updating mid-driver, effect on x-over? Need help

TommyVinegar

Member
2015-11-03 12:24 am
Hello,

I have a pair of old 3-way speakers I built a couple of years back. I used an off-the-shelf-crossover (nominally designed for 8ohm drivers), because I wasn't confident enough to build x-overs back then.

The quoted x-over points are 800 and 5000. That last number is especially high and is hard work for most midwoofers. It resulted in a slighty distorted, harsh sound in the upper midrange, but my amp was able to mostly equalize it out. Otherwise they sound fine and I used them happily.

Now i'll be moving these to a TV-room i'm building and they'll be the L+R in a 5.1 setup. I want to A: fix that upper midrange, and since i'm opening it up anyway B: replace the driver, ideally with the Peerless HDS-134 PPB I have in my center speaker, so the timbre matches a bit better.

As I cannot reach the x-over anymore I was thinking to add impedance to the driver using resistors, to force the x-over point down to, say, 3000/3500. Am I right in asuming this might work? Or do I have it the wrong way round? How do I calculate the value of the resistors, also taking into account the desired attenuation to match the woofer? Do I place them/it in line or parallel to the driver?

To further complicate the matter. When looking at the impedance graph for the Peerless HDS it (roughly) reads 11 ohms at 3000 and 16 ohms at 5000. Does this mean it will push the x-over point down anyways? Now, i'm thinking I should use a zobel to provide a more predictable load to the x-over AND add resistors to increase that load, to lower that x-over from 5000 to 3000. How would such a setup be wired-up?

This is getting a bit too complicated for me. Any suggestions are greatly appreciated.
 

AllenB

Moderator
Paid Member
2008-10-18 11:31 am
If you have an 8 ohm mid and put a resistor across it, say 8 or more ohms, you might be able to cut some of the higher frequencies. Taking it from 5kHz to 3kHz might be possible. Using this resistor is going to have a side effect similar to using zobel correction to the upper region of the band.

You won't be able to get too specific using just this method. It would be possible to add a limited range of extras to the crossover as it stands. Eg a higher order of filter, or notch filtering for a peak, impedance compensation or attenuation for a new more sensitive unit could be added after the existing crossover. An increase in the inductance for the primary filter could be put in before the mid crossover.

If you need more control you could bypass it altogether.