Looking to make LED strips pulse to bass hits (in my car)

[Pittcrew]

So basically,

I have 4 LED strips (each rated for 12v) that I’d like to make pulse to the bass of the music, such as in this video . There are a lot of misinformed people on YouTube who just hook up their LEDs directly in parallel with the subwoofers, which doesn’t work well at all and will blow the LEDs, so YouTube hasn’t been much of a help to me.

I’m just a first year EE student, not a full on EE like most of you guys probably are, so I was wondering if anybody would be able to help me design a circuit to make this work. A threw this together, it is by NO means even remotely correct, just a little visualization of what I’m trying to do here. I think I’m probably going to have to end up doing something with a transistor for the switch, and a capacitor to make the LEDs fade after the bass hits, but I’m not really sure.

Some things to note:

Each diode in that circuit represents an LED strip rated for 12v DC
My amp is 3000 W peak power
There are two subs wired in parallel with a total impedance of 4 Ohms.
Any and all help/suggestions is much appreciated!

Thanks

 

[omega_minus]

I've not made a circuit like this but you've got a start. One bit of advice, don't parallel diodes without a resistor in series with each one, regardless of how small. This is because diodes have negative thermal resistive coefficients, which means the hotter they get, the better they conduct, unlike resistors. You can picture what happens in a parallel configuration: slightly more current in one of the diodes heats it up a little. This makes it less resistive, which means more current. This leads to thermal runaway, and as you noted will blow the diodes. Some small ohmic resistance before each diode will bring this situation under control. You can also put some small dc voltage across each diode to get them close to turn on that way you won't need to have a signal over 0.7V to get turn on. Good for you designing this as a first year too!

 

[Pittcrew]

Thanks for the response! So something more like this?

 

[omega_minus]

Right. You can figure out the voltage drop across the transistor and then get the best values for the resistors so that the current allowed in the diodes is not too much. The 12 volt source minus the transistor voltage minus the roughly 0.7 volt drop on the diode gives the resistor voltage (minus any capacitor effects). Use ohms law solving for the resistor values since the manufacturer of the diodes will give the voltage drop and max current. To get an idea of what the added dc voltage across the diode would look like (to help with the 0.7V turn on) would look like, google amplifier circuit with crossover distortion compensation. I only have access to an ipad so i can't draw it, but you can use a voltage divider tapping off your battery to put about 0.5V across them (almost turning them on) so that any little signal will get them lit up somewhat. You might find this is not needed, so try without it first. If it seems unresponsive, try my suggestion. Where are you going to school?

 

[Pittcrew]

I'm currently on summer break, but I just finished my first year at Kansas State University. So far, I've only had one EE class, so my knowledge isn't too extensive.

The other thing I'm wondering, so the thing switching the transistor is going to be the speaker outputs on my amp. Is there a way to get it so it works at low volumes (minimal power from the amp) while still working at large volumes and not frying anything? Also, what would be the best way to get it to draw the least amount of power from the amp (as to not affect the subwoofers) while still working at the largest range of volume possible?

 

[omega_minus]

If i were doing it i'd use a MOSFET as the transistor. They don't draw current through the gate, the gate is a capacitor. A BJT has a diode as a gate and thus draws some current, which uses considerably more power. Also, as such, the gate resistor would not be required, as no real current will flow through it. Also without really analyzing the circuit there might be additional issues or improvements but I don't see why it won't work at a glance. In this configuration, only the battery will give power to the led's and the power to the speakers will not be affected provided the amp has a good enough regulator.

 

[Pittcrew]

If i were doing it i'd use a MOSFET as the transistor. They don't draw current through the gate, the gate is a capacitor. A BJT has a diode as a gate and thus draws some current, which uses considerably more power. Also, as such, the gate resistor would not be required, as no real current will flow through it. Also without really analyzing the circuit there might be additional issues or improvements but I don't see why it won't work at a glance. In this configuration, only the battery will give power to the led's and the power to the speakers will not be affected provided the amp has a good enough regulator.

Ok thanks a lot! You've helped me out a ton so far. So no resistor between the amp and mosfet. Here's the part I'm going to need some help for if you don't mind. What kind of mosfet should I get and where can I get it at? I found this website http://www.westfloridacomponents.com/MosFETS.html but I have no idea which one would work for this. And what do you think I should get in terms of size for the cap?

 

[omega_minus]

I would get an n channel power mosfet, the ones with the 3 legs inline and rectangular body. Many kinds will work and since they're really cheap, you could get a few of each of a couple types, in case you burn one up or a certain type works better than another. These have a big metal tab on them which have a screw hole for attaching to a heat sink. depending on the current draw it might be a good idea. Getting them together will save on shipping. To be honest the cap might not be needed. The transistor has dc forward blocking capability and since the only current passing through it goes to the diodes (we're not trying to filter out dc to get an ac signal riding on it) you could probably leave it out.

 

[Pittcrew]

I was planning on having the cap there to create an effect where after the bass hits the lights dim out instead if just turning off. Would that be the correct place in the circuit to put the cap?

 

[omega_minus]

Looking at your first post i see you wanted the cap for fading. In your current setup this won't happen. Remember caps are like springs. The current goes back out the way it came in, just like a spring pushes back from the direction it was pushed. So the current will tend to go back toward the transistor (which it can't get through). Instead you'd want the cap (and maybe a resistor) in parallel to the diodes to ground. A parallel cap will take some of the current going to the diodes when the voltage is high, and feed it back as the voltage drops. This will also probably make them fade in a bit too. The RC time constant of this network will determine how reactive the diodes are.

 

[Pittcrew]

So putting an RC in parallel with the LEDs will act as a low pass filter and create a fading effect?

 

[omega_minus]

It should, in principle. The way it is in the original drawing, it would eventually charge up to around 12V and then block anymore current. A series inductor (or better, the parallel cap) will give the fading effect. You will no doubt have to play with the component values to get it how you want it, but as long as you don't over volt or over current your led array, it should be fine.

 

[Bobbywhy]

Observation: LEDs and LED "strips" operate on Direct Current (DC). The power source in the above circuit is the 12 Volt battery. But the capacitor, C1, BLOCKS DC! The LED strips may emit some light for an instant during the capacitor's initial charging, but as soon as the capacitor is charged up current will stop flowing and they will go dark. How would C1 discharge during the opposite polarity of the audio signal? I suggest you re-examine your basic circuit plan.

Edit: I reserve the right to be totally wrong because after seven decades my brain may malfunction occasionally.

 

[Pittcrew]

Ok, so I've revised the circuit, let me know what you guys think

 

[Bobbywhy]

Pittcrew,

Now, that new circuit looks like it might even work! The "CircuitLab" edition you've used makes a very nice schematic drawing. Does it actually simulate the operation of the circuit to verify its function?

Edit: Applying the audio signal to the gate of the FET would turn the FET (current) and the LEDS on when positive and off when negative. You've asked for your LEDs to respond to the bass "beat", I think, and NOT the sine wave out of the audio amplifier.

Why are you trying to "recreate the wheel"? Why not just find some circuit designs that drive LEDs in synch with the bass beats that are already completed and functional? Then perhaps modify a working design to suit your specific requirements.

 

[omega_minus]

Bobbywhy has a point but I assumed you were like me and enjoyed reinventing the wheel for educational purposes (and maybe bragging rights). One observation. Your rc circuit is short circuited. The current leaving the diodes is at ground potential, it won't have any reason to then go through the rc and back to ground. You can always tell if something is shorted by seeing if both ends of the network occupy the same node. This means there cannot be any potential across it and hence no current. Instead put one end of it after the FET and the other end at ground, so the current will have the 'option' to go through the diode leg or the cap leg. The way to tell if two things are in parallel is to see if they share the same node on one side and share another (different) node on the other side. This means they will always have the same voltage across them. Series configuration means that current leaving one element has no other path except through the next element, like your resistor/diode pair. They are in series. Also realize that elements may be in neither series nor parallel configurations, those are just specific cases. Good work so far!

 

[Pittcrew]

 

[omega_minus]

That looks good, build it and see what you get. Keep me posted I'm interested in how it comes out! Again, good work!

 

[Pittcrew]

Thanks! Yeah I just got back from radio shack, unfortunately they didn't have everything so I'm going to have to order the rest online. But once I get it done ill definitely post a video/pics of the build and it in action

 

[Bobbywhy]

From a different thread here in the Electronics Forum, “carlgrace” posted this:

“Download LTSPICE from Linear Technolgy (linear.com) and simulate it. You should always simulate something before you build it, anyway.”

https://www.physicsforums.com/showthread.php?t=695226

 

[sophiecentaur]

It could perform better if you were to use a level measuring circuit rather than just feeding the LEDs with the audio signal. If you were to use a rectifier / detector circuit to determine level and use that level (comparator circuit) to switch the diodes on and off (perhaps with a bit of 'hold', so they go on and stay on at a 'syllabic rate' - rather than 'flickering' at the low frequency audio rate. It may be that the Sub Woofer signal is so LF that this may not matter but it would be more on the lines of 'disco lights' control for it to work on level.
It depends upon what you actually want to achieve, though. I may have got the wrong idea about what's cool in motor car sound these days. I never have had room in my boot / trunk for large speakers as it's always been full of other stuff. haha