Why do you need a speaker crossover - discuss

[sq225917]

I like arriving late in the day to a topic it means you bypass all the bullshit and interpersonal rivalry.

To answer Richards initial question.

Some loudspeaker designers, such as the previously mentioned Robin Marshall design the whole speaker, they are electrical, mechanical and acoustical engineers of the highest order and are able to design drivers, and load them correctly, acoustically, electrically and mechanically in such a way that they require no 'additional crossover parts' to allow the drivers to align in a sonically unobtrusive way at the crossover point. They do their math, calculate the correct slopes and everything works. I doubt you'd luck into this level of accuracy empirically without the help of measurement and calculations, it seems unlikely.

Robin isn't the only guy to take this approach REF 3a also make their midbass drivers this way, though they use simple slopes in their 3-ways to roll off into the bass driver. nevertheless it's the same approach on a grander scale. The point is that they make their own midbass units. Do you see the pattern yet.

Now Joe Bloggs speaker 'manufacturer' might not have the all round engineering prowess to design and make his own drive units. He's pretty much stuck with what he can get from the off-the-peg shelf at Tymphany etc, or what he can have slightly modded from the next shelf down. (Or more likely the case these days from the shelf marked 'custom' which is actually the off-the-peg shelf for liars.)

He's kinda screwed, he has to do something to make his driver choice meet in the middle, assuming he's not picked two drivers that were designed to work together by the manufacturer into a specific volume. So he has to build an ugly electrical crossover to tailor the X-O point and the slopes either side of this point.

That's the rub. You either use the drivers you can get, and stick them in a different box to everyone else and then cobble together a X-O to make them work as best as you can, or you design everything from scratch.

The third way would be to take two drive units that you like the look/sound off. Stick them in a box and then shape their responses by ear by acoustic or physical loading. I assume this is Richard's preferred approach to avoiding the evils of the crossover and it's as valid as any other approach.

[Dr Bunsen Honeydew]

It depends on your reference. If you like classical or acoustic music, it's easy to compare. Anything electronic (jazz, rock etc.) and you're into a whole world of preferences and lack of baseline.

I would add female vocal and especially piano and I will completely agree with you.

The taste for only rock and pop has spawned the home PA system approach that was the flat earth.

[Marco]

I do not have absolutism or simplistic thinking, I have experience and opinion, nothing more.

Yes and we respect your opinions and experience, Richard (it is precisely why you are here), but there are only a finite amount of times you can mention the same things before they become stale.

And that is the issue here, not that we don't rate your opinions or consider them valid.

So, moving on now, I'd appreciate your views on the question Anthony has just posed?

Marco.

[StanleyB]

Sometimes simplistic thinking beats experience and opinions. I have used that principle many times, and with great results.

[Dr Bunsen Honeydew]

Hi All,
ok, lets take this debate a little further and discuss what we are actualy doing when we use either cross-overs [be it mechanical or passive/electrical]
What are the goals we wish to acheive, and what affect might this have on the performance of the speaker/driver and cabinet as a whole ?
Anthony,TD...

How long have you been designing valve amps - it seems you have loads of experience. Without that experience would you produce what you do now, I doubt it, to actually "luck" onto something good is like winning the lottery.

Well for me it is exactly the same, you try things, you have ideas, some work some don't, you listen and you choose. It is that ability (or not) that you rise or fall on. That is the empirical approach to design. My pet hate is the white coat brigade, who see the white coat as a qualification and no one else is in their club. They are brainwashed into thinking that you can create something by purely applying objective criteria, they don't listen. They calculate objectively, they make prototype, they put them in an anechoic chamber and play test tones through them, and read off B&K graphs and say that's it and sign off the design to production. The outcome is more often than not a pigs ear.

Or you get the approach where you limit the design and aim for a specific set of criteria. The opposites in this approach I see as Linn speakers at one end and BBC at the other. They listen, but they have fixed criteria and narrow the potential. It was entirely because it was the BBC design that held the market when Linn first appeared that they took an opposite approach. It was perceived in the early 70's that BBC design was not good at rock and pop. PA systems were making an impact at live concerts (to the extent that you are no longer listening to the band you listen to the PA). Linn persuaded a PA amp manufacturer to join them in the marketing take over they entered into, and the home PA system was born. I personally don't like either approach, but given the choice I would prefer the Linn approach. Yet there were bloody good all round speakers being produced on the east coast of the USA at the time by Allison and Snell that were virtually ignored because of this nationalistic *EGO* thing that the Brits get involved in.

So going into specifics of this cabinet this, this driver that, just applies to that specific product / compromise. The real real question is conceptual.

[anthonyTD]

Richard,
I understand and agree with a lot of what has been written in your posts.
However, What i was trying to get at was, what are we physicaly or electricaly doing to a specific driver and cabinet combination to get the desired affect we require. then once we have established that, we can then descuss why one method might be more desireable than another, i think this may have been what was on your mind when you started this thread, please excuse my ramblings if that is not the case.
Anthony,TD...

[Paul Hynes]

It’s nice to see such interest in the line arrays, however, I do not wish to take Richard’s thread off topic with further discussion about them, so I am going to start another line array specific thread, either later today or tomorrow when I get a little spare time. I hope those who have shown interest here will also monitor the new thread and join in with their own ideas. I will answer all your line array posts, since my last post in this thread, on the new thread and restrict any further comments regarding this thread topic to the thread theme.

Paul

[Dr Bunsen Honeydew]

Richard,
I understand and agree with a lot of what has been written in your posts.
However, What i was trying to get at was, what are we physicaly or electricaly doing to a specific driver and cabinet combination to get the desired affect we require. than once we have established that, we can then descuss why one method might be more desireable than another, i think this may have been what was on your mind when you started this thread, please excuse my ramblings if that is not the case.
Anthony,TD...

Well when I started the thread I was thinking conceptually not specifically. But of course specific is important. All I can do is talk about my specific design. I see two ways to approach speaker design, first would be to start with the basic drivers and cab and then add thing as in mechanical damping and electronic filters to either get what reads flat anechoically or sounds the way you want. I use the listening approach, but once I get to that point I reverse it and start removing things or substituting to fine tune. I have a agenda to remove anything and everything I can from the signal path and still work. I don't see mechanical or acoustical properties as in the signal path. Signal path for me is an electrical term, mechanical things sit *on* the musical path, if that makes any sense. And very often especially if primarily having a damping effect they actually sit on the music and cover it.

Anything that interrupts the connection between the amplifier output stage and the speaker voice coil is decoupling the two from each other, and that can *only* interfere with the music by changing it. It is the compromise of what is needed to remove the imperfection of the object and getting too much of it which then sits on the music. It is just another yin - yang dichotomy. The one with too much is dynamically flat and bland but can be said to be accurate tonally, the other can make you wince but can't stop you wanting to bop or wag you feet I want both but you have to make some compromises, you just constantly try to find ways to remove those compromises.

[Marco]

It’s nice to see such interest in the line arrays, however, I do not wish to take Richard’s thread off topic with further discussion about them, so I am going to start another line array specific thread, either later today or tomorrow when I get a little spare time. I hope those who have shown interest here will also monitor the new thread and join in with their own ideas. I will answer all your line array posts, since my last post in this thread, on the new thread and restrict any further comments regarding this thread topic to the thread theme.

That's great, Paul, and much appreciated. It's that sort of professional, unselfish, attitude which epitomises what AoS is all about. You are a credit to this community, mate!

Marco.

[Dr Bunsen Honeydew]

It’s nice to see such interest in the line arrays, however, I do not wish to take Richard’s thread off topic with further discussion about them, so I am going to start another line array specific thread, either later today or tomorrow when I get a little spare time. I hope those who have shown interest here will also monitor the new thread and join in with their own ideas. I will answer all your line array posts, since my last post in this thread, on the new thread and restrict any further comments regarding this thread topic to the thread theme.

Paul

But stay here as well Paul as I for one would welcome your input on the topic.