Low pass filter gain compensation

[marcelomd]

Hi guys,

I'm playing with a variable frequency low pass filter (100Hz-5kHz). Just a simple potentiometer wired as variable resistor and a capacitor into an opamp.

The problem is that the variable resistor messes with the amplitude of the signal in the passband. Is there a (simple and/or usual) way to counter that and keep the pass band at unity gain?

With the values I'm using, the passband at the lowest setting is -9dB and -30dB at the highest.

Thanks a lot!

[Rob Strand]

To avoid attenuation you need the pot resistance (the largest resistance of the filter) to be 10 times *smaller* than the load. Yours looks like it is 30 times larger.

However, I doubt you will be able to do it with such a large frequency control range.  So what you will need to do is put a buffer between the low-pass filter and the circuit.

[marcelomd]

I'm trying to split my signal in two, high and low frequencies. I want to adjust both frequencies independently, a la Dug Pinnick's signature amp (that's where I got the frequencies from, but it's just a guideline).

My low side signal chain so far is [Voltage follower]->[variable low pass filter]->[inverting buffer]->[inverting mixer, with high side]. Pretty much like this. My filter is already sandwiched between buffers.

I calculated a 50k pot (and a 1k limiter) plus a 33n cap. I guess I can lower the values and keep the proportion.

Thanks!

[Rob Strand]

>[variable low pass filter]->[inverting buffer]

The inverting buffer is the problem.  It has medium input impedance.
This is what is loading the low-pass filter.   Hard to get around the problem.

In the link you gave the buffer is non-inverting.  It doesn't load the low-pass filter
and avoids the problem.

Try to change it to:
[variable low pass filter]->[non-inverting buffer]->[non-inverting mixer, with high side].

I guess I can lower the values and keep the proportion.

You will struggle to get it to work unless you have very large resistors on the inverting buffer (which cause noise).

[marcelomd]

Aaaaah now it works as expected. I thought a JFET opamp had enough impedance in this case.

I wanted to save a stage and still be phase correct at the output. I guess I'll put a simple common emitter after everything.

Thanks a lot!