passive to tame 8khz peak

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
i've got a peak at 8khz that i can't get out with the equalizer on the radio.
after putting the 8khz knob all the way down, it is still louder than the other peak.
i tried to raise the other equalizer knobs to further equalize the peak, but my 20hz and 31hz starts to distort with the equalizer setting up that high.

my only option is to put a filter in line to reduce the peak, because at first thought.. i figured the peak wasn't as loud and obnoxious when music was playing, but after listening for a while, i realized i was fatigued.. and the only thing i could think of was the 8khz peak causing the fatigue.
i knew it was there as if 'atmosphere' but i thought maybe it was louder with the test tone than with music .,and while that might still be true, the fatigue is there and i want to get rid of it.
because other than that.. the pressure inside the cabin is rather linear from the lowest lows up into the midbass on into the midrange.

the weather outside is too cold to pull out the computer to use the microphone for a note of how much louder the peak is from the rest.
but my first thought was something like a notch filter, and with that said .. i seen a filter calculator on the internet search results.
well i put in the value for the resistance of the speaker and the desired frequency.
that got me some values from the calculator.
i decided to check the radioshack website to see if they had anything at all locally.
they had a 100uh inductor and a 4.7uf capacitor.
my first thought was 'okay.. the cap might be a little lower than needed, but i can boost those frequencies since the equalizer already has them at full attenuation.. and i'll just unwrap the inductor a little bit to get the value to match the frequency'

well i went and got it all and hooked up the pieces to the soundcard to test with some audio analyzer software, but the first run simply showed a break in the phase at 6khz .. and after unwrapping the inductor a little bit, the results were much different than i expected.

there wasn't any attenuation at all.
i tried only the inductor, no attenuation.
i tried only the capacitor, no attenuation.

what the broken phase defines in the audio software, i don't know.
but regardless of a sine sweep or pink noise with the RTA, the attenuation wasn't there.

somebody want to offer some clarity?
words or build it for me, either one or both works here.
You will probably want to put the (notch) filter between preamp and amp, or, at any rate, before the power amp. That way you won't need high-current components.

An active filter, i.e. with an opamp or two, might work best, since it can be made to have a high input impedance and a low output impedance. The Texas Instruments site ( had a free download called FilterPro that was very good. But it could be done with just passive components, too. LT-spice is a free download, from, that would enable you to simulate various filters of all kinds.

But you would need to know how deep and how wide the notch needs to be, before you could design ANY filter that would work properly.

You do want a NOTCH filter. Bandpass lets everything through, in the band. A notch is like the inverse of a bandpass. It blocks everything around the notch's center frequency.

A notch filter can be made to be narrow or wide in its reach around the center frequency, and can be made to have more or less attenuation at the center frequency. So you will need to know what shape to make it. And that will require measuring the shape of the response that you want to correct, in order to be able to know the shape of the proper correction filter's response.
i'm glad i didn't make the mistake of listening only to 8khz..!

i think the saddest part is,
connecting anything that doesn't run on a battery to the auxilery input causes noise in the signal.
the only time i've ever connected something without a battery is a laptop and a desktop with microphone.

the only way to get a solid signal from the radio is:
1. burn a cd with pink noise
2. get some audio isolation transformers to break the connection

i'm a bit more excited about the number 2 option because i can use it again elsewhere if necessary.
and that way i can connect my laptop to the power invertor since the new battery quit working.
then i'll be able to use a vst equalizer to turn down 8khz further, but i was hoping and searching for a solution to get the 8khz down for if ever somebody has a reference CD they want to hear through the speakers.

the audio hum from the laptop sounded like a steady ground loop, as well as varation whenever the CPU usage changes.

i emailed behringer to see if their HD400 hum eliminator would get rid of both, because i don't want to go through the hassle of returning the item for a refund if it doesn't work.

there's no chance i'll be putting the filter before any preamp or amplifier, because i'm running two twelves rated at 600 watts RMS with only 150 watts RMS
(bandpass box with the plexiglass removed)
the front speakers are just loud enough to be heard while driving on the highway, so i dont see any need for seperate amplifiers (cant afford them either).

my monthly savings go as low as $6 with simply buying food & gasoline.

but i want to get this filter installed & working simply to be able to enjoy a flatter frequency response, as well as teach other people the power of calibrating the equalizer.

the stock 6.5 inch speakers were decent, but they lacked a tweeter.
these 6x8 speakers kinda fell on my lap.
they are 3-way speakers, and i can run them full range without distortion.
the speakers are the 50 watt RMS version rather than the higher 75 or 90 or 125 watt RMS type.

i soldered the speaker wires, had one backwards and had to re-do it.
then i put the fader back to the neutral position, only to find out the back speakers are out of phase with the front.
but with that, i came to realize the decay time inside the cabin is a lot better.
i flipped the phase of the subwoofer and started noticing details in the audio that simply weren't there before when the stock speakers where in phase with the back speakers and the equalizer done somewhat flat.

i figure i'm hoping behringer has something good to say, that way i can get those pieces of hardware and at least enjoy some flatter frequency response sooner rather than later.
and then when the weather improves, i can get the microphone in there to really note how many dB of difference the 8khz peak is.
i'm also hoping to grab an impulse response file to use it in a convolution filter to get the response a little bit better.

i'm kinda hoping when it is done, i'll be able to use the vehicle as an advertisement to see if the person wants their equalizer calibrated - to simply show them the difference with and without calibration.

i figure there's a lot of people out there that simply don't get their frequency response flat as they should when enjoying movies or television or music.
Where do you have that peak?
How do you know you have an 8KHz peak?
What is "the other" peak?
Why do you touch the 20 and 31Hz knobs to correct an 8KHz peak?


when i sit in the driver seat, i can twist my head and the peak doesnt seem to change.

i put some test tones on a cd for each of the equalizer bands, because i didn't have a calibrated microphone yet and i wanted to adjust the bands by ear to get it somewhat better than without.

the other peak is 800hz
but there is also a peak around 3khz too

the reason to touch 20 and 31hz knobs is..
if your turn 8khz all the way down and it still isnt enough,
as long as all the other knobs are neutral, boosting all the other frequencies will further reduce the 8khz because the gap between the amplitude increases.
i see there is a parallel notch filter and a series notch filter.

i dont get how either one of them can be connected like that, yet there isnt any confusion of the electrons.. because the connections look like they create a direct point of confusion.

which method is better?
are there pro's and con's ?
if i go with the parallel notch filter type, is the resistor simply the point that adjusts the attenuation?
(because if it does, maybe i'd choose a variable resistor to really help dial in the peak to get it flatter)
the email i got from behringer said as long as the noise was generated from the laptop the box would get rid of said noise.

i figure if the cost for parts to reduce the 8khz peak was $18 (plus shipping)
doubling that amount for the box to use the laptop gets me 31 bands of equalizer adjustment instead of simply 1.
i also get the option to play with an audio toys to make the sound more spacious.
and the laptop plugged into the power invertor will last longer than 30 minutes the battery gave me.

seems like the list of reason is rather strong going the route of the box.

then when i get the microphone out, i'll be able to use the impulse response file in the convolution plugin too.

the reason i dont use the laptop with the microphone is because the laptop's soundcard doesn't support the full frequency response.
it stops recording at about 3khz
but with 31 bands from the equalizer, grabbing only the mic and preamp might be something to do if the weather is around 40's or 50's (still too cold for the desktop).

but if somebody else ever wants their vehicle's equalizer done and there's noise in the connection, i'll already be with the box to eliminate the noise.
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.