VU Meter LED backlight voltage help...

[tomzarbo]

I'm in the process of working out the details for a retro-style radio for a friend of mine and I decided to go with old analog VU meters for the light show.
I managed to get the needles working despite the separate driver board I purchased having no documentation at all.

But, the board has two additional leads per side which are supposedly for voltage to drive the LED's built into the VU meter. The (contradicting) documentation on the VU meter states that the LED's want 1.9 to 3.3 volts, Using my adjustable power supply it looks better to me with around 2.0 volts or so.

Anyway, the LED's (there are 3 per side) are embedded into the VU meter and there are no specs available for them. I just need to figure out how to get the voltage down to 2.0 volts somehow, with a resistor I presume. Since I don't know what value, type, etc. the LED's are, I'm having a hard time figuring out how to hook them up.

I have two 9.5 volt outputs on the driver board to power the two LED banks on the VU meter itself. So I need to get that 9.5 volts down to 2 volts for the backlight LED's...
I figure I'm overthinking this, but is there a calculator or some way to easily figure out how to do this with resistors? I'm drawing a blank. Guessing with various values of resistors doesn't work because testing the voltage with a Multimeter doesn't show the reduced voltage until a load is hooked up to it.



I need to be careful of the LED's because they are embedded and non-replaceable and the VU meter was kind of expensive.
Thanks for any advice, I'll keep tinkering!

TomZ

[tomzarbo]

I found an online resistor voltage dropping calculator... yeah, I had to experiment a bit, but I came up with this for 9.5 volts...



A 470 ohm and 100 ohm 1/4 watt resistor in series for 570 ohms brings 9.5 volts down to about 2 volts even.

The two little resistors heat up to around 115 degrees which I don't THINK is too hot? (I used a digital food thermometer with a pointy metal probe and it took a few minutes to get up to almost 115 degrees) It feels hot when I touch my finger to it, and 3-4 seconds is long enough to feel uncomfortable, but I'm not sure if this is acceptable or not.

I've had it powered up for 20 minutes or so and nothing bad has happened, yet.

If I purchased some 1 watt resistors of the same value to do basically the same thing, would that extra wattage cushion mean that the resistors would not get as hot and therefore be a little safer?

TomZ

[Steve Lee]

1.5 VDC is the norm for these LEDS and I think you need some higher wattage resistors to be safe Tom.

[tomzarbo]

Hmm, I probably agree with you on the higher wattage resistors, but the brightness at less than 2 volts wasn't adequate.

TomZ

[Steve Lee]

Turn the overhead lights off - the VU meters illumination were never meant to be more than back light, sir.

Remember the rugburns of your youth with the receiver playing in the background?

Yeah, like that.

[tomzarbo]

Turn the overhead lights off - the VU meters illumination were never meant to be more than back light, sir.

Remember the rugburns of your youth with the receiver playing in the background?

Yeah, like that.

Okay, I gotcha. Makes sense.

I did turn the lights off and in a dark room, they don't need to be as bright as I thought. And I only have one side operational during testing also, they'll be more output with both L and R sides in action.

One of my concerns is that I plan on having this in close proximity to a single DIN car stereo which I will use as the head unit of the radio, and that has a 4.1" screen as well as some backlights for the controls as well, so I didn't want the VU meters to get lost in the haze. I plan to angle them up though, so that may help.

I still have to figure out which resistors will get this to work. Reducing the voltage on the power supply from 9.5 to 5 or 6 volts still gave me around 2 volts after the resistors, but it did dim the LEDs considerably. Maybe I'll get an assortment of PE's 1 watt resistors to play with.

These projects can get complicated and I'm trying to get all the pieces playing nicely with each other before I start building it.

Here is a quick mock-up of the basic idea. Boom boxes have always had a lot of 'bezels' and protrusions and extra doo-daddery on them, so I'm trying to do similar with this one to a lesser extent. Not my normal type of project, but this one's a custom special order so to speak. I plan on making grills for the drivers and PR's on the sides as well. Lots of jigs and special stuff with this one....




In order to keep things manageable, I plan on having the speakers be if not removable, at least separate in construction so I can figure out how to get the bits for the radio arranged in some type of reasonable order. I have voltage reducers, drivers, timer relays, possibly an internal power supply, etc... I'd like to get things all figured out and actually wired up in advance before even making the thing so I can be assured it will play well together.

Trying to get this one done by Christmas. Wish me luck!

TomZ

[Steve Lee]

Sounds like fun - Good Luck!!

[johnny5jz]

I miss the old boom boxes, but definitely do not miss the sound. This looks like a really cool project!!

[randyohoh]

White, Blue, & 525 nM "Traffic" green LEDs normally require 3 volts to get them to the rated light output. Red, orange, yellow, and the yellow-green "Grellow" 565 nM LEDs normally require 2 volts to get them to the rated light output. Running the LEDs hotter will increase the light output and decrease the life, and that can be dramatic or even instantly if substantially overdriven. Most T1 (3mm), T1 3/4 (5mm), and small (0603, 0805, & 1206) surface mount (smt) or (smd) LEDS run at 20mA. If LEDS are wired in series, their voltage ratings must be added, and in parallel, the currents added. I have an old version xl spreadsheet to calculate resistors which I was not allowed to attach.

[tomzarbo]

White, Blue, & 525 nM "Traffic" green LEDs normally require 3 volts to get them to the rated light output. Red, orange, yellow, and the yellow-green "Grellow" 565 nM LEDs normally require 2 volts to get them to the rated light output. Running the LEDs hotter will increase the light output and decrease the life, and that can be dramatic or even instantly if substantially overdriven. Most T1 (3mm), T1 3/4 (5mm), and small (0603, 0805, & 1206) surface mount (smt) or (smd) LEDS run at 20mA. If LEDS are wired in series, their voltage ratings must be added, and in parallel, the currents added. I have an old version xl spreadsheet to calculate resistors which I was not allowed to attach.

Thank you randy, that sheds some light on my situation. I may be able to guesstimate a better solution with your help. Thanks!

TomZ

[AEIOU]

It's times like this that an adjustable DC power supply comes in handy.

[unclejunebug]

You may also consider using a voltage regulator instead as their outputs are often adjustable. You may even find one with a fixed voltage output value that matches what you need. If there are any additional resistors/caps needed they would be clearly documented in the data sheet.

[neildavis]

You need to focus on the current through the LED, not the voltage. Look up "diode equation" on Wikipedia and you will see that the voltage across a diode, including an LED, is a function of the current. The curves are different for different types (and colors) of diodes. You need to keep the current within the rating of the LED, which is usually on the order of 10 to 40ma. At these currents, the voltage drop for a red LED is typically around 1.8V, and white LED's are 3-3.5V.

Calculating the required resistor is easy. Suppose you want to operate the LED at 20ma and your source voltage is 9.5V. At this current, the voltage across a white LED will be about 3V, so the voltage across the resistor will be 9.5V - 3V, or 6.5V. To get 20ma, just use Ohm's Law, where V=RI, or R = 6.3/.02, or R = 315 ohms. A 330ohm resistor will work fine. Since there is about 20ma going through the resistor and the power dissipation is I^2R, the resistor should be at least 0.132W. So, you should use at least a 1/4W resistor.

You can adjust the brightness by using more or less current. But don't worry too much about the voltage--it will vary with the current but usually you can assume the voltage to be constant in the useful operating region. Note that if you can find 5V somewhere, the power dissipated by the resistor will drop quite a bit. For a 5V supply and 20ma through the diode, you would need a 100-ohm resistor, and it would only dissipate .04W.

[tomzarbo]

Thanks guys for the good information.

Neil, I'll read this again when I get home. A 1/2 hour lunch break isn't enough for this to sink in.

I have a decent assortment of 1 watt resistors coming from PE -- they handily come in packs of 10 -- between that and the ones I already have, I should be able to get this working soon.

I'm really looking forward to getting the size and space requirements for the electronics part of this figured out. Then I can finally start to build it out.

Thanks again for the guidance.

TomZ

[neildavis]

Whoops...I didn't see your voltage measurement of 2V, which is a typical value of a yellow LED at 20ma.

At 9.5V, there will be about 7.5V across the resistor, which means you have 7.5/570, or 13ma flowing through the LED. You could probably drop down to a 390-ohm resistor if you want the display brighter (that would result in 20ma through the LED), or use a larger value than 570-ohm if you want it dimmer. With 13ma flowing through those 2 resistors in series, the 470-ohm resistor will dissipate (.013^2)*470 watts, which is .08W. So, the 1/4 resistor size is fine. A larger resistor (1 watt) will get just as hot since it will dissipate the same power, but it might feel cooler due to the larger surface area.

[randyohoh]

Here's a picture of the xl spreadsheet that I had previously mentioned. I AM NOT an electrical engineer, but as a mechanical worked with indicator LEDs for several years, had an EE help me.