Group delay in cardioid designs
 

Advice!!

I note one pro manufacture is creating cardioid subs with front facing driver and using a bandpass (4th order) for the rear facing, all in one cabinet.

As they are using "one amplifier" I am assuming they are not using an applied delay to the rear already out of phase output from the bandpass section of the cabinet.

I have modelled some designs of my own with Win ISD and Winspeakz and can get the responses of the front and rear sections to track each other in response but the band pass section is, by the nature of bandpass designs, tracking the front section with a difference in group delay of around +2ms.

Is this the same as applying a 2ms delay, because if it is, and correctly dimensioning the cabinet, I have cracked it!!

Not too technical a response please as I am an artisan not a scientist!!

I think 2 ms is minimum depending on wich center freq you want to attenuate.
this artical migth be usefull
btw delay aply's to the front element.

Thanks EPA,

Actually doesn't answer my question though!

If you have a fwd facing sealed box sub with a rear facing additional bandpass section resulting in two group delay curves that actually track each other but say, 2ms apart..is that the same as applying an actual 2ms delay to one of two identical monopole loudspeakers in cardioid configuration from a DSP or analogue (with delay) crossover?

If that is the case one only has to dimension the cabinet to create the cardioid/hyper cardioid response without the need to apply actual delay?

Thanks for the link..I use EV laps 11 for quite a lot of my research!

Yes, if the band-pass section has an acoustical output which is the same as the delayed "identical monopole" this would apply in the same way. Uniform delay is still uniform delay whether it comes from a DSP box or from using a band-pass enclosure for the rear drivers. The output from the band-pass box has an actual acoustic delay relative to the front-firing monopole, and there is a frequency range where that delay is fairly uniform.

However, since the band-pass has an intrinsic low-pass built-in as part of its response, its output will start rapidly dropping off, while the delayed monopole would still be going strong. At that point the effect from the band-pass will NOT be the same as the delayed monopole. So the lows would most likely be deliberately crossed over below that transition frequency, to avoid or reduce this issue.

Thanks Jack,

that is what I was hoping! The design I have tracks the group delays around 2.25ms difference from 30Hz to around 90Hz where an active 4th order will be applied.

Brilliant......thanks!!