Adjustable passive line level XO

[knifeinthesink]

It occured to me that, in theory at least, a passive line level low pass filter could be made with a pot to create a variable filter that would be useful for experimenting.

The filter is a resister in series with the amp and a parallel cap.

The recommended resistor value is between 5k and 10k and the cap is calculated with the following formula:

c = 1/2pi(Resistor)(Frequency)

Heres the site I got the info from:

http://www.t-linespeakers.org/tech/filters/passiveHLxo.html

I have a few questions though.

Is it important that the resistor be within the range of 5k to 10k as recommended.

My idea was to use a 1K resistor in series with a 10k pot (the 1K resistor is just to keep the filter from going into wierd frequency ranges) so that I could vary the filter from about 50hz to a couple hundred or so (Dont remember the exact frequencies as calculated but you get the idea).

Is there anything Im overlooking here other then the insertion loss that will cause this to go ary.

Also, the site I was reading about this on suggests using pots to balance the filters, but a pot in series with the low pass will change the frequency of the filter and after would change the frequency of the high pass, assuming its used. Where should it be inserted. Im also assuming a passive pre is not an option with these. Am I correct.

[Kal Rubinson]

1. I'd sooner swtich caps than vary the resistor but that's a personal bias.
2. If you want to do a LP filter and level control, an intervening buffer is a good idea.

Kal

[Amphiprion]

1. I'd sooner swtich caps than vary the resistor but that's a personal bias.

Is that to keep insertion loss constant? If so then I agree.

[Dennis H]

Heres the site I got the info from:

http://www.t-linespeakers.org/tech/...assiveHLxo.html

Those formulas are overly simplistic because they don't account for the amp's input impedance. Art Ludwig has more complete ones.

http://www.silcom.com/~aludwig/Sysdes/Crossove_Design.htm#Bi_amping

As Kal and Mark imply, and Art's formulas show, twisting the pot will change the volume of the signal as well as the frequency. That may or may not be a problem depending on whether your amp has a gain control.

My idea was to use a 1K resistor in series with a 10k pot (the 1K resistor is just to keep the filter from going into wierd frequency ranges)

That's a good idea for another reason. If you twisted the pot to zero ohms, the preamp would be looking at a dead short at high frequencies through the pot and the cap to ground. The fixed resistor should be the lowest impedance your preamp is comfortable driving.

[Davey]

Knife,

Pretty much everything is a variable with PLL designs. If you want something that can adjust a single parameter and leave everything else the same it's much easier to accomplish with an active (buffered) unit.

Dennis,

Peter calls it "Ramp" in the high-pass example, but he does kinda ignore it with the low-pass example. Although in that case the amp input impedance relative to the series resistors primarily causes the insertion loss.

Davey.