Inductor hi pass to filter low bass

Background:

I am running a Supravox 215 2000 EXC as a midrange driver.

The driver is designed to run full range, but is obviously limited in the high freq and low bass.

I have the unit open baffle with the baffle size acting as a 1st order filter into a 15" sealed bass unit (130L) @ 250hz. It is also augmented by a tweeter above 10KHz.

There is no electrical hi pass filter on the mid, although there is a broadband LRC notch filter (for the baffle peak) and an input resistor to match level with the bass.

Everything is great, but I do notice that low bass is causing the Supravoxs to buzz.

Question:

I really want to avoid putting a capacitor in series with the mids, and besides the value would have to be huge.

Can I put an inductor across the terminals after the pad and the RLC notch filter?

What value would I need to for an F3 of 80Hz?

Would there be any side effects, audible or otherwise?

Thanks

David
 

Curmudgeon

Member
2005-11-02 7:13 pm
The notch filter will not provide a significant impedance at low frequencies. Then the pad, if set to zero, allows the amplifier to see the inductor directly across its output; this is NOT A Good Thing.

I agree with your desire to avoid the series capacitor. I think the best solution might be to add a separate chamber internally to enclose the mid in an infinite baffle, sized to roll off the mid at 80 Hz. That will keep the cone loaded at low frequencies, and limit its excursion. It's currently free to move as far as it can at frequencies below its resonance.

In such a small enclosure, I find it important to reduce internal reflections off the walls, and of course keep the dimensions non-integrally related. (e.g. 8" X 16" bad, 8" X 13" pretty good. ) Prevent reflections, especially off the back wall with at least a 1/2" thick felt lining. (Felt with a high wool content, but not dense.) And I use acusta-stuf to help reduce standing waves and absorb reflections.
 
[wiki=I think the best solution might be to add a separate chamber internally to enclose the mid in an infinite baffle, sized to roll off the mid at 80 Hz. That will keep the cone loaded at low frequencies, and limit its excursion. It's currently free to move as far as it can at frequencies below its resonance. ]%[/wiki]

Curmudgeon, I like this suggestion! :) How do you calculate the size of this chamber? And I gather this same strategy can be applied to other full range drivers? I am wanting to do the same thing with the Alpair 10. One further question; will this small chamber cause a loss in efficiency of the driver?
 
Thanks for the reply.

The notch filter will not provide a significant impedance at low frequencies. Then the pad, if set to zero, allows the amplifier to see the inductor directly across its output; this is NOT A Good Thing.

The input resistor on the Full Range is 7 ohms. I think this is part of the reason for the buzz, in that the amp isn't controlling the cone in the bass.

I agree with your desire to avoid the series capacitor. I think the best solution might be to add a separate chamber internally to enclose the mid in an infinite baffle, sized to roll off the mid at 80 Hz. That will keep the cone loaded at low frequencies, and limit its excursion. It's currently free to move as far as it can at frequencies below its resonance.

In such a small enclosure, I find it important to reduce internal reflections off the walls, and of course keep the dimensions non-integrally related. (e.g. 8" X 16" bad, 8" X 13" pretty good. ) Prevent reflections, especially off the back wall with at least a 1/2" thick felt lining. (Felt with a high wool content, but not dense.) And I use acusta-stuf to help reduce standing waves and absorb reflections.

I definitely don't want to go this way. OB was my design brief. Lack of efficiency in the woofer has forced me to pad the Full Range driver down. It is designed to run full range in an open baffle, and has a high Qms.

Eventually, I would like to go fully active, which will give much better control over the drivers. I was just looking for a temporary solution.

Thanks

David
 

Curmudgeon

Member
2005-11-02 7:13 pm
@ Ant: There are a lot of simple box design programs, which take a few Thiele-Small parameters, Vas, resonant frequency, and Qts and give you a few choices varying box size vs. flatness and extension. I've only used LEAP for a long time, so I'm out of touch with what is available. Efficiency above the rolloff frequency is not affected. The infinite baffle is -12dB/octave, so the rolloff is 2nd order.
I despaired of finding a series cap that was affordable and good enough to match my ATI (Skaaning) mid, and this eliminated that problem rather neatly.

@dcathro I gather then that the pad is fixed; an L-Pad? If so, what is the shunt value? The inductor's impedance at 80Hz needs to equal the the parallel impedance of: driver impedance (nominally 35 ohms, but changing very rapidly so close to the resonant freq.)and the pad, whose output impedance is itself (assuming an L pad) the parallel resistance of the series and shunt resistors, assuming a low amplifier output impedance. At 80 Hz then, -3, and at 50, perhaps -5, depending on the resonant frequency of the actual unit.
 
@dcathro

If there is a 7 Ohm resistor in series to the midrange
you can go that way if efficiency does not matter and
your amp does not bother with a lower impedance
in the bass.

You might have to replace the resistor with a higher
wattage type ...

Since you are talking about a "preliminary" solution,
have you already tried a capacitor ?
 
Hi,

The short answer is yes you can use an inductor after the 7R series resistor
and in this case it will work far better than using an overall series capacitor.

The 7R series resistor is undamping the fundamental resonance and using
a further series capacitor will make it stick out even more, even if it rolls
off the upper bass somewhat.

The place to start is with a 4ohm (that is the correct impedance in this
case) car subwoofer crossover inductor, 7 or 8 mH for around 80Hz.
Place it before the RLC.

rgds, sreten.

41dASvDJdhL._SL500_AA300_.jpg
 
Last edited:
@ Ant: There are a lot of simple box design programs, which take a few Thiele-Small parameters, Vas, resonant frequency, and Qts and give you a few choices varying box size vs. flatness and extension. I've only used LEAP for a long time, so I'm out of touch with what is available. Efficiency above the rolloff frequency is not affected. The infinite baffle is -12dB/octave, so the rolloff is 2nd order.
I despaired of finding a series cap that was affordable and good enough to match my ATI (Skaaning) mid, and this eliminated that problem rather neatly.
Thanks Curmudgeon! I hadn't really thought of using Unibox for this reason so it was a simple process once you suggested this strategy. Came up with 4.6L Qtc .707, F3 92 Hz. This should work as it is only 2dB down at 80 Hz.

Sorry for OT Dcathro!
 
@ Ant; Glad you liked it! Only further comment would be to note that many prefer a slightly overdamped alignment, in the Qtc 0.6(?) region or so. Might be worth the experiment if feasible.

@dcathro; given the poor rolloff characteristics of a single pole, especially against the rising efficiency of the resonance region, perhaps a shunt trap would be better...
 
@ Ant; Glad you liked it! Only further comment would be to note that many prefer a slightly overdamped alignment, in the Qtc 0.6(?) region or so. Might be worth the experiment if feasible.

Thanks again Curmudgeon. Qtc of 0.6 gives an F3 of 97Hz and and Fb of 79Hz. Its only 3dB down at 80Hz. Volume is 6.6L. This is still workable but it will be a case of deciding on which one to go for, rather than making both. :xfingers: