Driver mounting - Front or rear of baffle - Benefits?

[synsei]

I note that some speaker manufacturers, for example Harbeth and Spendor, choose to secure their woofers to the interior surface of the baffle on some of their models. From a purely aesthetic point of view I have always considered this method of mounting drivers to look a tad untidy, however are there benefits to doing so?

[istari_knight]

Who better to answer than the designer ? :)

http://www.harbeth.co.uk/usergroup/showthread.php?958-Woofer-driver-positioning-on-baffle-in-front-or-behind-baffle

http://www.harbeth.co.uk/uk/index.php?section=products&page=designersnotebookdetail&id=18

[synsei]

Who better to answer than the designer ? :)

http://www.harbeth.co.uk/usergroup/showthread.php?958-Woofer-driver-positioning-on-baffle-in-front-or-behind-baffle

http://www.harbeth.co.uk/uk/index.php?section=products&page=designersnotebookdetail&id=18

Thanks for posting the links James, it's nice to see that a designer of what is a premium product range is so accessible ;)

My worry is that mounting the bass driver on the inside surface of the baffle might lead to some minor colouration induced by the edges of the aperture cut into the baffle, re: the thickness of the baffle material. I hope am am coming across clearly :lol:

[Reffc]

Hi Dave

the reason Harbeth and others have to do this is due to the higher order crossovers they employ giving rise to phasing issues. As AH points out, a remedy is shifting the driver position to correct the phase issues. Its more an issue with higher order x-overs and isn't an issue with say 2nd order crossovers where drivers are designed for flush mounting (2nd order = 180 degrees phase between tweeter/woofer so you just connect one of them up with reverse polarity). The advantages of going higher order are that (as long as time alignment is compensated for) off axis results can be better and steeper slopes can be employed. The big problem with higher order passive networks is their (necessary) complexity. Some people also don't like the way that 3rd/4th order networks can start to make speakers a little shut-in but if done properly, this shouldn't be the case. Anything more than 3rd order really belongs within an active filter network due to the complexities involved.

Baffle design and driver mounting are a whole design issue in themselves, structurally speaking!

[synsei]

Cheers Paul ;)

So I understand things a little better could I ask someone to explain the various 'orders' of crossover please? I'm sorry but when I read technical stuff I tend to skip over the bits I don't understand whilst trying to make sense of it from what I do understand. I'm stumped in this case though :eyebrows:

[Yomanze]

Cheers Paul ;)

So I understand things a little better could I ask someone to explain the various 'orders' of crossover please? I'm sorry but I when I read technical stuff I tend to skip over the bits I don't understand whilst trying to make sense of it from what I do understand. I'm stumped in this case though :eyebrows:

That's a very big question...! Have you tried this? Lots of diagrams & nice descriptive language in amongst the technical bits. :) http://www.rane.com/note160.html

[synsei]

That's a very big question...! Have you tried this? Lots of diagrams & nice descriptive language in amongst the technical bits. :) http://www.rane.com/note160.html

Thanks Neil but I think that's one for later when I have more time, a Brandy by my side and I am in a more relaxed mood... :lol:

[Reffc]

It's a fairly complex area David as although you can calculate the values needed for various orders, the driver electrical and mechanical characteristics and the cabinet design can all influence crossover design. In a nutshell, a crossover is required for anything but a full range driver to separate the various frequencies to the drive units best able to handle that particular frequency range.

In the case of a two way driver example, a first order crossover has one component in the bass (an inductor) and one in the treble circuit (a capacitor) although it can have more (padding resistors) if it needs attenuating to bring output down to match the bass.

Inductors' reactive impedance rises with frequency so the value you chose for a crossover is chosen to be equivalent to an open circuit, or open loop therefore passing no signal beyond a certain frequency.

Capacitors' reactive impedance drops with frequency, so it requires a certain frequency for any signal to get through and as frequency rises, a value is chosen to "roll on" a tweeter (say) at a particular frequency.

In the case of a first order crossover, the roll-off of the inductor and roll-on of the tweeter happens at 6dB/Octave (an octave is the interval between two particular frequencies at half or double that particular frequency). 6dB/octave is a fairly shallow slope and there are some drivers where the gap between the bass rlling off and tweeter rolling on forces the bass into distortion or the tweeter into distortion, so a steeper slope is required.

A 2nd order crossover provides a steeper slope (12dB/Octave) and is the most commonly used passive crossover network as it tends to be the best compromise between desired slope and complexity of higher order crossovers, particularly in respect of phasing and time alignment (more later on that).

2nd order,as the same suggests, have two components in the network for each frequency range: the bass has a series inductor on the positive input line and the a capacitor shorted to negative (or signal return). The tweeter has a series capacitor on the positive input and an inductor shorted to negative (signal return).

Getting back to phase: the result of introducing a first order crossover is to introduce 90 degrees of phase shift. Consider drivers in phase as pushing together and pulling together. 180 degrees out of phase and one will be on axis whilst the other is at maximum extension. Its not often than any time delay circuit is introduced for first order crossovers although some manufacturers shift driver voice coil (driver) position to introduce a slight time delay to one unit.

2nd order crossovers have the drivers at 180 degrees out of phase so that the polarity of one driver is simply reversed to bring both exactly back into phase. This, and the 12d/B per octave slope is arguably the best compromise where passive crossovers are concerned although that does depend on the design and off axis performance of the drive units. 3rd order crossovers can give better off axis response if Butterworth arrangements are used and symmetrical "MTM" driver arrangements are used but using a Linkwitz Riley 2nd order also gives very good off axis acoustic radiation patterns and response is flat across the crossover frequency.

The main advantages of 3rd order slopes is that they roll on/off at 18dB/Octave so offer lower distortion (depending on drivers selected) and better tweeter protection from lower frequency (higher power) signals. These days, there are many tweeters that can drop very low, down to 1500Hz for example, and the advantage of using a tweeter this low is that it usually has better off axis response and lower distortion than asking a bass unit to deliver at the upper end of its frequency spectrum (as mid/bass units can get "beamy" off axis).

So crossovers exist to ensure that the correct frequencies get to the correct drivers, that protection is given to the tweeter and to manage driver distortion and acoustic radiation patterns, depending upon design.

There are various types of crossover such as Linkwitz-Riley, Butterworth, Bessel etc and these are named after those people who came up with their characteristics and design. Even order crossovers are usually best in Linkwitz Riley and odd order using Butterworth. The type of crossover is a complex subject and to keep it simple, the above is probably all you need to understand the basics.

There are other circuit considerations such as bass circuit Zobel networks and high frequency padding resistor networks (L-Pads) but these, again, don't need to be explained for the basic understanding of what a crossover does.

[YNWaN]

In an ideal world a drive unit should produce an absolutely straight line output and then instantly drop to zero as it passes over to the next drive unit, that should also have zero output beyond it's usable range and be at exactly the same level as the other driver. In the real world drive units do not behave in this manner at all. Ist and 2nd order crossovers mean that at the crossover point both drive units are producing the signal and continue to do so both above and below the crossover frequency, gently dropping away. 3rd and 4th order crossovers have increasingly steep crossover slopes, which is good in the sense there is less overlap, but bad in the sense that the join between the drive unit outputs must be more seamless (which drive units rarely allow) and this can lead to the sound changing character as it hands from one driver to another. Higher order crossovers also need quite a few more components which is why 4th order crosovers are often active and placed before the power amps.

[synsei]

Until now my understanding of crossovers has been limited to knowing that crossovers split the signal between the various drivers and that they feed them with the correct frequency range for that driver. I am a little wiser now, thanks Paul :)