.

Warm regards,

WORKFLOW

If you are averse to WASTE (of energy) then you should use an Inverter with input V_dc-=110 and Output V_dc=2.

You can modify a standard smallest Spot Welder IGBT Inverter- to above parameters.

Dear MUKUL,

Thank you for your response. Your suggestion is also very good and very positive. But frankly admitting that I am not aware of standard smallest Spot Welder IGBT Inverter.

Kindly tell me something more about this.

Warm regards,

WORKFLOW

Workflow,

Bad Idea!

It defeats most of the purposes of using LEDs to use them in that way (consuming 50 X the wattage of the LED in the resistor). Additionally, the 2 Watt rating is the absolute maximum wattage, not what you want to run it at, let alone exceeding it like you are. For continuous duty in a closed housing, I would use a safety factor of at least 100%. Meanwhile 12 of these "heaters" will be dumping about 26 watts into your small enclosure! That's ridiculous. Plus, you will likely encounter resistor failures, long before the LEDs burn out, although when you run resistors too hot, you speed up their natural tendency to change resistance value over time (least change with the wire wound type) and that could potentially cause premature LED failure or dimming.

There is nothing wrong with wiring the LEDs in series if you run them at a nice safe voltage, and they are high quality ... they will last a very long time, and if one should fail, just replace the whole ring of them, but that would be rare.

That way if you stick with the resistor approach because of space constraints or cost is a factor and efficiency is not, at least you will only be dropping between 86-100 Volts once, instead of 108 Volts between 5 and 12 times. Then you will have the room to use several resistors in series and/or parallel to spread the heat dissipation. Whatever the wattage calculates out to for each value, double it at least. As a result you will never be dissipating more than 2 Watts of heat, spread out over 4-6 (or more) Watts of total resistor ratings.

That would be the way to go, considering your space constraints. "Wasting" 2 Watts maximum is acceptable for driving 5 or more LEDs since the whole lighting arrangement is only consuming 2.2 Watts whether you have 5 LEDs or 12.

Regards, Greg

Dear Guru GregG,

Many many thanks for valuable guidance. Realy your name is the indicator and plays vidal role to bond CR$ Community. If I will be able to do a little bit service as you are rendering to the Community I will be lucky enough to do so. So will try to follow your dedication and service.

I have not tried out LEDs in series connection but I will work out on this suggestion as it do not interfere with the basic concept of space, utility, cost factors, and easily implementable. I will try out various combinations and come to final conclusion. And if it fails then I will try out the ideas of MUKULMAHANT, Bill, and rcapper.

But hopes the problem will be sorted out as per your suggestion.

Hopes we should extend such suggestion to get the platform to become more stronger among the community.

Warm regards,

WORKFLOW

Workflow, if you put your LEDs in series and want to protect the whole chain against failure of one single LED then put a Zener-diode with lightly higher voltage than the LED in parallel. If the LED fails, the Zener diode overtakes the chain current. Another advantage is that every LED is protected by the ZD against reverse voltage damage.

I`ve done this very often with Pt-100-Sensors in multi-sensor-matrix-fields and it works fine.
Regards Uwe

Here is an example of a switching supply chip that would run off the power line and produce low voltage power with minimal losses. It may not be the best one for your application but I offer it as an example to point you in a direction that you may discover a suitable solution. TI and others make many versions of similar parts. http://focus.ti.com/docs/prod/folders/print/ucc2888.html

Why dont you try a capacitor of about 104 /(600v)to reduce the voltage to suit for led? you must get 110ac easily. or from front of the dc. its small size even if you serial with a resistor.