Active Crossovers

[Rudy81]

Ever since I listened to some Jubilees with active crossovers, I have been trying to learn about this process. I have been toying with the idea of actively filtering my Khorns just to evaluate the differences between active and passive crossovers. I currently have the ALK universals with JBL 2470 drivers on ALK Trachorns and APT 150 tweeters.

I have looked at the EV DX38 which seems to be a favorite processor around here, but for three way filtering it seems a cost prohibitive experiment. I have been studying the EV DC-ONE which is 2 in and 6 out. I am also studying amplification for the project. Since I am rather ignorant on the subject, I am here to ask the Oracle's for help.

Has anyone tried the EV DC-ONE and will it work for this application?
What amps would you guys recommend for the project?
Do any of you use capacitors to protect your drivers from the amps? (Read that one somewhere)
Any tips on how to best do this?
Do you think it is worth the effort, particularly if you have already done this.
Does time alignment make a big difference? (Not trying to start a war, just trying to educate myself)

I'd love input on this idea, pro and cons etc. I'm starting with my Khorns since I know how they sound with passive networks. Eventually, I may go Jubilee or Jamboree with the process.

EDIT: This thread has become rather long, but has a tremendous amount of information on active filtering for the novice. It has occassionally gone off into other tangents when I needed help with side issues so I will come back now and then and create a sort of index so you don't have to get bogged down with other issues.

Begin with next post for active crossover discussion.

Go here for discussion on protecting drivers by using capacitors between the driver and the amp : http://forums.klipsch.com/forums/t/128607.aspx?PageIndex=10

Skip here if you are having problems with hum, hiss, noise or cable issues with an active filter: http://forums.klipsch.com/forums/t/128607.aspx?PageIndex=13

[Coytee]

For the sake of you fooling around, I've got an idea for you.

You're right in that the Dx38 won't be of much help to you for a stereo pair of 3-way speakers. It's a 2 in, 4 out unit. However, let's not forget that it's indeed, a 2 in, 4 out unit!

Nola in New Orleans has my second Dx38 as he's been curious as to what some of the fuss might be about. I don't know how long he'll have it as he received it just last week and has probably, yet to take it out of the box.

Here's my point...(offer)...

When he's done he is currently slated to ship it back to me and all I'll do is put it into a closet for a couple months until I'm ready for it downstairs.

If instead, you'd like me to have him ship it to you, then after he's done, you'd be welcome to play around with it so you would have essentially, no cost in your playing around. It might answer some questions for you and it might create others.

It's a 4 out so what you could do is take your (for example) "right" input in to it...you'd have two "right" outputs. Next to the inputs are some parallel outputs so you could take the #1 output (right) and redirect it to #2 input which would give you a 2-in, FOUR out as long as you are fooling around with a single channel.

This would also allow you to do a mono left with a mono right, using any configuration you'd want to use.

You can find these on Ebay for about $600/750 each in the used market. $750 seems to be about the right price for decent units. I did however, luck into some ebay theft (snicker) and bought two of them for $500 each and a third unit (this one) for $400!!

So, their price isn't as horrendous as it may seem however, this unit clearly won't fit your needs unless you obtained two of them.

So... if you'd like to have it sent to you after Nola is done, feel free to let me know and we'll make that happen. Realize though, I told him that I had ZERO need for it back quickly and as far as I was concerned, he could have use of it for several weeks without me even thinking about it. If you want to play around relatively soon, I can't say he'd be done by then.

Legit offer though.

As for your questions:

Which amps? I'd say what ever amps you like. You can mix/match tubes, solid state how ever you want. If your crossover unit has XLR outputs you might find some benefit in having amps with XLR inputs but you can also get around that with adapters. (I liked the McIntosh unit I had since it had both inputs)

I don't have any capaciters between drivers/amps, not even a fuse

how to do it? yeah... open your checkbook, buy some Jubilees, buy some nice tube amps for the top half, maybe a beefy solid state for some meat and you're ready to go.

Is it worth the effort, particularly if you have already done this (what is "this" biamping? active?) I went from passives/Khorns to active/Jubilees. In my opinion, the difference in sound is easily worth it. (total package, maybe passive/Khorn to active/Khorn might not)

Alignment: Actually, I think I'll amend my above comment. I personally think the signal alignment (how does one align time []) Anyways, I think the signal alignment is one of the things that helps account for the noticable improvement in sound. I'd personally speculate that a signal aligned, Khorn (which means biamp K400/K77 and K33, or triamped) would be a dramatic improvement.

As I recall, one of the biggest shortcomings of the Khorn is the disparity of signal alignment. In my situation, I had to be about 30' away from my Khorns (ALK ES networks) before I suddently had an "OMG" moment where the sound was distinctly better than being closer to them. With the Jubilees, I can be 10' away from them and have the same sound quality it took 30' for my Khorns. Yes, I'd expect this to be a great improvement.

Which actually, brought up another thought... if your ALK crossovers will allow...you might be able to use the Dx38 in stereo even though it's a 4 out. You might be able to keep the K400/K77 on one channel and the woofer on the other channel. Using the Dx to shove the HF section back several snippets in time will probably garner more improvement than worrying about aligning the tweeter/midrange (or so I speculate)

If I were you, I'd nab my Dx for a while, after Brian is done. I'd try a side by side, using the Dx in 3 way mode while playing in mono. I'd then reconfigure things, putting the K400 back with the K77 and then try it in stereo, allowing the ES5800 (or what ever you have for the HF crossover) deal with the mid to tweeter exchange, using the Dx to manage the signal exchange AND delay from the woofer to the HF section.

[Rudy81]

See, this is why this forum is so great. Everyone trying to help. Coytee....many thanks, and I'll take you up on your offer. I have another local member who has offered to lend me a spare DX38, so I can do both speakers. I'm going to start looking around for a six or seven channel amp at a reasonable price for the experiment.

I will take great care of your gear. I may have a thousand questions or so as I work on this. Frankly, I have my Khorns sounding sooooo good with the ALK universal, the Trachorns, JBL drivers and the Eminence tweeters. However, the more I learn about active crossovers, the more I want to try them and capitalize on the potential improvements. Such as time alignment and a steeper slope setup. I at least want to try it. If I like it, my next step will be to add a Jub or Jamboree bass bin to the mix and see what results I can get. The beauty of the active crossover is that I can change the parameters at will to suit whatever drivers and bass bins I'm using. I have slowly learned of these benefits.

I'm still not sure this is the way to go, but am willing to learn something new.

Thank you again for the offer, and no rush at all. I first have to learn a whole lot more. Let me know and I'll PM you my address.

[Mike Dubay]

Rudy,

Currently I'm working toward active biamping of my cornscalas. For an intial experiment I'm going passive biamping, using a solid state Adcom for the woofer duties, and a SE KT-88 tube amp for the squawker/tweeter. Using the basic AA crossover between the tweeter and squawker. I have listened to Khorns with the SE tube amp and duly impressed with the mid range clarity, along with the nice tight bass. My cornscalas just don't have enough sensitivity to go SE for the woofer.

To me solid state for the woofer and tube SE for the mid and tweeter is at least worth the experiment. It should give me the best of both worlds, tight bass, and the smoothness in the midrange.

If this does not work out, then back to passive crossovers and a full range tube amp.

Any other recommendations on active crossovers? I'm looking at the DBX234 XL, 2 in 4 out.

[Rudy81]

I'm at the bottom of the mountain looking up, trying to figure out how to start up the hill. However, I am considering the DX38 or the DC-ONE from EV. The DX38 will require two units to fully control three drivers in stereo. The DC-One, seems to to have 2 in and 6 out, which is ideal for a three way application. I'm reading the manuals as I type this.

[Rudy81]

For those of you who have the DX38, can you briefly explain how one sets the mid freq. bandpass crossover setting? I have been reading the manual and am not clear on how it would be set up in a 3-way setup. I easily see how to set the LF setting as well as the HF setting. However, how are the mid points set? I can see that there is a factory option for 3-way plus full range, but details are not sufficient in the manual for understanding. Drinking from this fire hose is tough!

[DrWho]

I would avoid the dbx units...the GUI is very akward and their cheaper units don't sound very good at all (even for PA use). Once you get into the price range where they start sounding decent, there are other better options available.

One thing to be aware of when going with different amplifiers for the LF and HF is that different amps have different gain and phase responses. You will need to account for this in the gain structure and xover settings.

I've heard that the Ashley Protea processor sound real nice. They have two versions, one with 4in 8out and the other with 3in 6out. Looks like 6 EQ's per input and 4 per output? Might be kinda lacking in the flexibility department.

EV has a new DC One that looks interesting, 2in 6out and I believe the same filter options as the Dx38. The thing I find interesting there is the AES/EBU input...which would allow one to keep the signal digital until after the DSP, which would be a much better signal chain for signals starting out digital. Apparently one can connect SPDIF to AES/EBU too:
http://tcsupport.custhelp.com/cgi-bin/tcsupport.cfg/php/enduser/std_adp.php?p_faqid=1305&p_created=1067521162&p_topview=1

The DEQX PDC-2.6 HD is supposed to be pretty killer too and actually intended for the audiophile market (with the audiophile pricetag of course: about $2800 USD). I wonder how much manual control is given to the end user, although I'm sure many would find the automatic features quite interesting.

Anyways, just throwing out some ideas for those that don't wanna go the Dx38 route. The Dx38 is actually the best sounding active xover I've run across.

[DrWho]

For those of you who have the DX38, can you briefly explain how one sets the mid freq. bandpass crossover setting? I have been reading the manual and am not clear on how it would be set up in a 3-way setup. I easily see how to set the LF setting as well as the HF setting. However, how are the mid points set? I can see that there is a factory option for 3-way plus full range, but details are not sufficient in the manual for understanding. Drinking from this fire hose is tough!

When going 3-way, you're going to have a lowpass and a highpass for the MF. The LF just has the one lowpass and the HF has just the one highpass. Set the lowpass for the MF to be the same type, frequency, and slope as the highpass for the HF. Set the highpass for the MF to be the same as the lowpass for the LF.

[Rudy81]

Dr.Who. Thank you. Seems simple enough. I have been studying the manual on the EV DC-ONE and it seems to be very versatile when it comes to a stereo 3-way application. I agree that the less you mess with the signal, the better.

My current idea is to use a DX38 or DC-One with a Parasound six channel amplifier. This solves the issues of different amplifiers for the novice. I'm trying to keep the learning curve as manageable as possible. I am also planning on some protective capacitors in the tweeter and mid driver lines to prevent the potential problem.

From the perspective of an 'experiment', I think this would at least teach me the basics of active filtering, allow me to play with time alignment, and be rasonably inexpensive.

Does this all seem reasonable, or am I missing the whole point?

[Coytee]

The DX38 will require two units to fully control three drivers in stereo

Although your larger scale thinking is right on, you could probably cheat a little. Example, I once set an industrial Heresy on top of my Jubilee bass bins. I wanted to add some punch to them!

I "biamped" the Heresy's putting the LF signal to the Jubilee bass bin, crossed at 80hz, put the HF signal to the Heresy and let the passive inside the Heresy (a 3 way) take it from there. I ultimately had a 4 way speaker using the Dx which is designed for 2-way.

So, if you were able to put a passive between your midrange and tweeter, you could in essentially, biamp the 3-way speaker using the 2 outputs. One for the woofer per se' and the other output going to a passive which splits the mids/highs.

Just food for thought.

[DrWho]

Sounds like you've got your bases covered.

The only other
thing I would add to the fire hose would be the ability to measure what
you're doing. Just getting good acoustic summation through the xover
passband is extremely difficult to do by ear. Going the extra mile and
EQ'ing by ear is even more difficult. I've always felt the real power
of these speaker processing units is the ability to easily match the
shape of the filters with the frequency abberations of the
speaker...which allows one to correct the phase response along with the
magnitude at the same time. When you get it just right, everything
clicks into place...I dunno how else to describe it.

I don't bring it up to discourage anyone, but I've personally
never been able to get to that click point doing it just by ear. On the
flip side, dialing in a system while measuring is still extremely
listening intensive. The measurements just help you get there easier.

One
of the nice things about the Jubilee is that Roy has already done all
the measuring for us, so it's just a matter of plopping in the right
values. I would love to see that become standard practice for all
speakers and then moving to actives wouldn't be so painful.

Btw,
many of the klipsch passive xovers have passive equalizers built into
them as well. Ideally that would be replicated in the active xover...

And just a quick tip...caps on the MF and HF are a great idea. You can take it one step further and only hook up the woofers first to make sure you don't hook a tweeter to the LF output. You can even hook the woofers to the MF and HF output to make sure they don't have any LF in them too. There are so many opportunities to hook things up incorrectly that I always go through these steps.

[Rudy81]

Coytee: Good point and certainly to be entertained while running the experiment. BTW, I have been thinking about the audio chain in active crossover and have another question. If I use a lossless digital source, send it to my pre/pro which converts it to analog, which then sends it to the DX38 or other crossover only to be converted back to digital for processing then back to analog sent to the amps. What 'losses' are involved in that entire process? It would seem to me that this back and forth conversion process can't be good and at best you would only 'break even' with quality components. Any experience or opinions on this? Man, this is getting complicated fast.

My brain is beginning to hurt.

How much of a difference would you expect to hear between a good passive crossover, like the ALK, and an active setup? I am beginning to wonder if the effort and expense will really produce worthwhile results? My whole goal was to experience the 'time alignment' issue. The more I look into this, more I see both the potential pros and cons. No free lunch as usual.

[Rudy81]

Sounds like you've got your bases covered.

The only other thing I would add to the fire hose would be the ability to measure what you're doing. Just getting good acoustic summation through the xover passband is extremely difficult to do by ear. Going the extra mile and EQ'ing by ear is even more difficult. I've always felt the real power of these speaker processing units is the ability to easily match the shape of the filters with the frequency abberations of the speaker...which allows one to correct the phase response along with the magnitude at the same time. When you get it just right, everything clicks into place...I dunno how else to describe it.

I don't bring it up to discourage anyone, but I've personally never been able to get to that click point doing it just by ear. On the flip side, dialing in a system while measuring is still extremely listening intensive. The measurements just help you get there easier.

One of the nice things about the Jubilee is that Roy has already done all the measuring for us, so it's just a matter of plopping in the right values. I would love to see that become standard practice for all speakers and then moving to actives wouldn't be so painful.

Btw, many of the klipsch passive xovers have passive equalizers built into them as well. Ideally that would be replicated in the active xover...

I absolutely agree that measuring is way more productive than just listening....at least for me. I have difficulty 'remembering' what one setup sounded like vs. a change I made half an hour earlier. I currently use REW to take room measuremets. The other issue is that my pre/pro has Audyssey built in and will attempt to compensate for major room anomalies.

I am beginning to suspect that the differences are not a night and day issue, but more of an incremental difference. It will be interesting just to play with it to see just what I can do with such a system. I need to win the lotto so I can have more time to play with this stuff.

[DrWho]

If I use a lossless digital source, send it to my pre/pro which converts it to analog, which then sends it to the DX38 or other crossover only to be converted back to digital for processing then back to analog sent to the amps. What 'losses' are involved in that entire process? It would seem to me that this back and forth conversion process can't be good and at best you would only 'break even' with quality components.

That's why proper gain structure is so important...the artifacts of moving in and out of the digital domain typically show up in the noise floor. If you can keep the digital noise below the noise floor of your room, then the listening studies indicate that you likely won't be able to tell that the extra conversions were taking place. If the signal stays digital into the processor, then the signal path stays the same...just that there are more D/A conversions happening in parallel.

I personally find that the tradeoffs outweigh the artifacts generated by passive xovers. Having the amplifier directly coupled to the driver offers better electrical damping, and then you don't have issues with inductor hysteresis creating distortion and shifting the frequency response around with the music.

Always gotta pick your compromises.

For what it's worth, I would worry more about introducing a DSP stage when starting with analog sources than if I were starting with digital sources.

[ben.]

y'know, it might not be the worst idea to start with an analog active crossover when first getting into this stuff. Getting the idea of properly multi-amping is much easier with them and that would avoid the additional mind-bending of learning to use a relatively powerful pro audio device.

Pass over the entry level Chinese stuff - they won't sound very good for hifi. Something like Ashly's XR series (I think - their step-up series) sounds much better. If you are in a decent sized town, you could probably rent one from a PA house for very little money to see what you think.

[Coytee]

There are so many opportunities to hook things up incorrectly

(Mr. T voice)

Pity the fool who solders up his XLR wires backward and plugs them in out of phase with the other channel....[]

[Coytee]

How much of a difference would you expect to hear between a good passive crossover, like the ALK, and an active setup?

To specifically answer your question with a non-answer, I notice you have a LaScala? Do you have two of them?

I have a pair of Al's ESN networks for LaScalas as well, ESN 600T as I recall? Mated of course, with his ES5800. I happen to have three LaScalas so I have three of these. One of them (currently on ebay with evidently no interest at my asking price) has never been installed and is sitting downstairs.

I would be happy to send it to you so you could have this as well to play around with. If you have two LaScalas you could go fully active on one of them and fully passive on the other.

Now...if you were reading between the lines, you noticed I said I had it on ebay! No disrespect to Al but I'm slowly evolving over to the dark (active) side.

I'm planning on a center channel and am planning on taking my third LaScala and converting it to a JubeScala with the K510 getting stuffed inside the top section. I could use a passive for this however, because of my feelings now on some of this signal alignment stuff that I've experienced in my own home with my old non-aligned Khorns and now my aligned Jubilees.... I personally think that focus of sound is worth the tradeoff that might be there.

If you can take your current LaScala outdoors, mono, put some music on and simply start walking backwards in a direct line of sight.... there will come a place where you will suddenly notice a more coherent sound from your speaker. I'd suggest this is the distance that it takes for all three drivers to finally meld together into a single wave of sound.

If you then put an active on there and align the signals, I'd contend that you will find this same sweet spot much closer to the speaker. In my case with the Jubilees, they sound as good while being 10' away from them as the Khorns did while having to be 30' away from them.

Right now, I have no plans to take "my" LaScalas active. I bought them new in 1979 and even with the improvement in sound of the JubeScala, could NOT bring myself to hack into them to convert them. I would have to sell them and buy 2 more LaScalas and hack into those that were not my companions for the last 30 years. Funny how we get attached to things. My third LaScala was obtained as an empty cabinet and I don't care how much (bad) I hack it up.

[Rudy81]

Coytee: Once again, thank you for the offer. I only have one La Scala, which is my center channel. But, I understand your idea. However, you mention that I should experience better sound as I move farther away from the LaS if outside in mono. I also recall reading that othes have experienced that 'better' sound with Khorns. Now, I don't understand why at a much farther distance, the sound is better. Are the frequencies from the individual drivers traveling at different speeds? If so, this time alignment thing is only going to work for one specific, very small, place in space. I suspect that this distance thing with a passive crossover has more to do with space acoustics rather than driver alingment.

I never realized this was such a complex topic! I can now see why mass marketed speakers are generally passive. The general public's head would explode if they had to deal with all this stuff and no manufacturer would sell much of anything.

I know what you mean about hacking into a La Scala. I had to cut into mine to fit the Trachorn. Very difficult to start cutting with the saw knowing it would not be a reversable procedure.

[Rudy81]

Here's a few theoretical questions for you experts. While at the gym this morning I was contemplating the whole driver/time alignment thing. I realized that I was listening to the perfect driver alignment system....my headphones. No driver alignment, no room acoustic problems, just a direct and very short distance between the driver and my ear drum. Now, my headphones can't hold a candle to my Khorns as far as acoustic experience in terms of the magic of 'stereo', the sound stage and pure experience.

That thought made me question the whole driver alignment thing. If driver alignment was that important, we would all be running around trying to match the 'sound' for our headphones. Think of the savings in amps, speakers, cables etc.

So, what am I missing here? I certainly don't feel the driver aligned sound of my headphones is superior to my non-aligned Khorns.

I still plan on playing around with this, but am beginning to question the theory itself. Of course, I may be full of it.....

[Coytee]

I also recall reading that othes have experienced that 'better' sound with Khorns. Now, I don't understand why at a much farther distance, the sound is better. Are the frequencies from the individual drivers traveling at different speeds? If so, this time alignment thing is only going to work for one specific, very small, place in space. I suspect that this distance thing with a passive crossover has more to do with space acoustics rather than driver alingment.

Ok...well...I'm certainly no expert. If you want to know how to make a killer cinnamon sour cream coffee cake or discuss credit call spreads, I'm your guy. This other stuff... I still don't have my head around. So, with that understanding:

I guess you're right, sound travels at the same speed. To me, this means there is no "window" where you have a brief moment where they are all in in synch because the HF signals will always lag the LF signals. What I speculate happens outdoors is, as you step back 20/30/40 feet, you get away from the distinct sound from the woofer and the distinct sound of each HF driver. Perhaps like taking your glasses off and the words get more blurrey as you pull back. As the sound blurrs....what is really happening is the definition between each driver is starting to blend together like the words on the page, creating the illusion of a more singular sound wave.

If you can indeed, align the signals, then this mixing of the sound doesn't really need to happen as much since they'd all be coming at you in the same time reference.

Also... some might suggest simply shoving the horns (K400) farther back or the tweeter farther back. I'd wonder if this method might create even more issues because of the sound then bouncing off the top of the LaScala

Seems to me that once you properly align the signals, they will be in synch for as far out as you might hear them (trees in the way, not withstanding [])

When I took my LaScala outside (here's my old thread http://forums.klipsch.com/forums/p/74175/730166.aspx#730166 )

I walked around to various parts of the field and listened to that entire CD (Ohio State Marching Band). It was very intersting what I heard. That poor little LaScala was belting its guts out too, I really had it blistering loud.

Drag your LaScala to your doorway and try listening to it. Doing this will remove all room issues and what you'll be hearing is really more of what the speaker is actually doing. I guess one way that I'd want to put it (at the risk of it sounding offensive to someone)

Why would anyone want a speaker in their room that takes 30' of distance to reach its maximium potential when in fact, they could have that SAME speaker in the same room, hitting its max potential at 10' distance if you only aligned the signals to it?

That logic alone is one easy reason why I want my center channel to be fed with an active signal. The above link is part of what helped me reach some of these understandings for myself. (I'm kinda slow & need to learn it the hard way)