LEDs pulsing to music...

[Dark Shade Slayer]

Hey all,

I have been browsing around for schematics etc for a circuit to pulse LED's to music, and have found only ones that involve a TIP31 transistor. Now I enjoy building things with what I have, and I don't have a TIP31... So I tried to find a suitable substitute for it. Because these are LED's, they draw very little current (around 20mA if I understand correctly (each of course)), so I thought for just one or two LED's maybe a 2n3904 would work. I made a circuit with it and it did not work... the LED's are on all the time, or don't come on at all depending on how I connect them, but they never flash. So I proceeded to try a 2n2222 (I assume that's what it is, it says MPS2222A and has a Motorola symbol on it...), it didn't work, I then tried what I believe is a 2SC2274 (it just says C2274 on it), and it didn't work... Seeing a pattern yet? Anyone have any ideas as to why this might not be working? I think my logic works out, but maybe someone else would say otherwise (which is fine)? If anyone wants to see the datasheets I can provide them, as for schematics Google has a bazillion of these things and they are almost all identical.

Thanks,
Dark Shade

 

[Reloadron]

That circuit floating around on the internet leaves a little to be desired. If it is the one I am thinking of, it does not even include a base resistor for the transistor. You also do not mention what you are using to drive the transistor? Some use a line out signal and some use a speaker lead?

If you do a Google of "Color Organ Circuit" or "Color Organ Kit" you should get some decent results. The 12 volt versions that drive small incandescent lamps should be easily hackable for what you want. There are countless threads in this forum about that TIP 31 circuit which does not really work well when it works at all.

This link should be of some help and remember if you just want for example bass response the circuit can be reduced in parts count.

Ron

 

[Dark Shade Slayer]

Thank you very much for the link, and sorry I was not too specific in my first post. I am using a 3.5mm jack with a male-male cord from whatever device I want to get input from as my source for the project (likely to be my computer later on, but no need to accidentally damage it in testing so I use a portable mini-disk player for the testing phase... this comes after blowing up my ipod ). I can understand almost all of the circuit, but am a little confused as to exactly what the virtual ground does. I just did some reading, and I think it is to keep the voltage steady throughout the circuit? If so, couldn't a 12V voltage regulator (such as the lm7912) in a small circuit be used? If not, can I substitute an lm358 instead of the chip they use (TS921), as I believe it just requires a basic op amp? Oh and also, I am 99% sure this would work (assuming pinouts are the same(I'll be checking before building it)), but I would also be using an LM324 quad op amp.

Thanks,
Dark Shade

 

[Reloadron]

OK, before I forget, the virtual ground. Audio is not a steady state signal, it is an AC signal constantly going above and below a 0 volt ground reference. The op amps used in circuits like this use a single supply voltage so an incoming signal can't swing below 0 volts (actually not quite zero). So what is done is place a voltage like 1/2 of Vcc at the non inverting input (+) of the op amp. Now at the output the input signal can swing above and below virtual ground. That being an over simplified explanation.

The reason they like using the TS921 is because it is a Rail to Rail op amp. That means the output can be driven to the voltage level at the rails. You really want a rail to rail op amp if possible. If you can get any of these they should work, I am not 100% on that. The Rail to Rail is preferred. This is a good link and mentions a kit available from Jameco Electronics.

The LM7912 you mention is a Neg Regulator? You want a LM7812 which is a Pos Reg and would work fine.

Ron

 

[Dark Shade Slayer]

Lol, looks like I've got to triple check my component names before clicking "post". If I someday get around to buying parts, I will keep those in mind. Meanwhile I did a bunch of substituting of parts (this is probably why it doesn't quite work) and came up with the attached schematic (I also included the original). When the circuit is plugged in (12V wall adapter), it blinks the LED's then they fade and wont come back on... any ideas (other than I subbed to many parts )? I decided to only make one part of the circuit for simplicity.

Thanks,
Dark Shade

 

[Reloadron]

When you say they come on, blink and fade what happens to the supply voltage at that point in time. The circuit also really needs the right op amps. Also, on some op amps the power pins are not the same so you may want to check the data sheets.

Ron

 

[audioguru]

The TS921 "virtual ground" opamp is not needed. The two 100k resistors voltage divider and 0.1uF filter capacitor can easily provide a reference voltage to the extremely low current (+) input of the filter opamps and the voltage and current do not change.

The filters have a bandwidth that is too narrow at the peaks and too wide further down because they are very simple.
The high frequency filter should be a second or third order Sallen-Key Butterworth highpass type and the low frequency filter should be a second or third order Sallen-Key Butterworth lowpass type. The mid filter can be a combination of both.

EDIT: Here is an example of the frequency response of the "bass" filter. Notice that it has no deep bass.
The other filters have the same shape.

 

[Reloadron]

Thank you AG for finding this. What is U1?

Ron

 

[audioguru]

Thank you AG for finding this. What is U1?

Ron

U1 in my simulation is any half-decent opamp. I would use a TL071 single, TL072 dual or TL074 quad audio opamps.

 

[audioguru]

Here is another example of the mids frequency response of the Multiple Feedback Bandpass filter in the project (a whistle), compared to a Sallen and Key Bandpass filter (a real bandpass filter):

 

[Reloadron]

Thank you AG. Nice on the frequency responses also.

Ron