Howdy guys. I'm trying to make a subwoofer out of 2 small cones that I took out from a pair of old computer speakers. I don't need something that will shake the house, just a little compensation.
anyways, I decided to use op amps to build low pass filters. I have tried a 2nd order low pass filter circuit (got the circuit diagram from the data sheet of my quad op amp chip), it does amplify the low frequencies but there's a LOT of noise in it. That makes me wonder if a subwoofer circuit has more than just low pass filters in order to eliminate the noise? Thanks.

3/19/2006 4:08:25 PM

What opamps did you you use? Try the OPA2134 - made for audio and notoriously stable.

3/19/2006 4:40:24 PM

burr brown

3/19/2006 6:01:33 PM

when you say "noise" what sort of noise do you mean? I'm not aware of too many old computer speakers that will make many noises other than "fapfapfapfapfapfapfap" below 60Hz or so

3/19/2006 6:38:32 PM

well, the noise is like distortion. it's like when you go to your media player's equalizer and turn all the bass frequencies to 14db, except much worse.

3/19/2006 8:29:48 PM

Show the circuit and values you used, with power supply voltages.

3/19/2006 8:40:22 PM

these are the threads that i like to see around here

3/19/2006 8:54:54 PM

check your gains (Vout / Vin). is this using AC or DC voltage supplies? what is the impedance of the cones you're using? you may want to be sure that there is a more significant dB drop at the cutoff frequency...matter fact, what is the cutoff frequency you have it set to? specs would help us out. thanks.

3/19/2006 9:03:48 PM

thank you guys for the help. I'll make info available as soon as possible.

3/19/2006 10:33:26 PM

Hmm, I'm hungry. I guess I'll go in the kitchen and start rubbing sticks together.

jk

3/19/2006 10:41:45 PM

OK I posted the circuit diagram in my photo gallery. The chip is an LM324 quad op amp. The power supply is 18V DC straight from a big capacitor on the circuit board of that old speaker set, but I have tried 6V DC from 4 batteries too. There's no way I can find the impedance of the cones without an oscilliscope, except the resistance is 4ohms. I haven't derived the transfer function of that circuit diagram yet but I'll do it as soon as possible. Thanks.

P.S. (silently whisper) where is esgargs when you need one?

[Edited on March 19, 2006 at 11:28 PM. Reason : dadadada]

3/19/2006 11:26:44 PM

I'd build an active crossover, you'd need a highpass, lowpass, and bandpass

3/19/2006 11:33:43 PM

cool thread

3/19/2006 11:50:18 PM

The problem is your opamp.

Get a NE5532 or something better and burn the one you have.

You are using quite possibly the worst opamp for audio there is.

3/20/2006 8:00:21 AM

^duh....

I hate radio shack.

3/20/2006 8:14:17 AM

You're trying to do this with a single supply, which is easier from a power supply point of view, but generally more difficult from a audio quality point of view. I would highly recommend implementing a split supply setup.

Read this for your knowledge

http://instruct1.cit.cornell.edu/courses/bionb440/datasheets/SingleSupply.pdf

3/20/2006 10:09:47 AM

^there's only one power supply pin on the chip, do I use the ground pin for the negative supply?

3/20/2006 10:39:02 AM

Did you read even the intro to what I just posted?

They mention that specifically to check your data sheet. Since it doesn't take long I'll do it for you.

The LM324 as stated is shit for audio, get something else, but as it is, it was designed with single supply (straight from the data sheet it says this) in mind. Nonetheless, it will run off of dual per the data sheet

Quote :

" Wide power supply range: Single supply 3V to 32V or dual supplies ±1.5V to ±16V"

I can probably dig up some high quality burr brown Op amps somewhere in my parts bin that I'll give to you at a nominal fee as I probably won't be using them in the next 5 years.

[Edited on March 20, 2006 at 10:53 AM. Reason : Sorry, as for your question, yes, use the gnd pin for the negative supply.]

3/20/2006 10:52:41 AM

^sorry, I have been having class all day, so didn't get a chance to read it yet, but thanks for the info!

how much are they? I don't wanna go to radio shack again or order from digikey.

[Edited on March 20, 2006 at 4:19 PM. Reason : aaa]

3/20/2006 4:18:24 PM

you can get free opamp samples from nationalsemiconductor.com. they have all sorts of opamps

3/20/2006 4:24:46 PM

Let me look and see how much I paid for them and I can let you have them for similar or a little less.

I work over here in Cary off Harrison, so if you want I can probably leave them at the desk for you to pick up once you paypal me or whatever.

I'll be in touch (or PM me in case I forget )

3/20/2006 4:52:02 PM

^I don't have paypal account, and I dont' think I'll get one either...

3/20/2006 4:56:26 PM

This is the op amp I have

http://focus.ti.com/lit/ds/symlink/opa1013.pdf

5.51 in singles from digikey.

The problem is, I thought I had more than one of them.

Let me get back to you if I find the others (at least 1 other, I think I have 4-5 of them).


Edit:

Actually, let's back up quite a ways. Can you detail what all you are trying to do, and what you are using, specifically as your source (your PC I assume), your processing (pc + these filters?), and your amplification (this is what I am interested in most).



[Edited on March 20, 2006 at 11:26 PM. Reason : x]

3/20/2006 11:21:05 PM

^well, what I'm interested in, first is how much are they? then I wonder how its(your amplifier) performance will be at the low frequencies (<150Hz, but didn't find clear answer in the data sheet), and I guess I need to make a high pass filter to lower anything below around 40Hz to reduce noise (when I play with the equalizer in the Media Player, high gain at the 31Hz range produces the most noise)

I'm planning to make a 4th order low pass filter with 2 op amps, then a high pass filter with 1 op amp, and amplify the final signal and feed into the speakers. So I need at least 3 op amps since I have 2 other transistor amplifiers.

the signal comes directly from the computer, one branch goes into the regular sound system, and the other one goes through the low pass filter -> high pass filter (if needed?) -> amplify -> subwoofer speaker.

I don't know what gain I need, however I measured the Vout/Vin ratio of the regular sound system at modest volume, and it's about 30 or so.

3/21/2006 1:53:38 AM

Quote :

"^well, what I'm interested in, first is how much are they?"

$5.51 in singles from digikey + shipping. I thought I had more but now that I think about it, I probably only bought one since it has 2 per package. Like mentioned, you can get free samples sent right to your door so this might be a better route to take.

Quote :

"then I wonder how its(your amplifier) performance will be at the low frequencies (<150Hz, but didn't find clear answer in the data sheet)"

It's in there, but if you are going to be an engineer, you better start learning how to read data sheets and understanding what the graphs and information mean. Of interest to you are the Common Mode and Power Supply Rejection Ratio's versus frequency graphs. Over your frequency range they are well over 100dB, which is quite good.

Quote :

"and I guess I need to make a high pass filter to lower anything below around 40Hz to reduce noise (when I play with the equalizer in the Media Player, high gain at the 31Hz range produces the most noise)"

There is a lot of bass content below 40hz (heck, most of the good stuff lies right around here). With a proper low pass filter (tuned to between 80-120 for this application) you won't need any filtering below.

Quote :

"I'm planning to make a 4th order low pass filter with 2 op amps"

Why bother with 4th? Is your sub matched correctly with your mains so that such a harsh roll off is going to blend well? Or is this just a "4th is better then 2nd" because it gives a steeper roll off type of thing?

I like your desire to play around with an active filter for this project but it might be more worthwhile just to slap a passive filter in there and call it a day. Computer supplies (and indeed the sound cards/chipsets) are often nasty from a noise point of view. Your average subwoofer/computer speaker setup will have a lot more circuitry in front of it to reject the noise. I suspect that even with a proper split supply, good components, and good construction techniques, that you will still get noise in this system. I know because I have done it. This stuff isn't easy.

3/21/2006 8:48:26 AM

its not that simple

3/21/2006 10:27:51 AM

Edit that or I'll be sure to have bobby suspend you for trolling in here you worthless peice of shit.

When are you going to grow the fuck up?

[Edited on March 21, 2006 at 11:00 AM. Reason : x]

3/21/2006 10:59:54 AM

i enjoy reading threads like these. gg on trying to make something like this yourself

3/21/2006 11:21:20 AM

i say go passive also. you can build some quality xovers for $20 in parts. there is lots of free design software on the web.

3/21/2006 11:39:43 AM

Thank you people for the help. I have read through the posts and now I'm considering about passive filters. I'm going to the library and find some books about passive filters. I hope only resistors and capacitors will do it, for I don't like big huge inductors. Hey State409c, you said you have done it before, did you do the same thing I did?

3/21/2006 5:17:27 PM

I actually built a power amp based around a Sanyo STK4191V module. The construction wasn't perfect in that I didn't use the best grounding plane or layout techniques possible and when hooked to a PC or a corded CD player the 60hz hum was audible enough on quiet passages to be quite annoying but when hooked up to a battery powered player it was silent. I thought I took enough time with my first effort that I wouldn't have that problem but I wasn't successful. I didn't feel like redoing it as there are even simpler (less pins) chip amps available now (do a search for gainclone and be amazed). I use it as a easily portable amp for various debug purposes, etc.

It depends on what kind of crossover you want to build but a low pass is going to requrie an inductor no matter what. But, if you aren't going to be driving a lot of power through them, then you can skimp here, but, it's a nearly trivial matter to get some magnet wire and wind them yourself.

While you're at it, read up on "thiele small" parameters, because to do your own passive crossover yourself, you are going to need to know some basic driver characteristics, and this means you are going to need a multimeter at least, and a signal generator (your PC + some sine wave generating program will work).

And you probably can find most everything you need on the web without having to hit up the library.

For instance
Crossovers
http://www.the12volt.com/caraudio/cross12db.asp
Measuring theile small
http://sound.westhost.com/tsp.htm

3/21/2006 7:23:25 PM

^Thank you very much!

by the way, are passive filters going to perform better than active filters in the noise aspect?

[Edited on March 21, 2006 at 10:47 PM. Reason : sss]

3/21/2006 10:40:55 PM

^yes

http://www.the12volt.com/caraudio/cross12db.asp#lp

this need not be complicated

3/22/2006 8:51:41 AM