An Overview of Filters in A Crossover Network

April 07, 2020

In electronics, filters serve a critical role in many common applications such as audio electronics, radio communications and power supplies. We use filters to block or pass a specific range of frequencies. The filters can be either passive or active.  We usually divide filters into four main types, depending on which frequency components of the input signal they pass on to the output signal. The four types of filters are low-pass, high-pass, band-pass, and band-stop.

In this article, we will take a closer look at the band-pass filters and focus on their application in the crossover networks of the stereo system loudspeakers or the loudspeaker enclosures that are found in our homes. Individual loudspeakers have different efficiencies for various ranges of frequencies. For example, a woofer speaker, which is a large diameter speaker, typically has more efficiency for low frequencies or bass tones than the high-frequency signals.  

Therefore, we need filters to direct the signals to the appropriate speakers in the loudspeaker enclosure. The utilization of low-pass filters, band-pass filters and high-pass filters prevents signals of a specific frequency range from getting through to a particular speaker, depending on the designated frequencies for each speaker.

In Figure 1 below, we have a diagram of a passive crossover three-way system. It contains speakers, namely Woofer, Midrange (Squawker) and Tweeter. The Woofer speaker is the most efficient from 0 Hz – 630 Hz, Midrange is operating from 630 Hz - 8 kHz, and Tweeter from 8 kHz and up.

Coming into this crossover network is the audio signal spectrum which typically runs from 20 Hz up to 20 kHz. A low pass filter passes the frequencies from 0 – 630 Hz over to the Woofer. In the middle, the band-pass filter passes a band of frequencies that exist from 630 Hz up to 8 kHz. The high-pass filter passes any frequencies 8 kHz and above over to the Tweeter.

Frequency Response Curves 

The response curves for the three different filters are shown below. Figure 2 (a) illustrates the response curve for the low-pass filter, which starts at a frequency of 0 Hz along the X-axis, with a bandwidth of 0 – 630 Hz. At the point of 630 Hz, the output achieves a value that is 70.7% of the maximum output value at the lower frequencies. This point is called the cutoff or critical frequency (fc) point.  

In the high-pass filter – Figure 2 (b), the response curve has no response at the low frequencies, but at a frequency of 8 kHz, we have a response that is above 70.7% of the maximum value. This response is the output or the bandwidth of the high pass filter.

In the center of the crossover network is the band-pass filter – figure 2 (c), which is designed to have a very sharp, defined frequency response. A band-pass filter is equivalent to combining a low-pass filter and a high-pass filter. The cutoff frequency (fc) of the high-pass section becomes the lower frequency limit in the passband f1. The upper frequency in the passband f2 is the result of the cutoff frequency in the low-pass section. The passband, or the bandwidth, is the difference between f2 and f1 points. The f1 is designed to be 630 Hz in this case, and the f2 is 8 kHz.

So, there are three individual filters, with three separate responses, and the filters are directing the appropriate frequencies to the respective speakers. Depending on the design and purpose of the band-pass filter, the bandwidth may be very wide or very narrow. Overall, the crossover network has a net of a combined response of the three filters.

We hope this has been helpful to you as a Technician or a student entering the field. If you have any questions about the Electronics or the Electromechanical Technician programs you can reach one of our Program Consultants toll-free at 1-888-553-5333 or by email at [email protected].

Comments

Submitted by Mark Lucitt (not verified) on Mon, 09/12/2022 - 09:22

There should be a mention of phase changing in the crossover. The first order LP and HP filters each contribute 45 degrees of phase change but cancel each other out because one is positive and one is negative. The midrange will have some phase change and his is usually a compromise to get the steeper octave frequency cutoffs.

In reply to by Mark Lucitt (not verified)

Submitted by Iris on Tue, 09/13/2022 - 16:26

In this article we looked at a general overview of bandpass filter and did not cover the phase response of the filters since this content may need to be discussed solely on its own. If you desire to have an in-depth understanding of phase relation and response in active filters here are a few links:

https://www.analog.com/en/analog-dialogue/articles/phase-response-in-active-filters-2.html

https://www.analog.com/en/analog-dialogue/articles/phase-relations-in-active-filters.html

 

Add new comment