Axiom Home Page Forums General Discussion Advice from Axiom Owners Crossover on the EP subs.

Hi sirquack,

I will expand on this when Ian returns to the Axiom plant (10 days) and I visit and get some listening-test time.

The comments were added to the article on the suggestion of Ian Colquhoun and Tom Cumberland, who designed the filters in the DSP circuits in the amplifier.

I know it contradicts the conventional wisdom that one should avoid "cascading" of crossovers, but I was assured that it's worth experimenting with and that, depending on the AV receiver or AV processor's bass management, it may produce a smoother transition. I'm not sure I hear any difference in my own installation so far.

Regards,

Thanks Alan, I look forward to your input.

That's what SHE said.

I'm glad we are a bunch of comedians around here, versus a bunch of grammaredians.

Usually when I tweak crossovers its with an eye to the SMS real time display so I can see its effect on FR. Blending the sub/speakers to achieve flat FR on either side of the crossover point depends on various factors such as slope of the low pass filters , phase, level etc. The conventional wisdom about not having two low pass filters in tandem is that it plays havoc with phase at the crossover point and can cause irregularities in FR at and near the crossover frequency. Having a brickwall filter like we have in the EP subs also means a discontinuity in sub's response above the crossover point compared to the more typical 24db sloping Linkwitz-Riley 4th Order low pass filters you find in most other subs.

That brickwall filter works well if the sub is level matched to the speakers, you are not running the sub hot and if the speakers have flat FR. Room acoustics not withstanding. As we all know Axiom speakers are engineered to linear FR. However most people (not me) like running their subs 3-5db hot so that alone will mean a discontinuity in FR at the crossover point because of the DSP brickwall low pass. Also depending on the processor filter it may or may not introduce phase anomolies or possibly intermodulation distortion.



Perhaps there is something about the DSP chip that avoids the usual problems with using two low pass filters? Conceptually I can see why it could work but only when the sub and linear speakers are level matched and the processor low pass filter doesn't cause phase problems. Anyway, I'm looking forward to some clarification on this one.

..-..

I would expect that the DSP filters in the Axiom subs and receivers are implemented using finite impulse response filters. This method avoids phase distortion.

How does running the sub hot create a discontinuity in frequency response?

It doesn't if the receiver crossover is set lower than 100hz and the sub is left in bypass since you continue to get output above the receiver crossover point and below 100hz. Setting the processor crossover at 100hz or more, means no rolloff just the brickwall response at the sub. There is no output above the brickwall DSP low pass filter unlike the typical rolloff you see with fourth order filters. You can see the steep clifflike drop in amplitude from the FR graphs.

So running the sub hot and crossing at 100hz or close to it means a steep drop in response at or slightly above 100hz, a discontiuity in overall FR if you will depending on how hot you run the sub above the speaker level. Its less of an issue the lower you cross and not a factor at 80hz receiver crossover having a 24db slope with the sub crossover bypassed. However, it is an issue as you cross closer to the 100hz brickwall because, except for some nominal port resonance, their is no output above 100hz.

Discontinuity in response may be a problem if you are trying to blend small speakers that are rolling off above 100hz because there is no output coming from the sub to fill in above 100hz. A firmware revision which raises the low pass brickwall to 120hz will mean more of the output from the sub above the receiver crossover point. In addition higher crossovers help when blending two way speakers or bookshelf monitors.

Thanks to those of you that have brought up this issue.
I too would like to see an increase of the brick wall to 120 Hz or more. Today's receivers are increasingly flexible and allow a wide range of crossover frequencies. Of course this is assuming that the FR of the EP500 above 100 Hz is good enough to allow for this.

It would be interesting to hear from Axiom concerning the reasons for designing the brick wall at 100 Hz. I am guessing it relates to the standard of setting all crossovers at 80 Hz. However, the 80 Hz crossover is not always appropriate.