advertisement


Why did crossovers end up in speakers, rather than preamps?

marshanp

ellipsis addict
The JR149 thread got me thinking... it has always bothered me that amplifier designers go to the trouble of producing low distortion output across the whole frequency range, only to have a large part of their efforts thrown away/strangled by a crossover positioned after the power amplifier.

When most domestic speakers had a single full-range driver this obviously wasn't an issue... but when more than one driver became the norm in high end domestic systems, why did it not also become the norm to split the signal prior to power amplification and allocate each driver its own power amplifier - one designed to excel in the frequency/impedance range over which output would be demanded of it?

Economics may have played a role, obviously... but I'm talking about fine (maybe single manufacturer) systems aimed at those with deep pockets and a taste for demanding music, not run-of-the-mill systems.

I think I would have stuck with single driver speakers (Voight corner horns, for example) in that era, had I been sufficiently well-heeled to afford hi-fi of that calibre.
 
The crossover is determined by the speaker design, as you say, so on a practical level alone, to put them anywhere but in the speaker would create chaos and/or entirely bespoke amp'/speaker combinations.
 
Les, he of Avondale recommended I place my externalcrossovers as close as possible to my integrated ampbecause he thinks it sounds better than if they are next to my speakers, so I did which enabled me to connect my 2-way speakers using 4-core Van den Hul and Neutrik speakercon plugs which is neat. I believe low level detail and transparency improved but I wouldn’t argue with anyone.
 
Active crossovers were only affordable once the NE5532 appeared
So to implement a crossover at line level is only possible with sophisticated ICs? Whereas to implement it at speaker level is relatively simple and doable with discrete components? That would certainly explain it...

The electrical simplicity of the Voight corner horn driven by a valve amplifier still appeals... if only I had two corners available!
 
If active operation was the norm then every possible amp pack and speaker combination would have to be made together to work together so no box swapping or upgrading possible really. Also not possible to use manufacturer A's amps with Manufacturer B's speakers most of the time.

Also extra power amps cost more to make than speaker crossovers over most of the price range.

Can't see why the NE5532 or any other op amp is a game changer. XO's are just filter circuits and its entirely possible to make them out of passive components at line level to do exactly the same job as the generally rather larger passive components in the speaker XO.

One of the advantages of active filters is that you can actually make them do a better job at lower cost quite a lot of the time since large wound components and large capacitors can be eliminated.
 
Many active speakers have multiple amps with electronic crossover to do the job. Not as exciting as having an amp in the rack. And you still need a control or preamp anyway.

A general purpose crossover in your preamp would need to be fantastically complicated to cater for all loudspeaker types with variable crossover frequencies, variable slopes, what about correction circuits for frequency humps and dips in drivers. And then what happens when you buy a 3 way speaker......?

a fantastically flexible electronic crossover starts to look a like several parametric equalisers....and most hifi buffs have a pretty negative view about those.

In the modern age (now) you can go digital and do all the work there to feed your power amps. Such boxes are available to buy. You still need microphones and a lot of detailed information about the drivers in your speaker boxes though.
 
...In the modern age (now) you can go digital and do all the work there to feed your power amps. Such boxes are available to buy. You still need microphones and a lot of detailed information about the drivers in your speaker boxes though.
Is it possible to look at the components in a conventional crossover and "translate" them into parameters suitable for input to a digital crossover for use at line level?

It had occurred to me that if so, this could be a viable and cost-effective upgrade path - perhaps the only one available to people who have very good sources, but who cannot accommodate large speakers.
 
Yes and no.

You can easily enough figure out the type of filter and crossover frequencies, but "in speaker crossovers" actually deal with quite high currents which their active equivalents usually don't so the resistance of the coils and their interaction with the driver coils are actually important aspects of their operation. It usually isn't 100% obvious how driving a drive unit directly with a power amp will sound compared to driving it via a filter with some big wound components.
 
Can't see why the NE5532 or any other op amp is a game changer. XO's are just filter circuits and its entirely possible to make them out of passive components at line level to do exactly the same job as the generally rather larger passive components in the speaker XO.

One of the advantages of active filters is that you can actually make them do a better job at lower cost quite a lot of the time since large wound components and large capacitors can be eliminated.
In the 70s I would have buffered the filter to 600 Ohms and used a LCR design using Mullard pot core inductors. These were quite big and expensive. Not trivial to design either before computer simulation became common.
Far too many amplifier designers were only copying App Notes and would not have had a clue
 
That's interesting, and the reviewer was clearly impressed!

Having looked on the usual auction site, I suspect that chances of finding a Tri-channel system intact are close to zero... even though it's only a year older than me. A pity; it would make an interesting restoration project for somebody with Skills (not me, then).
 
In sq car audio systems, pretty much all systems are active, so a digital crossover and then all drivers individually connected to a speaker, compared to running passively the differences are enormous, the sound quality takes a leap! I have even used a studio rack mount 3 way active (passive) crossover from Rane modified for in car use - holy crap that sounded superb. Also a stadium EQ by Alto audio converted to 12v use.. again that was the nuts.. this is still in service in the Porsche with three class A amps...

I’m very surprised that active doesn’t get more attention at home, but I honestly think it’s because people would be blowing drivers all the time by using the wrong x-over points, speakers going back to manufacturers for new tweeters all the time... I don’t think they’d be over the moon about this, so to protect ‘their’ product a passive x-over is an insurance policy.

I have three way Tau’s atm and would love to try them active, I have the amps to do it, but don’t want to blow the unobtainable mids! I know what I’m doing with XO points, and won’t attempt it, all it takes is one wire or RCA wrong and it’s bye bye speaker.
 
The JR149 thread got me thinking... it has always bothered me that amplifier designers go to the trouble of producing low distortion output across the whole frequency range, only to have a large part of their efforts thrown away/strangled by a crossover positioned after the power amplifier.
.

Why do you assume that a crossover ‘strangles’ the signal? It’s more accurate to think of it as a device that sends the relevant part of the signal to the right place.
 
In a lot of budget (or even not so budget) designs that's certainly the case.

Yonks ago Wilmlow did in improved crossover for Kef Concertos which by comparison to the original was quite a lot bigger. They still sell it!

Prize for most efficient PXO design ever must go to Mr Epos for the ES14. One capacitor!

In fairness, Naim who certainly used to extoll the virtues of active systems, actually build quite good PXO's
 
Car audio is constrained by the 12 V power and a need for electrical efficiency, which favours an active approach
 
I recall Derek Hughes making external crossovers for the LS3/5a, called Circable.
Ken Kessler raved on about them, as he would.
I recall someone on PFM bought a pair and was very impressed with them.

They were £600 a pair though the Hi-Fi News Accessories Club.
You did you own work wiring them to your LS3/5as.
I think it was back in the year 2000, but I cannot remember exactly.
 
I favour doing the necessary filter(s) passively at line level. My speakers only require first order on the tweeters so doing it was easy. A polystyrene cap in the power amp input and job done. By using the resistor to ground that was already there I didn't run into any impedance issues.
 


advertisement


Back
Top