properly decreasing voltages

[cannikin]

I have a general question about decreasing voltage.

I have a sifam meter with a Lamp rated 12V 1/2W
The VU lamp supply coming off my PCB is 38V. So I have to decrease the Voltage and get it down to 12V MAX.

the VU lamp positive side I have 150 ohm 1/2Watt resistor
On the negative side I have another 150 ohm & 100 ohm 1/2Watt resistor.

Which decreases the Voltage to 7.5Vdc. :thumb: my intention for decreasing it to 7.5V is to extend the life of the lamp.

But the resistors Are getting pretty hot to touch...

So this leads me to believe there's a better/more efficient way to achieve the same results.

ANy pointers would greatly be appreciated. :roll:

 

[CJ]

Your bulb filament is not getting hot enough. Tungsten filaments change their resistance from a piece of coat hanger at power up, to many ohms when hot. Not very linear. This is why your bulbs at home always pop during power up. Have you ever seen a house light just go out after it was on? The initial inrush usually gets the bulb before it actually burns out.

Since your bulb is cold, it is drawing more current than you calculated and thus, stressing the resistors.

If you run your bulb at 10 percent less voltage, you will increase it's life by at least 50 percent. Automobile lights run at very inefficient voltages to keep them lit longer. I just replaced my low beams after 240,000 miles. They still lit, but were getting a bit dim. Either that or I had to many bugs from the summer time on the lens.

 

[bcarso]

Of course CJ is right about the large temperature coefficient of tungsten lamps.

In terms of your general question, anything you use that conducts continuously, like a resistor or a transistor with a constant voltage drop across it, will be turning that power of the voltage drop times the current into heat.

However, if you use switching techniques and supply your resistive load with pulses of a certain duty factor you can get quite efficient. When you do this with a bunch of parts including reactive components for short-term energy storage it becomes a switching converter.

I wouldn't advise bothering with any of this to dim a lighbulb, especially in an audio environment. Besides the complexity, the switching process tends to generate conducted and radiated noise.

 

[cannikin]

Thanks for your input.. yet another learning session:thumb:

[quote author="bcarso"]I wouldn't advise bothering with any of this to dim a lighbulb, especially in an audio environment. Besides the complexity, the switching process tends to generate conducted and radiated noise.[/quote]

One thing .... I'm actually not doing this just to dim the light. I need to get 30V down to the 12V which is the rating of the lamp.

 

[Svart]

12v zener and a resistor. aka zener shunt regulator. good stuff

of course we'll need to know the current draw of the light at 12v to properly do this..

 

[pstamler]

Well, let's do some sums. You're reducing the 38V source voltage to 7.5V, so you're dropping 30.5V in the resistor string. The total resistance of the string is 150 + 150 + 100 = 400 ohms. So the current being drawn is 30.5V/400R = .07625A.

How much power does each resistor dissipate? P = I^2 * R; we already know the current is .07625A, so each 150R resistor dissipates (click, click) .872W, and the 100R resistor dissipates .581W. Half watt resistors are too small and will eventually burn out. Try changing the resistors to 2W units, and you should be fine. They'll get equally warm, but they're built to take it.

Or use two 100R 2W resistors, in series, in each leg.

Peace,
Paul

 

[CJ]

LM317 in the 220 package with min pot/knob on the voltage adjust would be cool. You could adjust the pot for bulb brightness for those special moments in the studio, like when the cops show up, you kill the lights and you need a dim night light so you don't trip over the cables while you are sneaking off to lock the door.

 

[Svart]

i see that CJ was over at my studio the other night.... :green:

 

[PRR]

> the resistors Are getting pretty hot to touch... So this leads me to believe there's a better/more efficient way to achieve the same results.

Any linear technique with the same result will have to waste the same heat. No Way Around It. Doesn't matter if you do it with resistors, linear regulators, or a barrel of salt-water.

You have 38V. You want (rounded) 8V. 8/38=20% is doing useful work; 30/38= 80% is just dead waste.

Oh, you could try to source 24V lamps that fit the same place. De-rated to 16V is only 58% waste. Or a 48V (rare) or two 24V in series: hardly any waste at all.

Or find a 6V transformer just for the lamps. At these light loads it will probably deliver a bit over 7V, just what you want, no huge waste required.

The alternative, time-chopping the supply down, is extra work and usually mucks-up audio unless you spend more time than the silly bulbs are worth.

Feeding 2V LEDs from 38V is also wasteful, but LEDs are so much more efficient that the total waste heat will be similar. Life-span of an LED is much better than a cheap bulb, even with stiff derating.

0.872 Watts in each 150 ohm 1/2 Watt resistor sure is WAY too much. One of my first lessons in power rating was a dead 1-Watt cathode resistor that had been dissipating just about 1 Watt, in a hot corner under a 50C5 output tube. Died in a couple years. We replaced it with 2 Watt: cost a nickle more than a 1 Watt, but lasted forever. If you don't have a Corporate Accountant whipping you for every penny in your design, never run resistors past half their rating.

One 390 ohm 5-Watt or two 220 ohm 2-watt (or "two 100R 2W resistors... in each leg") seem like plausable practical parts.

> They'll get equally warm

The surface temperature will be lower, 0.8W in a 2W versus 0.8W in a 1/2W. Of course the power is the same, and the heat coming off it is the same, but (for same material) the 2W will be much bigger than the 1/2W so the surface temperature is lower, the paint/varnish stays intact, your finger suffers a first-degree burn (pain and redness) instead of a second-degree burn (blister). Also most good wire insulation can lay on a half-power resistor "OK"; a resistor running 1.5 times rating will melt about anything except Teflon.