A little crossover help please. (1 Viewer)

[Dustin B]

In another thread on this forum I was explaining why you should crank the crossover on a sub when you are using the crossover in the reciever.

The sub in question in a Paradigm PDR10. The suggestion is that setting the PDR10's crossover (which I think is 2nd order) to 100hz and also using the receivers 80hz crossover results in a flatter frequency response than cranking the PDR10's crossover and just using the receivers 80hz one. Here is part of the post from the other thread.

20hz through 90hz, where it starts to falloff.

I know highpass rumble filters on subs can cause different levels of boost above the filters corner freqeuncy (is that the right term?). But can a low pass filter cause a decrease in output below its corner frequency like this quote is suggesting. Ignoring the cascading crossover problem for now.

Any thoughts?

 

[Dustin B]

Not even an opinion :frowning:

 

[Richard Greene]

You wrote:

"The sub in question in a Paradigm PDR10. The suggestion is that setting the PDR10's crossover (which I think is 2nd order) to 100hz and also using the receivers 80hz crossover results in a flatter frequency response than cranking the PDR10's crossover and just using the receivers 80hz one.

I know highpass rumble filters on subs can cause different levels of boost above the filters corner freqeuncy (is that the right term?). But can a low pass filter cause a decrease in output below its corner frequency like this quote is suggesting. Ignoring the cascading crossover problem for now. "

My reply:

Sharply limiting output above 80Hz. with an 80Hz.

24dB/octave low pass filter makes a subwoofer close to

sonically invisible (subs sounds better when their location is sonically invisible). If a receiver has an

80Hz. 12dB/octave low pass filter and the subwoofer also

has a 12dB/octave low-pass filter that can be set at 80Hz.

using BOTH at the same time is the least expensive way to achieve 24dB/octave low-pass filtering at 80Hz.

(the circuit used 24dB/octave filters is actually two 12dB/octave filters cascaded -- so there is nothing inherently wrong with cascading two 12dB/octave filters).

Rumble filters are high pass filters.

Filters don't boost anything, they only cut SPL

at different rates per octave. (6, 12 and 18dB per

octave filters are down 3dB at their turnover frequency

and 24dB per octave filters are down 6dB (3dB + 3dB)

at their turnover frequency.)

 

[Dustin B]

I'm not sure if you understood my question. The claim was that setting the low pass filter on the PDR10 (which is variable from 50hz to 150hz) to 100hz would result in the subs response between 50hz and 90hz being attenuated 2-4dB allowing for a flatter overall frequency response from the sub than if the variable crossover was set at 150hz.
All the while the receivers 80hz crossover is in effect. So if there is no problem with a pair of 12dB/octave low pass filters at 80hz, what about a 12dB/octave at 80hz and another at 100hz?
My view was (and still is with my current knowledge level, unless someone can change that knowledge level such that I will then believe differently) that setting the crossover on the PDR10 to 100hz would not attenutate the 50-90hz range and that combining a 100hz and 80hz 12dB/octave filters on the same speaker was not a good idea. The other poster was recommending it.
Still on topic but a little off, I must be a little confused about rumble filters. Here is a quote from Adire's site on their AVA250 amp:

But what really sets the AVA250 apart from the rest of the competition is the built-in high pass rumble filter. We call it the Sane Ultralow Rumble Filter (SURF). What is SURF? Well, it's a second order high pass filter. Nothing unique there! But what sets SURF apart is the way it's designed: low Q, low frequency!
SURF uses a Q=0.96 filter with a corner frequency of 18 Hz. The result is flat response down to 14 Hz, with a VERY mild bass boost of just 1 dB at 25 Hz. SURF eliminates the need to manually rebuild your plate amp, just to eliminate that annoying excess boost (in some cases, as high as 6 dB at 30 Hz!). SURF works with this amp to keep the headroom around down deep, and to keep the bass natural and tight.

Also are the resistors being changed at the end of this PDF file not part of a high pass filter? If they aren't, what are they a part of?
http://www.partsexpress.com/pdf/300-794.pdf

 

[Wayne A. Pflughaupt]

Dustin,

A crossover is suppose to attenuate the audio signal by a pre-determined amount beyond the designated frequency (i.e., -6dB/octave, -12dB/octave, etc.). It would be expected for a crossover set at 100Hz to have some effect at 90Hz. But if it is having an affect on output all the way down to 50Hz, that crossover is frankly a piece of junk.

I don’t know where the fellow you mention gets his information, but if for some bizarre reason this 80Hz, 100Hz configuration with this sub worked for him, it would be an anomaly specific to his room. There’s not much chance it would be good advice for everyone.

Bottom line, Dustin, I agree with your contention that his advice is not good. Response between 50 and 90Hz is very much room dependent. The best way to flatten response in that region is with an equalizer.

Richard is right that 24dB/octave slopes are best. Personally, I’d like to have them steeper.

The purpose of a rumble filter is to roll out the extreme lows, not boost response in certain places. It sounds like this is something designed specifically to enhance the performance of this particular amp.

Regards,

Wayne A. Pflughaupt

 

[Pete Mazz]

I used to do something similar when I had a Sony receiver that had a fixed Xover of ~120 Hz. In order to stop the sub from reproducing higher freqs, I would set the sub amp Xover to ~120 Hz. The combination yielded the best compromise I could get with the limited flexibility of the Sony.

Pete

 

[Dustin B]

So then my initial understanding was correct. Low pass and high pass filters in and of themselves will not cause any boost, and will attenuate the signal above or below their corner freqeuncy at a rate dictated by their order (6,12,18,24 dB/octave). Not a small amount over a broad range of frequencies.

The rumble filters on those subs are a modified high pass filter that can cause the boost reported in the PDF file.

Can anyone explain to me what actually happens when a pair of filters with different but close corner frequencies cascade with each other?

 

[Richard Greene]

The rumble filters on those subs are a modified high pass filter that can cause the boost reported in the PDF file.

RG:

If there is a boost, the circuit is an equalizer plus a high pass filter, not solely a high-pass filter.

Can anyone explain to me what actually happens when a pair of filters with different but close corner frequencies cascade with each other?

RG:

The frequency response measured in an anechoic chamber is a little less smooth.

In a real rooms where reflections and standing waves cause wild frequency response deviations, I doubt if you could hear a difference between an 80Hz. 24dB/octave low-pass filter and two 12dB/octave filters cascaded (one at 80Hz. and the other at 70Hz., for an example).

In general, if you have an 80Hz. 12dB/octave filter in your receiver, ANY additional low-pass filtering at 80Hz. or higher will usually make a subwoofer sound better ... even if the second low-pass filter is at a considerably higher frequency such as 100Hz. This is a cheapskate method to get better subwoofer performance and not perfect ... but it's better than using a subwoofer with lot's of output over 80Hz. I wonder if people who criticize cascading filters/crossovers ever experiment with this and make measurements at their listening position.

Experiments with 48dB/octave filtering have not revealed any

significant advantage over 24dB/octave. -- I suppose that's why there are so few active 48dB/octave crossovers manufactured.