Choosing values of caps and resistors in high pass filter

[shot]

I know this is probably trivial question for most of you, but as a self thought enthusiast I couldn't find a place where this is explained, so I have to ask here for some explanation.

I want to put a simple passive high pass filter at the output of a unit I've built (CLX-VU to be exact), since I have some problems with sub frequencies it creates when processing low percussive material. The amount of energy below 20hz is too much and is triggering other dynamic units in chain  needlessly.
So I'll strap a RC filter at about 15hz at the both output pins.

And here's my question:

If I go and calculate this filter using passive HPF formula fc= 1 / (2*pi*R*C) I can get 15hz cutoff using 100nF cap and 100K resistor. But the same frequency can be obtained if I use 100uF cap and 100 ohm resistor!
What is the practical difference?
I know both will work.
My gut feeling gears me towards 100nF film capacitor rather than using bigger 100uF electrolytic, but I want to know more!
Or is there any other R/C combination that for some reason is more adequate in this situation?
Line output where I'm putting this filter is impedance balanced and is used mostly on unbalanced insert of my console (so going back to console's insert receiver) but sometimes it will be driving daisy-chained into other gear's input. If that is to any concern when choosing values.



Luka

 

[JohnRoberts]

adding a simple passive filter in series with an normal I/O interface must also factor source impedance of the outlet (in series) and input impedance of the following input in parallel. This will work for a fixed combination of products but results may get a little weird if you mix and match gear.

JR

 

[PRR]

> I can get 15hz cutoff using 100nF cap and 100K resistor. But the same frequency can be obtained if I use 100uF cap and 100 ohm resistor!

Can your source even drive 100 Ohms? (Most won't sound happy.)

Is it really "100K" if the box it feeds is 22K or 10K (very popular input impedances on modern boxes)?

Here's a dart-toss. Most gear will drive 1K, and most gear will show 10+K at its inputs. Take the Geometric Mean. (Multiply together, then take Square-Root.) 1,000*10,000= 10,000,000. Square-Root of that is about 3,000.

So 3,000nFd (what I might call 3uFd) and a 3.3K resistor (round-up, as you will see).

Key computation: 3uFd into 3.3K is 17.7Hz, acceptable.

With complications:
3.3K||10K = 2.48K, F3 becomes 21Hz
3.3K||22K = 2.87K, F3 becomes 18.5Hz

Since in-general, "nobody" has speakers that go below 50Hz, and the -3dB@21Hz case gives -1dB@42hz, and *all* rooms have +/-3dB or worse bass-bumps, the 18Hz-21Hz zone is not wrong.

In basic commercial caps, 3.3uFd is more common than 3.000uFd. So everything is 1.1 times lower(0.909). The "21" becomes 19, which is closer to your original (and perhaps arbitrary) spec.

OTOH, you could by a 10-pack of 1.0uFd, and try 2, 3, 4 etc in parallel until you find a happy balance of boom and freak-out.

As John says or implies: a single simple passive R-C is hardly a sharp tool, and highly affected by what you connect to.

 

[shot]

Wow! Yes!
THANK YOU!

 

[Mondy]

PRR,  I tip my hat to you mate!!

 

[Philbo King]

I know this is probably trivial question for most of you, but as a self thought enthusiast I couldn't find a place where this is explained, so I have to ask here for some explanation.

I want to put a simple passive high pass filter at the output of a unit I've built (CLX-VU to be exact), since I have some problems with sub frequencies it creates when processing low percussive material. The amount of energy below 20hz is too much and is triggering other dynamic units in chain needlessly.
So I'll strap a RC filter at about 15hz at the both output pins.

And here's my question:

If I go and calculate this filter using passive HPF formula fc= 1 / (2*pi*R*C) I can get 15hz cutoff using 100nF cap and 100K resistor. But the same frequency can be obtained if I use 100uF cap and 100 ohm resistor!
What is the practical difference?
I know both will work.
My gut feeling gears me towards 100nF film capacitor rather than using bigger 100uF electrolytic, but I want to know more!
Or is there any other R/C combination that for some reason is more adequate in this situation?
Line output where I'm putting this filter is impedance balanced and is used mostly on unbalanced insert of my console (so going back to console's insert receiver) but sometimes it will be driving daisy-chained into other gear's input. If that is to any concern when choosing values.



Luka

Electrolytics like 100uF have really sloppy tolerance, like -20/+50%. Best to use a polypropylene or similar with smaller C / bigger R. The impedance of the circuit also plays a big role in R vs. C selection.

 

[ruffrecords]

The practical difference is the input and output impedances of the filter. The input source needs to be not more than say 10% of the filter impedance and any circuit following the filter needs to have an input impedance of about 10 times that of the filter. If you do not do this then the filter characteristics and be significantly altered by the driving and following circuits.

Cheers

Ian

 

[Matt Syson]

Any band filtering must be done in a place where the input and output impedances are defined and 'inaccessible' to other equipment, so AFTER an input stage but BEFORE an output drive stage. Meaning proper bandwidth control must be done inside 'your' circuit and out of band interference must be carefully evaluated so smooth roll off below and above your intended design goal. Allowing frequencies of say 100kHz or greater to run riot inside an audio preamp must not be allowed, particularly now as there is so much high frequency 'noise' generated by mobile phones etc.

 

[JohnRoberts]

This thread has been quiet for years...

JR