sensitive low frequency filter

[electrodoc2]

Hi,
Long time hobbiest first time op amp user!

PLEASE would someone help as i cant get to grips with the formula for bandpass filters/gain or anything regarding electronic filters.!

All I want is a phisicaly small high gain low frequency filter circuit which will use a xtal/moving coil mic to make a bulb flash to the bass beat of music (appx 40hz - 100hz) the voltage would be 3-9 volts and i'd like to use a dual opamp with a single supply line. unless theres a transistorised version anyone can think of.

Thanx in advance.. Rich

 

[MP1]

You are looking for a low pass filter? What is the cut off frequency that you desire and how much cut? These are necessary questions before you can build a filter, whether it is built with op-amps or transistors.

MP

 

[electrodoc2]

Hi MP.
Cut off frequency is appx 100 - 150 hz 12db gain would be good or 18db but i want to use a max of 1 dual opamp.

 

[MP1]

Here is the best you will probably get with two op-amps. I have provided the formula in the drawing. You already have the cap values, you just have to figure the resistors from the frequency of cut-off.
Just round the resistor calculated figures to a number that is an actual resistor value.
The gain is 1.586 and has to stay there for the circuit to work properly.
If you need help with this, let me know.
MP

ThirdOrderLowPass.pdf

 

[russlk2]

This circuit will provide the lowpass and adjustable gain:

View attachment 35130

 

[MP1]

Yes, this is a single op-amp filter with a gain stage ahead of it. So, if you are going to limit your design to only 2 op-amps, you have a choice. Do you want the more attenuation of signal outside of the frequency pass band or do you want more overall gain of your circuit? If you want both, you can just add another op-amp stage for gain. The more stages of frequency filtering you add, the more attentuation you will have of the frequencies you want to reject.
Have Fun!
MP

 

[electrodoc2]

thanx alot peeps!, i can add a transistor on the end of that to flash the bulb, so that should do fine.

cheers.. Rich

 

[Kevin Weddle]

This circuit is not biased properly. The load impedance of the opamp is too high. It will work in the end, but the response will not be that of what is intended for an opamp. You have to DC bias it correctly before you use a signal. The current ouput of the first stage is simply to low.

 

[MP1]

Are you referring to the one I posted or the one russlk posted?

MP

 

[Kevin Weddle]

Both. Where is the ouput bias? You need a resistor to ground or a negative voltage. What you have is something other than what is intended for an opamp. What happens is an increase in voltage on the emitter causes an increase in voltage on the base. The base has to be able to maintain it's voltage against the signal. This is why it is grounded. It's like running differential signals when one signal dictates over the other rather than producing a sum. Plus, you need the output bias current.

 

[MP1]

No. That is not a problem. This is ok for an op-amp. You cannot look at an op-amp as if it is a transistor.

MP

 

[Kevin Weddle]

The reason it works that way is because the external current dictates the voltage. This is, apparently, always the case with an opamp. I think it's interesting that it appears to work, when you should actually bias it right. You know, of course, that the opamp will work differently depending on how you bias it. In other words I can make the thing appear to obey one law and not another or disobey both laws.

 

[electrodoc2]

Hi i couldn't get it going anyway.. I could use 4 opamps and i want the signal to be amplified from a microphone, it must be using a low voltage single supply and then to flash a bulb to the bass beat of music. Any other ideas would be most marvalous. Thanx

Rich

 

[MP1]

Let's see what you have so far. It might just be a simple modification. Any chance you could put a sketch up here on the forum?

MP

 

[audioguru2]

MP: Isn't your circuit a 4-pole filter?
Rich: Russ's classic 3-pole filter will filter the bass very well but will not have enough gain for a microphone unless it is placed right up against a very-loud speaker. But don't simply increase the gain of the 1st stage becuse it will overload during loud vocals.
Add a variable-gain stage to the output of Russ's circuit which will be the same as his 1st stage except replace the 100K feedback resistor with a 220K volume control in series with a 10K resistor. Couple into the new stage with a 10 microfarad cap.
Of course you must cap couple your mic to the input, and bias an electret mic. This circuit probably won't work with an xtal mic.
You must also cap couple (10 microfarads) the variable-gain stage to your bulb-driver transistor and use a 10K resistor in series with this. Put a reversed diode across the base and emitter to allow the cap to charge and discharge equally.
Use a 1000 microfarad cap from Vcc to ground.
Now your bulb will flash (not very loud since it receives only 1/2 wave rectification) to the bass.

 

[audioguru2]

OOps, the bulb won't be very bright, not loud.
Why not use a super-bright LED in series with a current-limiting resistor instead of a bulb? Then adjust the resistor for the maximum brightness.

 

[MP1]

It is a third order lowpass filter. The gain is fixed so I am not surprised that Rich is having a problem with the gain. Although you can add gain to the output, if he is using a passive input such as a microphone, then he is going to need some amplification at the input. Microphones give out a very weak signal. Another option is to use a powered mic.

MP

 

[audioguru2]

MP, let's hope that Rich can get the circuit working.
Don't you count 4 equal RC networks on your filter? It looks like two 2-pole filters cascaded, but the first 2-pole filter isn't quite a multiple-feedback type since the feedback resistor doen't connect to the junction of the 1st resistor and cap, but instead is connected as a simple integrator.

 

[MP1]

If I am not mistaken, the term 2 pole or twin T refers to two networks that look like a T built from capacitors and resistors where the bottom of the T is connected to ground. If the two components that make the top of the T are resistors, then you have a low pass filter whereas if the two components that make up the top of the T are capacitors, you have a high pass filter. In this type of circuit, you would cascade as many T's as needed to get the roll-off desired. With an op-amp for each section you do not lose the amplitude.
The high order filter is a little different. The second section should look familiar as a sallen-key high order design. The first stage is a simple first order low pass network which definately has the T network you were referring to. There are a lot of designs for filters, but the most popular that you might find is the Chebyshev and the Butterworth. Many others are one variance or another of these.

MP