RL passive high pass line level filter

[noviygera]

After experimenting with the a 0.005uf capacitor on the amp input I would like to try make the RL version of the high pass filter (6db/octave slope). Just to see how different they sound...

Anyone can recommend a calculator or formula to get the resistor and inductor values?

My amp has input impedance of 91K ohms, crossover freq is 335Hz, first order.

Thank you,
Herman

 

[AndrewT]

Search crossover calculator
Using 91k & 335Hz, I get C=4n7F or L=47.7H

You are not going to easily hear the difference for yourself. Finding an ~47H inductor will be either very difficult, or very expensive.

 

[DF96]

A 47H inductor is likely to be further away from an ideal component than a 5nF capacitor. Also bigger, heavier and more expensive. Better at picking up hum.

 

[noviygera]

Are we on the same page?

I'd like to clarify:

I need to calculate the values for a RL HIGH PASS filter. the one that has a resistor in SERIES and inductor as the SHUNT (across positive and negative). Line level.

NOT a low pass filter with a huge inductor.

Do I still need the 47h inductor?

Herman

 

[dave slagle]

you just need to use the formula z=2*pi*F*L and solve for your desired resistance value. Something around 8K should work and nets you a 3.8hy inductor. Since the inductor will be harder to source I would get that first and simply size the resistor to set the crossover frequency. Say you found a nice audio quality 4HY inductor, then 2piFL tells you you need an 8.4K resistor.

 

[noviygera]

you just need to use the formula z=2*pi*F*L and solve for your desired resistance value. Something around 8K should work and nets you a 3.8hy inductor. Since the inductor will be harder to source I would get that first and simply size the resistor to set the crossover frequency. Say you found a nice audio quality 4HY inductor, then 2piFL tells you you need an 8.4K resistor.

Dave,

Thanks for your response. From this formula I see that the lower the L then we get a lower resistance value. So why not use a small inductor, lets say 0.2mH or 0.0002 henry.

at 335Hz this gives a resistance value of 0.42ohms. Are these values usable for a line level filter or am I missing something?

Thanks,
Herman

 

[AndrewT]

yes, you are missing something.
You line level source has to be able to drive a 0r42 load.
Not even a headphone amplifier can do that.

You originally gave a 91k load. That's the one that needs the enormous inductor.
Reduce the load to 8k3 and you need a very big inductor.

BTW, my original component values were for 100k and 334Hz

 

[DF96]

You could try doing some reading on first-order filters. Guess what - that is how we learnt to do it.

 

[dave slagle]

hey,

as andrew T mentioned, your source needs to be able to drive the load under worst case situation which will be low frequencies where the inductor impedance is low. The series resistor needs to be sized to prevent the source from taking a dive. Most sources are happy with values in the 8K range and in turn the 8K value is low enough to not be bothered by the 91K input Z of your amp.

dave

 

[bear]

What sort of amp?

If it is a tube amp, then the place to do all this is inside the amp, not at line level at the input, unless you think you will want to swap it in and out. But even then, you can set up a shunt/switch depending on if you use a change in capacitance or an inductor and where you do it. (you can do it in a solid state amp too, but it takes a little more to get that done right)

If your amp has GNFB then you can also consider using the feedback loop to do the same HP job.

The good news on your line level thing, is that if you look around you will see that people have been there before, a long time ago, and companies like UTC and others made chokes for this job that are small - look at the "ouncer" line for example. Also you could use the winding of a suitable "mic" or other line level matching xfmr, if it has enough inductance for the job... or Dave Slagle can wind you a pair, no doubt! 😀

 

[Mark.Clappers]

You can also use a simple sim to check the frequency response, input / output impedance, group delay, ...

I use this one a lot: AADE Filter Design and Analysis


It's free and easy to use..

 

[noviygera]

"fast" capacitors

What sort of amp?

If it is a tube amp, then the place to do all this is inside the amp, not at line level at the input, unless you think you will want to swap it in and out. But even then, you can set up a shunt/switch depending on if you use a change in capacitance or an inductor and where you do it. (you can do it in a solid state amp too, but it takes a little more to get that done right)

If your amp has GNFB then you can also consider using the feedback loop to do the same HP job.

The good news on your line level thing, is that if you look around you will see that people have been there before, a long time ago, and companies like UTC and others made chokes for this job that are small - look at the "ouncer" line for example. Also you could use the winding of a suitable "mic" or other line level matching xfmr, if it has enough inductance for the job... or Dave Slagle can wind you a pair, no doubt! 😀

Thank you for the info, Bear. I am not informed enough to use the windings of an existing transformer, like the ouncer, all I see are the pri and sec impedances, not sure how to convert this to inductance values I need. Is there a formula for that?
I am using a solid state amp and I do not know how to integrate the filter circuit "inside". But I was told by one fella that the line level capacitor quality may change the sound characteristics. So if the "KimberCap" I tried is not "fast enough" I will try some Wima FKP2 type caps or vintage PIO for reference. Not sure that "fast" is the accurate word to describe the capacitor but indeed the sound is being subjectively rounded, transients softened. Will take it from there I hope, when I get other caps.

Best regards,
Herman