Variable filter

Status
Not open for further replies.

Dr.EM

New Member
Hi. This is a very preliminary design for a sort of variable low-pass filter I came up with. I have little knowledge of such things, but i'm certain some of you smarter people will know whether it will work, or how to improve it, etc.

My thinking behind it is simply that the op-amps will act as bufers (the resistor values are guesses). The variable resistor then changes the charge and discharge time of the capacitor; higher resistance = larger charge/discharge times = lower cutoff point.

Well, thats essentially all. I am aware that there are some values missing, so if anyone could help find some appropriate ones, that would be great (also, the 5k and 10k arn't ideal, I would rather have no gain or loss in this circuit, and an input imedance of 10k or so).

Thanks for looking!
 
That won't quite cut it. The output impedance of the first op amp is essentially zero, so changes in the potentiometer will have no effect on the output of the circuit. Actually, if you turn the pot down to zero ohms, the op amp may become unstable and oscillate.
Below is a simple buffered single-pole variable lowpass filter. If you already have a low source impedance, you can get rid of the first op amp. Without it, you can still have an input impedance of at least 10k by simply making that the value of the fixed resistor.
 
Cheers for that! I am sort of looking for a synth filter as part of an fx unit project, but don't need resonance or CV really.

BTW, you mention "single-pole", what exactly is that, and what difference would it make to have more poles?

I also noticed the below circuit online, but i'm not entirely sure of how to wire it up. If I wanted to try this one out (I think its a 12db one?) where would I connect what is labelled as "Vr", or could I leave that unconnected? Would I also be able to leave out the CV in? And also, is a dual polarity supply needed, I assume it would be?

Cheers again
 
In technical terms, poles are the roots of the transfer function. This simple filter (the one I drew) has a pole at f=1/(2*pi*R*C). More poles causes the rolloff above the cutoff frequency to have a steeper slope. However, optimum multiple-pole filters are more complicated than a simple cascade of single-pole filters. Filter design is a complex subject.
For the schematic you posted, the circuit that generates Vr (2 R's and a C) is only necessary if you use a single supply. It connects to the bottom end of the 1M resistors. If you use dual supplies, omit the Vr voltage divider and cap, and simply ground the lower end of the 1M resistor.
The CV In operates the same as the Freq. pot. You could eliminate the CV input and just use the pot for tuning.
 
Ah, so this single pole filter would be a 12db (?), then a 2 pole= 18, 3 = 24 etc?

I understand now about the Vr thing, thats pretty useful, as it means I can use a single supply (presumably to run this off of a 9v battery). So, to use single supply, I would just connect the "-" end of the battery to ground(s), and the "+" to "+9v, v+ and the op amps + terminal"? Then, connect Vr to Vr, and Vr to the ground on the 1M variable res?

I attached an image of how I think I should be wiring this up to use single supply because, well, it makes everything a lot more understandable

Thanks for all your help, btw, you've been very helpful!
 
A single pole filter rolls off at 6 dB per octave, a 2 pole at 12, etc.
You've got the schematic right, except the 1M level trim pot needs to stay connected to GND as it was originally. It will potentially screw up the bias if you connect it to Vr.
Note that the LED is being used as a voltage-controlled resistor. It will probably not emit enough light to be visible.
 
Ok, that makes sense.

If its only 6db, the effect surely wouldn't be very obvious, without resonance on at least? Or is this one 12db?
 
Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…