When first-order filter cannot thoroughly eliminate the unwanted frequencies, outside passband, we need second-order to do instead. And the second-order seems to be popular choice found in many crossover designs because it may be sufficient for this task. However, it’s obvious that many hi-end or expensive speakers tend to employ higher orders, such as fourth-order, in these days. What are the requirements for using higher-order crossovers when second-order is sufficient to eliminate the out-of-passband frequencies? When to know the second-order is no longer sufficient and the speakers need higher-order crossovers? Any signs or symptoms?
What are the requirements for using higher-order crossovers when second-order is sufficient to eliminate the out-of-passband frequencies? When to know the second-order is no longer sufficient and the speakers need higher-order crossovers? Any signs or symptoms?
If out of band suppression goals are met, and good ripple-free summation takes place throughout the xover region down to say -24dB per side, I'd say no need for higher order.....al least on-axis.
If off-axis shows too much lobing due to excessive separation distances between drivers, higher order can help reduce the frequency range that lobes
Personally, I draw the line at 4th order for IIR; I think any order above that needs to be FIR lin-phase. (and even 4th order benefits)
When to know...?
Apart from the on and off axis summations, I'd say....
Is there any excursion distress at the bottom end of the range, either audible or indicated by rising THD of the drivers span?
A higher order can help there, but depending it can also hurt. Higher order will put more energy into the xover corner (the xover knee will get sharper). So if that knee range is the frequency range talking back, the xover frequency will need to be raised to use higher order.
Likewise for the high end of drivers response, is the driver going into breakup mode to produce high frequency needed for ripple free summation down to -24dB?
On aside I've found interesting...in comparing the thermal load presented to a driver using low orders and high orders using the same xover frequency, it makes very little difference. This is with pink. Music of course may differ by who knows how much..
@presscot
See this example of my lastest 3-way speaker built some weeks ago.
I had to use a mid-range with a horrible breakup above 4-5kHz.
So I selected a 3rd order for the low pass to cut aggressivelly above around 3kHz.
Regarding the low frequencies, the mid-range had a better behavior, so a 2nd high pass was enough.
Each situation will require a specific slope.
See this example of my lastest 3-way speaker built some weeks ago.
I had to use a mid-range with a horrible breakup above 4-5kHz.
So I selected a 3rd order for the low pass to cut aggressivelly above around 3kHz.
Regarding the low frequencies, the mid-range had a better behavior, so a 2nd high pass was enough.
Each situation will require a specific slope.
Don't forget about phase response.Any signs or symptoms?