AVS Forum banner
1 - 20 of 32 Posts

·
Registered
Joined
·
9,202 Posts
Discussion Starter · #1 · (Edited)
I'm curious how many here run active versus passive crossovers on their speakers.

Active advantages:

1) Precision frequency division that is unaffected by speaker driver impedance. Which we know can vary as much as 300% or even more across the audio band.

2) The ability to easily use 24 or even 48db phase aligned filters that are very difficult to implement in a passive design - practically impossible.

3) With modern digital crossovers you can precisely time align the drivers and correct phase response. (well sort of!)

4) Allows you to tailor the amp technology to the driver. Example is a low power pure class A on the tweeters. Bass amp quality not as important and thus more options available.

5) Greatly reduces IM distortion within the amplifiers.

6) Reduces power demands on each amplifier as they don't have to amplify the entire audio bandwidth. Translates to much improved clipping performance.

Disadvantages:

1) Need a power amp per driver.

2) Interconnect wiring needs increase, especially expensive if you use high end name cables.

3) Cannot, IMO, set up a biamped or triamped system properly without analysis equipment, i.e. REW or the like. And of course the need to study the theory and setup process.

4) Especially difficult with mismatched amps as the amp gain must be accounted for in the balancing.

There is also the fact that a well designed passive crossover can easily trump a poorly implemented avtive system, in fact quite likely as a good active system is quite difficult to get balanced. But there is no doubt that a well tuned active system can provide clarity not possible with passive designs.

Thoughts on active crossovers in high performance systems?
 

·
Registered
Joined
·
904 Posts
I'm curious how many here run active versus passive crossovers on their speakers.

Active advantages:

1) Precision frequency division that is unaffected by speaker driver impedance. Which we know can vary as much as 300% or even more across the audio band.

2) The ability to easily use 24 or even 48db phase aligned filters that are very difficult to implement in a passive design - practically impossible.

3) With modern digital crossovers you can precisely time align the drivers and correct phase response. (well sort of!)

4) Allows you to tailor the amp technology to the driver. Example is a low power pure class A on the tweeters. Bass amp quality not as important and thus more options available.

5) Greatly reduces IM distortion within the amplifiers.

6) Reduces power demands on each amplifier as they don't have to amplify the entire audio bandwidth. Translates to much improved clipping performance.

Disadvantages:

1) Need a power amp per driver.

2) Interconnect wiring needs increase, especially expensive if you use high end name cables.

3) Cannot, IMO, set up a biamped or triamped system properly without analysis equipment, i.e. REW or the like. And of course the need to study the theory and setup process.

4) Especially difficult with mismatched amps as the amp gain must be accounted for in the balancing.

There is also the fact that a well designed passive crossover can easily trump a poorly implemented avtive system, in fact quite likely as a good active system is quite difficult to get balanced. But there is no doubt that a well tuned active system can provide clarity not possible with passive designs.

Thoughts on active crossovers in high performance systems?
I've run active in several systems with good results. I used a DEQX with some custom line arrays and Seaton subs, and more recently, PRO Audio Technology speakers where passive isn't an option.

With the DEQX it took a lot of work and many iterations(and a visit by Mark Seaton) to get things dialed in, but the end performance was materially better than the passive crossovers that came with the arrays (even with the incremental A/D and D/A).

I never had any luck mixing amps (unless you count the dedicated amps in subwoofers) as there was always an audible "discontinuity"

The incremental expense of running multiple amps, multiples of cabling, etc. isn't trivial.
 

·
Registered
Joined
·
904 Posts
The LT20 are fully Active 3 Way with the Filter Cards installed into the amplifier not the speaker.
That's similar to the Pro Audio Technology speakers I'm using. The total package (amps with built in crossovers and speakers) is very cost effective.

I should have said up above, going active can be very expensive, particularly if you end up buying "audiophile amps and cables" for every driver.
 

·
Registered
Joined
·
2,524 Posts
I started off with passive ATC 50s and switched the passive crossover out to the amp packs which have active crossovers/amps for each driver. Based on my experience I ended up buying all active ATCs for the remainder of my ht setup (Anniversary 100s and C6 centre).


The overall cohesiveness and dimensionality were improved with the active approach. Even though in theory the bass control should be better with active I had no complaints with the bass control and power using a Krell FPB600 amp on the ATC 50 passive. But overall the active approach was "better" and certainly more cost effective.
 

·
Registered
Joined
·
341 Posts
I definitely would add the biggest disadvantage of going with an Active crossover: COST!

On the advantage side, watching Curt set up my system was very instructive. I think being able to time-align the drivers very precisely in your room, as well as playing with the phase made for a spectacularly performing system. Going from the outboard analog active crossovers to the Trinnov Altitude solution made a remarkable improvement in the sonic performance.
 

·
Banned
Joined
·
17,607 Posts
Hans Zimmer prefers active Quested nearfields for multichannel , despite comparing to a very nice flatter liner passive crossover inside same speaker in the context of 5.1 the combined effect was cleaner on active.
 

·
Banned
Joined
·
260 Posts
Cost of the amplifier is the biggest issue, but if the setup is done right once it will provide decades of enjoyment. Comparing the Active to Passive for the past 2 years, I believe Active is the way to go for Home Theatre setup if you can afford it.


Quested active systems work by placing the Filter cards inside the amplifiers directly. On the LT20 Setup I have Each AP1550 Amplifier runs a Single Speaker (In mono block form) . Channel A for bass (Twin 10inch drivers) , Channel B for mids and tweeter.


I think Pro Audio uses Class D Amplifiers for there setup.
 

·
Registered
Joined
·
579 Posts
I love electronic crossovers :)... they're the easiest way to go if you like fooling around mixing and matching different loudspeaker components. A wacky old pro-speaker reconing service near Canal Street in NYC ("AST") got me hooked on them.

I've been using the Ashly Protea 4.24c for many years:




And here's one of the first models I used in the early '80s, and I still have it though the pots/switches are hopelessly worn/corroded now. [Pioneer D-23]

What a lovely sounding unit it was in it's prime.



Right now I'm pushing:
Subs: NHT 1259 [12"], 3ft³ sealed (8)
Midbass: McCauley 6222 [10"], 1.5ft³ vented (8)
Midrange: Dynaudio D-52AF
Highs: Scan-speak D-2905
Mostly Crown amps, configured for 240VAC.
 

·
Registered
Joined
·
1,743 Posts
For DIY, IMO fully active is the only way to go, especially if you're using an unproven driver array.

Buying commercial gear is another story, if someone has taken the time to work out a decent passive crossover, go it it and say the money on the extra channels.

.....I am now a DIY convert, so its active all the way for me. I currently use QSC DSP's and Amps:



For my two ways, plus the sub:

 

·
Registered
Joined
·
579 Posts
There are neat ways to cheat with the concept without buying the actual active crossover unit, maybe Glimmie has heard of them, by just connecting the passive filter elements between the amps outputs and the drivers. It worked okay the last time I [briefly] tried it and I think it was used by some old speaker company, maybe Bozak?

The other trick is to just series connect a passive capacitor to the amp input, but you can only get 6dB/octave. The cap's value also depends on the amp's input impedance, an interesting complication. Speaker Builder magazine had articles on this years ago.
 

·
Registered
Joined
·
9,202 Posts
Discussion Starter · #13 · (Edited)
There are neat ways to cheat with the concept without buying the actual active crossover unit, maybe Glimmie has heard of them, by just connecting the passive filter elements between the amps and the drivers. It worked okay the last time I [briefly] tried it and I think it was used by some old speaker company, maybe Bozak?

The other trick is to just series connect a passive capacitor to the amp input, but you can only get 6dB/octave. The cap's value also depends on the amp's input impedance, an interesting complication. Speaker Builder magazine had articles on this years ago.
Yes, you can make passive line level crossovers but in this day and age (actually since the 1970s) you can do much better with OPAMP active filters. IIRC, Altec made some passive line level units in the 1950s for theatrical sound systems that now go for outrageous prices on EBAY and Audiogon. Why I don't know because they are actually not that good compared to modern active designs. But it's the old legacy name association thing I guess. Just like 1920's Western Electric 43A theater amplifiers going for $30-40K.

You can buy analog and DSP based active crossovers for less than $400 from pro audio dealers. But this being the high end forum I would recommend the Marchand Electronics units. These are very good analog designs.

As for the digital side, I know the Trinnov can easily provide active crossover capabilities. Some of the other high end processors may be able to as well?
 

·
Registered
Joined
·
1,183 Posts
1) Need a power amp per driver pair.

2) Interconnect wiring needs increase, especially expensive if you use high end name cables.

3) Cannot, IMO, set up a biamped or triamped system properly without analysis equipment, i.e. REW or the like. And of course the need to study the theory and setup process.

4) Especially difficult with mismatched amps as the amp gain must be accounted for in the balancing.
1. One amp per pair or per four drivers or however many channels your amps have.
2. You'll RTA it so any db differences in power due to cables will be EQ'd away. Buy the cheapest.
3. Its even more difficult to figure out proper passive crossovers, with active crossover you can change it on the fly with RTA software looking at the results instantly. Testing where the crossover points should be is infinitely better than laying "the best laid plans". Also, if you haven't been able to try a hundred different passive crossovers, you can't really say if its a good one using the drivers to the fullest.
4. An hour of fiddling with the RTA software and some EQ and you have it figured out. Again, its much more difficult to do this with passive crossover. If I were to make a passive crossover my horns would pop at 1/17th the power level of the mids, barring a pre-planned "best laid plan" on what resistance to add in series to cut the power level the horns are given (if I feed the horns and mids the same power the horns are 12db above the mids). Timing differences in amps can also be dealt with, but most people have the sense to use equal timing amps.

I chose active crossovers for these reasons:
1. No need to pre-plan the perfect crossover points.
2. No need to get stuck in what I chose either, I can therefore find better combinations if I upgrade particular items (at first I'll use the 70hz-1khz drivers as lows since the crossover on the low output can reach up to 1khz, the crossover to the horns, then add dedicated subwoofers later).
3. Passive crossovers don't make sense at 1800w RMS (build thread to come as I get close to finishing). The heat alone would potentially be a problem. Let alone finding large enough parts. Substantial amounts of amplifier cost in just heating the passive crossovers would be lost.
4. The cost of the passive crossover would be greater than the cost of an active crossover. One active crossover can be as little as 62 USD (that's what I paid for my 3-way crossover for rack, used, including 21 USD shipping).
5. I already knew I wanted a proper 2x31 band EQ in the rack along with XLR and speakon cables, just so that I can hide it all in another room like I have with my PC (real Norwegian and Scandinavian thing, sleeper cars, sleeper PCs, sleeper stereos, all stuff that look like nothing out of the ordinary, but which can break your neck and make you deaf if you give it a little loud pedal).
6. filtering after the amplification doesn't make sense. Why filter hundreds of watts when you can filter a couple milliwatts?
7. Calculating a passive crossover and sourcing the parts to handle the power levels I'm using in my next build was too much work. Instead I bought a microphone with flat response and then I'm going to just use RTA software to figure out stuff for me (just look at the response, see where the best crossover point is based on testing ten different points).
 

·
Registered
Joined
·
9,202 Posts
Discussion Starter · #17 · (Edited)
1. One amp per pair or per four drivers or however many channels your amps have.
2. You'll RTA it so any db differences in power due to cables will be EQ'd away. Buy the cheapest.
3. Its even more difficult to figure out proper passive crossovers, with active crossover you can change it on the fly with RTA software looking at the results instantly. Testing where the crossover points should be is infinitely better than laying "the best laid plans". Also, if you haven't been able to try a hundred different passive crossovers, you can't really say if its a good one using the drivers to the fullest.
4. An hour of fiddling with the RTA software and some EQ and you have it figured out. Again, its much more difficult to do this with passive crossover. If I were to make a passive crossover my horns would pop at 1/17th the power level of the mids, barring a pre-planned "best laid plan" on what resistance to add in series to cut the power level the horns are given (if I feed the horns and mids the same power the horns are 12db above the mids). Timing differences in amps can also be dealt with, but most people have the sense to use equal timing amps.
1) Yes if stereo amps. However with monoblocks and strictly it's still an amplifier per driver.

2) I am by no means a supporter of expensive cables. But as this is the high end forum some here would not dare use commodity wires.

3) Agreed.

4) Read the paper over in ESP. It goes into why different amp gains must be compensated for or the crossover point will drift at different power levels. I don't know what you mean by timing differences in the amps. These are irrelevant at audio frequencies. The electrical delay is only going to vary between dissimilar amps a few microseconds at most. Now if you mean the time alignment of the speaker drivers, yes that should be compensated for.
 

·
Registered
Joined
·
1,183 Posts
1) Yes if stereo amps. However with monoblocks and strictly it's still an amplifier per driver.

2) I am by no means a supporter of expensive cables. But as this is the high end forum some here would not dare use commodity wires.

4) Read the paper over in ESP. It goes into why different amp gains must be compensated for or the crossover point will drift at different power levels. I don't know what you mean by timing differences in the amps. These are irrelevant at audio frequencies. The electrical delay is only going to vary between dissimilar amps a few microseconds at most. Now if you mean the time alignment of the speaker drivers, yes that should be compensated for.
2. Spend money where it has an effect. If I have low end equipment that don't have less effect than the really expensive stuff its still an high end system. You can buy "high end" electricity if you want. That's electricity with even less frequency and voltage variation than the national standard, but that wouldn't have an effect on the quality of the sound, so its not mentioned in this high end forum. No one goes "Oh well you're playing those 10 grand speakers on normal national standard electricity, that's not high end!". :rolleyes:

4. Timing differences can occur when amps do more or less digital work on the sound. Analogue amps due to the way electricity works, are instantaneous (well, speed of light). What the signal in is, is exactly what the signal is out less than a nanosecond later. doesn't matter how many things you line up in your rack of equipment if its all analogue. The signal still goes through at the speed of light. But once you introduce digital processing in the amp, then its no longer instant. It converts the analogue input, which takes time, then it records the digital information to a buffer memory, hands that over to a processor in one neat chunk per processing cycle, which takes time, then the processor hands the finished material onto another buffer, which then converts it back to analogue signal, which takes time, before amplifying it with an analogue amplifier, which takes no time (or amplifies with a class D/H amplifier without converting digital to analogue first, this can be done on the fly as fast as the signal is input from the buffer). The time between the input analogue signal and the output amplified signal can be as much as 20ms quite easily, if particularly unlucky with the particular processor used on the amplifier (you rarely will know what processor is used, its not among the stated specs anywhere I know of). It can happen with pro-grade amps with digital signal processing as well, but long story short this is the reason lab gruppen amps are so expensive, they need proper DSP.
Mixing and matching amps with different DSP could lead to different timing, also mixing and matching amps with DSP with amps without DSP.
I will read what you proposed I read.
 

·
Registered
Joined
·
9,202 Posts
Discussion Starter · #19 ·
4. Timing differences can occur when amps do more or less digital work on the sound. ......

Mixing and matching amps with different DSP could lead to different timing, also mixing and matching amps with DSP with amps without DSP.
I will read what you proposed I read.
Well yes, If your amps have digital processing built in, you can have significant delays. Come to think of it QSC had DSP packs that attached to the back of their amps. So in this case you did have to use matched amps.
 
1 - 20 of 32 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top