RE: External crossover units

Milo

Riaan C wrote: , but I tended to pick up a bit of a rolling feedback from time to time. Muso acquaintance that dropped by had a look and opined it was because I had no crossover.

No! The lack of a frequency divider has no impact on the directionality of the system.

Alan Ratcliffe wrote:
Crossovers have different roll-offs (how fast they change the level of the frequency after the cutoff point), measured in dBs per octave. Generally, the steeper the roll-off (the higher the number) the better. So 24dB per octave is better than 18dB, 18dB is better than 12.

4th Order slopes introduce more group delay than 1st,2nd and 3rd order so 24dB slopes are not better than 18dB slopes! The slope used will depend on the physical displacement of the two systems and the delay necessary to align them for the intended coverage area.

Alan Ratcliffe wrote: It's an easy calculation to figure out - 1ms per foot or 3ms per metre (the speed of sound).[/li][/list]

The delay needed will depend on the temperature and relative humidity and cannot be assumed. An impulse response of the two signals need to be measured and then aligned accordingly. The calculation will also only work for a single position in space and for someone on the other side of the venue the system might not be perfectly aligned. This is a trade-off and the most important area should have priority...

TomCat

Milo wrote:
Riaan C wrote: , but I tended to pick up a bit of a rolling feedback from time to time. Muso acquaintance that dropped by had a look and opined it was because I had no crossover.

No! The lack of a frequency divider has no impact on the directionality of the system.

@ Milo.......Hi and Welcome to the Forum.....trust you will have lots of fun here.

However....Personally....I would beg to differ.........Different loudspeaker drivers based on their diameter will have different directive indexes at the same frequency....possibly creating some off axis funnies.

There is also the issue of the interaction of different driver types covering the same frequency bands, possibly creating off axis lobes that can be picked up by any microphone within that lobe area

By applying an appropriate crossover to separate the loudspeakers will reduce these off axis lobes and could possibly have improved the rejection of feedback by the stage microphones.

In the case of Riaan's PA application, we do not know the physical placement of the loudspeakers relative to the sub woofers so anything could have happened off axis.....By minimizing frequency overlap, the system would have improved clarity and definition....and lobing of the loudspeakers could have been more controlled.

Anyway... the science and application of crossovers, delay in loudspeaker systems etc is a complex one. At the end....How good does it sound......

Generally....I believe that a suitable crossover would have helped in a number of ways.

BTW @ RiaanC.....Did you get the Rane.

If you did, Prosound are the dealers for Rane and they will have replacement PSU's available.

AlanRatcliffe

Hi Milo, Welcome to the forum.

Milo wrote:
Alan Ratcliffe wrote:
Crossovers have different roll-offs (how fast they change the level of the frequency after the cutoff point), measured in dBs per octave. Generally, the steeper the roll-off (the higher the number) the better. So 24dB per octave is better than 18dB, 18dB is better than 12.
4th Order slopes introduce more group delay than 1st,2nd and 3rd order so 24dB slopes are not better than 18dB slopes!

Good point. So basically you have a trade-off between steeper slopes and shorter delay?

The slope used will depend on the physical displacement of the two systems and the delay necessary to align them for the intended coverage area.

Interesting. So in a smaller system like the one under discussion, I gather you would use the gentler slopes to keep the delay introduced by the crossover short, while in a bigger system you can use steeper slopes as you are going to be compensating for the delay caused by displacement anyway?

Alan Ratcliffe wrote: It's an easy calculation to figure out - 1ms per foot or 3ms per metre (the speed of sound).
The delay needed will depend on the temperature and relative humidity and cannot be assumed.

In a larger system covering a wider area, yes. But in a relatively small system in a relatively small room, I think the approximation should be sufficient.