2 pole low frequency shelving filter design

[Northy]

Hi,

I'm trying to design a 2 pole low frequency shelving filter to boost the low frequencies in an audio circuit.

I've modeled this circuit and it models correctly, but I want to run it on a single supply op amp (LM324).



Should I connect R15 and R16 to 1/2 the supply voltage to make it work correctly?

Thanks,

Northy

 

[Arouse1973]

Hello Northy
Yes you can, but you can also connect a capacitor to the bottom of R15 to isolate any D.C. Where is your filter components?
Adam

 

[BobK]

I don't see how that circuit does any filtering.

Bob

 

[Arouse1973]

I don't see how that circuit does any filtering.

Bob

Yes quite, as asked.
Adam

 

[CDRIVE]

Ha, I assumed he's using the data sheet's Cin value of the inverted input for "C diff".

Chris

 

[Northy]

I must admit I don't understand how it works, I'm trying to figure that out too. I was told by a friend to use the design from this web page:

http://www.linkwitzlab.com/filters.htm#5

 

[BobK]

Well, you didn't copy well then.

The circuit you posted has 2 stages, both containing capacitors in the feedback network. These will be filters. The circuit you posted is just an amplifier.

Now, in response to your original question. Make the resistors that connect to ground, two resistors of twice the value, one to ground and the other to v+. You will also then need series capacitors on input and output.

Bob

 

[Northy]

Ooops, sorry that was me messing around with the simulation software, I'd deleted the capacitor. I can assure you that the circuit I built had it.

So I should try this:

 

[Arouse1973]

That looks better. Have you considered a bandpass filter? You could the choose a range of bass frequecies that would be amplified. Have you chosen a frequency and gain you want?
Adam

 

[Northy]

What I'm trying to achieve is ~7-10dB of gain at 300Hz and down to 0dB gain at 1kHz and above.
Could I achieve that with a bandpass filter?

Thanks,

Northy

 

[Arouse1973]

Ok LP filter is ok. I thought you might have wanted a band of bass frequencies boosted and not just from 0Hz to 300Hz. I would probably use a MFB filter for this circuit.
Adam

 

[BobK]

5V is too low for use with an LM324. Its output can only swing to within 2V of V+, so your output could only swing from 0 to 3V which is not enough for line level audio, presuming that is what you are aiming for. I would use 12V, or at least 9.


Bob

 

[Northy]

I've actually sort of got it working, but I'm struggling to work out how to bias it properly, could someone explain how to do it for a single supply design please? I'm happy with the output only being able to swing ~0V to ~3V as it's going to be fed into an amplifier stage.

Thanks,

Northy

 

[BobK]

You already have it biased at 2.5V. This would allow it to go only 0.5V to the + and 2.5 to the minus.

You could redo it to bias to 1.5V in order to get the maximum symmetric swing.

To make it 1.5V the two resistors should be chosen such that it is a voltage divider in the ratio of 1.5/5 and the parallel combination of the two resistors comes to 2.2K.

With R1 being the resistor to ground and R2 the resistor to 5V:

R1 / (R1 + R2) = 1.5 / 5.0

R1 * R2 / (R1 + R2) = 2.2K

I'll let you solve that.

Bob

 

[Northy]

I've been playing with the bias voltage (feeding it in with a bench supply), and it seems to need ~0.5V to center the output. That's confused me a little,

Is there any reason why the Inverting design would be better to use?

Thanks,

Northy

 

[BobK]

How is the -0.5V applied? Are you using that instead of ground? That makes no sense.

Bob

 

[Arouse1973]

I've been playing with the bias voltage (feeding it in with a bench supply), and it seems to need ~0.5V to center the output. That's confused me a little,

Is there any reason why the Inverting design would be better to use?

Thanks,

Northy

Where has your 2K2 resistor gone?
Adam

 

[Northy]

This is the circuit that I'm playing with at the moment. It seems to work if I feed about 0.5V into V_Bias. I'm confused why it is only 0.5V though.


Thanks,
Northy

 

[Laplace]

I'm confused why it is only 0.5V though.

V_Bias sets the '+' input voltage. R13 & R14 form a negative feedback voltage divider for the '-' input. Because of the 100 dB DC voltage gain of the LM324, the offset between the '+' and '-' inputs will be in the range of 10 μV. So the common assumption is that both op-amp inputs are at the same voltage, in this case V_Bias. So the quiescent point of the amplifier output must rise to a value that, when divided by R13 & R14, equals V_Bias. Is that anywhere near the Q-point necessary for the LM324 to operate from a single supply of 5 volts?

 

[Northy]

Sorry you've lost me a little bit there!

Northy