# RC HPF filter noise?

#### benbiles

##### Member level 3
Hi , I'm trying to work out if the resistor part of an RC passive HPF introduces noise to a signal path.

I would like to use a digi pot to adjust the filter sweep position but wonder if the THD of the resistor is a factor?

I will be using the resistor (digipot ) as a shunt between the balanced lines on an audio input of a preamp.

I can see if you pass the signal directly through a resistor it would introduce noise, but how about in RC filter configuration ?

I would place 2 x capacitor in the hot and cold signal path lines and the digi pot between them in a U configuration.

The idea is to have simple adjustment of a HPF from somewhere between 40khz -> 200khz

#### KlausST

##### Super Moderator
Staff member
Hi,

I'm confused. Do all these sentences talk about the same situation or different situations?
If you talk about different situations, then please divide them with 1) 2) 3)...so we can answer them using this numbers.

Don't use a textual description, draw a schematic. Hand drawn is O.K.

******

* a resistor introduces noise. No matter how you connect it.
* THD of a resistor should be ignorable. (But ceramics capacitors are known to cause THD)
* digipots come with datasheets. Read them. Use digipots only that specify noise and THD.

If you use an RC on an audio input, then the source impedance (the output impedance of the device that generates the audio signal) will modify the cutoff frequency.

Klaus

#### FvM

##### Super Moderator
Staff member
A 40 - 200 kHz HPF for an audio signal? Seriously?

Apart from this point, it's useless to discuss noise and THD without considering the signal level. Also source and load impedance play a role.

I you are specifically referring to resistor noise, it can be exactly calculated. E.g. any SPICE simulator does.

#### benbiles

##### Member level 3
sorry , i am trying to understand if the resistor introduces THD into the signal path and how. not noise.

yes the idea is to make a low frequency cut adjustable from 40khz -> 200khz or maybe higher..
yes I am looking at datasheets of various digipots ( resistor ladders ) and the THD they introduce.

I understand the impedance seen at the input would add to R1 to modify the RC filter.
can I add the RC in between preamp stages to work around that?

my question is , does R1 introduce harmonic distortion to hot and cold signal path and how to calculate?
What do I need to know about the resistor in order to make the calculation?

How is the resistor distorting the signal?

the input would be from a microphone so in the +/-mV range

impedance of the microphone say 100ohm?

the preamp input impedance is 9.2K

my attempt at drawing with a mouse was not great.. next time i'll try a pen!

Last edited:

#### FvM

##### Super Moderator
Staff member
O.k., microphone circuit. I presume a useful HPF cut-off frequency will be in audible range, e.g. 40 - 200 Hz rather than kHz? Or are you designing a bat recording studio?

I don't expect that resistors inject significant THD in mV level range, neither simple resistors nor internal resistors and analog switches of a digipot. Microphones are usually nor operated with matched load impedance, mirophone amplifiers (e.g. mixing deck inputs) have typically multiple kOhm input impedance. Filters are looped-in to the signal path after the pre-amplifier.

Some microphones already include a switchable high-pass filter, adding another passive HPF may cause an unwanted frequency characteristic.

#### benbiles

##### Member level 3
O.k., microphone circuit. I presume a useful HPF cut-off frequency will be in audible range, e.g. 40 - 200 Hz rather than kHz? Or are you designing a bat recording studio?

I don't expect that resistors inject significant THD in mV level range, neither simple resistors nor internal resistors and analog switches of a digipot. Microphones are usually nor operated with matched load impedance, mirophone amplifiers (e.g. mixing deck inputs) have typically multiple kOhm input impedance. Filters are looped-in to the signal path after the pre-amplifier.

Some microphones already include a switchable high-pass filter, adding another passive HPF may cause an unwanted frequency characteristic.
:smile::smile: Yes !! 40 - 200hz bass cut !!! its not an ultrasound studio device sorry !

I am trying to mimic the HPF of a microphone but make it adjustable.
many Sound Devices audio mixer recorders have this function.

I don't know if they positioned the filter after the preamp , in their manual it says before. I don't think the added 100ohm or 200ohm of the microphone makes enough difference to the filter if you choose the right caps. I think the idea is if you filter out large bass rumble before the preamp you can have more headroom ( dynamic range ) for usable audio.

I'll take a look at microphone impedance's and see what the difference is , maybe there some odd ribbon mics that are 50ohm or something crazy.

Its a shame i don't have my sound devices 633 mixer recorder with my right now , at this point i'd pull it apart and take a look where the filter is!

DS1882 has roughly 0.003% THD+N across 20hz-20khz but i wonder if its necessary to use such a high spec. for this kind of filter.

I would use electrolytic caps for the filter not ceramic.

#### KlausST

##### Super Moderator
Staff member
Hi,

I would use electrolytic caps for the filter not ceramic.

Best is a film/foil capacitor for audio.

Klaus

#### benbiles

##### Member level 3
Hi,

Best is a film/foil capacitor for audio.

Klaus
I'm currently using Nichicon KZ Series Electrolytic Audio Caps for DC blocking on the output of preamp to ADC and for 48v DC phantom power blocking and they seam OK to me although I'm sure there is better! My mixer sounds way better than my expensive sound devices mixer recorder at least.

The preamp IC datasheet i'm using recommends DC blocking caps between preamp and ADC rather than a buffer.

can you recommend any non electrolytic makes i should try ? I'd love to take a listen to some alternatives and see if I can notice the difference in an AB setup.

I have an oscilloscope with a basic FFT that i think I can use this to measure the THD of a sign wave passing through the filter.

I'll just build the filter with the digipot and give it a try. I have little confidence my cables and setup will be able to detect very low 0.003% THD though but who know , maybe i'll learn something if i try

- - - Updated - - -

ok, my scope is in fact 8bit so no chance of measuring THD in a signal of 0.003% !!!

I'll take you advise and make sure I choose a digipot that reports both THD and Noise in the datasheet and go from there.

#### KlausST

##### Super Moderator
Staff member
Hi,

an electrolytics capacitor is made for DC voltage across it. It has a "+" terminal and a "-" terminal. Never connect the the other way round.

So if you usit it as 48V blocking capacitor there will be always "+" at the positive terminal. It never becomes negative.

But in your case it will see AC. Sometimes positive, sometimes negative. No job for an electrolytics.

*****
The NICHICON web site and datasheet advertise them for "audio". Are you sure this means "low level signal voltage"?
It could be used for an audio DC supply as buffer, it could be used in a car audio for DC blocking to the speaker..

Klaus