• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Active Crossover...One Size Fits All??

ic-racer

Super Member
I have 3 projects that I can combine into one. I build guitar amplifiers and have gotten pretty good at making PC boards (in fact I'm etching one today). I have some Yamaha ns a836 speakers that I would like to use as a test bed for speaker cabinet/crossover modifications and I have way too many power amplifiers that I have fixed up. I have no interest in selling them, so they sit unused.

I had an idea to build this active crossover and run the Yamaha speakers with 3 of my stereo power amplifiers.

http://archive.siliconchip.com.au/cms/A_30278/article.html

My concern involves knowledge that passive crossovers are 'speaker specific' but this active crossover seems pretty generic and has no controls.

I'm interested in some comments by those of you that use or have used active crossovers.

Thanks.
 
Last edited:
Register to hide this ad
In order for an active crossover to be truly generic it has to be variable; variable frequencies, slope and level. Whenever you are restricted to set frequencies, slopes and levels a crossover will only be optimized for one specific use. For generic use those factors should be variable. An example is the Behringer CX3400 (~$130). It's a three way active electronic crossover with variable crossover frequencies and levels. There are other more expensive active crossovers that also allow slope adjustment. An example is the two way Bryston 10B (not cheap!).

FWIW: I use a Behringer CX2310 2-way crossover.
 
That's a lot of effort to be stuck with a LR 24dB slope. You may like this just fine or maybe not.
My thoughts are to duplicate the original passive slope and modify to liking/measurements from there.
 
I have been using active xovers for over 9 months now- I am SOLD on their benefits of phasing, freq. adjustment, levels, etc! I have been a custom speaker designer for many years using thr tried & true passives- but when I finally broke down and bought a used, good, Rane 3 way elec. I was blown away how much cleaner my systems sounded and the power resevers as well. Now almost everything I build is with elec. xovers in mind. Once you go there you will not fancy going back. You can mix, match amps for detail and effect, swap preamps as well- a LOT of advantages.

The xover you are looking at is a very nice one- with set lrvrls. IF i were to build one I would want to have several sets of xover point to exchange.

DC
 
I use an active (Electrovoice Dx38, 2 in, 4 out).

I was once at the engineering facility for Klipsch with several guys and one thing we did was a quasi "shootout" between a whopping 3-4 active crossovers and a couple passives.

Bottom line... if you get a cheap active crossover, our rather unscientific experiment indicated you might end up with lessor sound quality. We were fairly united that the more expensive units sounded better (around $1,000 price point). The units we listened to were were DSP units and also allowed for signal (time) delay.

Most but not all of the guys in the crowd were engineers of one type or another, with the bulk of them being EE's. (I am not an engineer)

One of them went on to say later on that after going through the process we went through he began to feel as though the active crossover represents the heart of the entire system.

I thought that was an interesting point.

We also compared the EV Dx38 which new, cost about $1,200 (now discontinued) but, you can find them on Ebay for 500/700, to a no cost spared passive crossover ($3,000 in parts alone for the pair).

I personally could not tell a bit of difference between the Dx and the $3,000 passive. Does this mean the active sounded as good as the $3,000 passive or the $3,000 passive sounded as good as the Dx?
 
Very interesting Coytee. I was wondering about that too because the equipment from DEQX is so much more expensive than a simple Behringer DCX.

What are some of the factors involved? Would the combination of an ultracurve DEQ to the DCX make a poor man's DEQX? At least you could then control dips, bumps and other problems related to the room.
 
I have been down the Beringer xover route, they tend to sound harsh and thin to me. Now a Rane, Ashley, EV or even a DBX to me sound much smoother.
DC
 
The easiest thing here is to get either a MiniDsp (use your computer to configure it) which has pretty much any flexibility you need to adjust slope/rate/alignments that no analog can practically accomplish. The Behringer DCX is another great option, and it's transparent for any level of set up if you use it's analog inputs; but be warned - if you use it's digital input you must send a native 96kHz/24bit PCM signal; it's internal re-sampler chip should never be used (many people make a mistake of sending non-native PCM to it). But, it's analog input has no issues and there is no reason to not use it. There are also many 'upgrade' options you can purchase to replace it's analog output board and other options, which are really not going to accomplish anything in regards to real SQ, but some options can add specific features (like 6 channel analog volume control) that some people may find optimal. The DCX is very powerful, but the MiniDSP does have ultimately more flexibility.

I can't imagine any reason so get the much more expensive DSP units unless there is some very specific function/feature that you require that is only available on said high priced units. Either that, or you may just want something that costs more for it's name or aesthetic value(all perfectly valid reasons).
 
The miniDSP discussions seem to center on getting it set up correctly. Haven't tried it yet, but was wondering how hard it really is to get it right for your room.

Presently I am working on a triamped open baffle speaker and I am wondering if going digital, and which particular solution, is the right way to go or if I just don't have the proper measurement setup to properly do digital and should just be happy with 2 analog crossovers and room correction panels and call it a day.
 
The miniDSP discussions seem to center on getting it set up correctly. Haven't tried it yet, but was wondering how hard it really is to get it right for your room.

Presently I am working on a triamped open baffle speaker and I am wondering if going digital, and which particular solution, is the right way to go or if I just don't have the proper measurement setup to properly do digital and should just be happy with 2 analog crossovers and room correction panels and call it a day.

Personally, I prefer the DCX because you don't have to link a PC to it to tune it/effect changes.

Also, if you are not comfortable at first, you can set the DCX to use (From it's panel control) it sort of like an analog standard crossover, albeit with far more flexibility. Later on you can get used to the PC GUI(which is excellent, btw) and easily tune in precise settings. The front panel GUI can look confusing at first, but it is really simple and once you understand it you would have no problem.

The DCX GUI can be downloaded from Behringer and run in virtual mode without a DCX so you can evaluate it's operation.

http://behringerdownload.de/_software/DCX2496_V1_16.zip

Oh, one additional thing: the MiniDSP has RCA inputs and outputs and is directly interfaced with regular consumer gear. The DCX has XLR inputs and outputs. A RCA to XLR adapter is needed for the input, and the DCX has a specific variable analog pre-amp so you can directly use consumer or professional line level inputs. However, the outputs require XLR to RCA adapters AND 10-12dB passive attenuators because the outputs, though they have trims, are optimized for professional gear only in so far as the voltage ranges the operate within, Therefor the attenuators are required to have proper signal to noise ratio AND to prevent potential over-driving of the consumer RCA inputs. Parts Express does have inexpensive Dayton brand passive attenuator to go inline between RCAs or they have a higher build quality Harrison Labs equivalent for higher cost.
 
Last edited:
If you go the digital route you are coverting any analog input to digital and then back to analog, after processing, then sending the signal to the amp(s) via the multiple DAC's provided in the crossover. Really cannot see how this can be ideal situation for a crossover in hifi, esp when using analog inputs. (If you are an analog guy)

Not just speculating, I have gone the digital route via Behringer DCX2496 and ran back to their analog CX3400. Alot of super cool features in the DCX did not add up to better sound for me, even when inputting digital in to the DCX. I want to upgrade active crossover someday, haven't decided the route as of yet. Passive line level to me seems the cleanest way a guy could go, no DAC's and conversions, no opamps or cheap pots. Anyone got experience here they could share???
 
If you go the digital route you are coverting any analog input to digital and then back to analog, after processing, then sending the signal to the amp(s) via the multiple DAC's provided in the crossover. Really cannot see how this can be ideal situation for a crossover in hifi, esp when using analog inputs. (If you are an analog guy)

Not just speculating, I have gone the digital route via Behringer DCX2496 and ran back to their analog CX3400. Alot of super cool features in the DCX did not add up to better sound for me, even when inputting digital in to the DCX. I want to upgrade active crossover someday, haven't decided the route as of yet. Passive line level to me seems the cleanest way a guy could go, no DAC's and conversions, no opamps or cheap pots. Anyone got experience here they could share???

This is a real concern for me too. In discussions with AK-er Poultrygeist, what he does is run the midrange direct to an efficient speaker (98db Eminence Beta) from a tube amp and then he runs the same signal into a sub woofer (98 db Eminence Alpha via SS class-d amp), then another analog crossover to BG Neo 3 planar tweeters via a T amp. He tuned the outfit by hand/ear. I want to go digital to use a microphone and software to tune the dips/bumps to the room.

On the face of it, one would need both a digital crossover and something to deal with timing, phase and parametric equalization. Maybe a DCX plus a miniDSP? or a DCX and DEQ? Ideally one could use a fancier solution like DEQX but that is 5 to 6 times more expensive for the cheapest DEQX, an order of magnitude more for their "normal" tools.

It would be good to hear from those that have been down this road before.

About the OP - it looks like one size does not fit all. Cost wise, an all analog solution is an order of magnitude cheaper than an all digital one IMO.
 
If you go the digital route you are coverting any analog input to digital and then back to analog, after processing, then sending the signal to the amp(s) via the multiple DAC's provided in the crossover. Really cannot see how this can be ideal situation for a crossover in hifi, esp when using analog inputs. (If you are an analog guy)

Not just speculating, I have gone the digital route via Behringer DCX2496 and ran back to their analog CX3400. Alot of super cool features in the DCX did not add up to better sound for me, even when inputting digital in to the DCX. I want to upgrade active crossover someday, haven't decided the route as of yet. Passive line level to me seems the cleanest way a guy could go, no DAC's and conversions, no opamps or cheap pots. Anyone got experience here they could share???

I have used the CX2310 and CX3400 in the past, and these are indeed excellent devices(for basic xover use) that are transparent, used correctly. The parts used in these are surprisingly good, but you are indeed going through op-amps in that unit, NJM4580 is memory recalls. Of course, this is a very good part, and should be transparent, used correctly, but many hi-fi people like to say it's not transparent, with no other evidence than is typical with any other part(say wire even). However, the lack of ability to fine adjust/tune the slopes separately from each other is absolutely not ideal for a hi-fi unless you happen to have drivers that are near flat and with very little phase deviation for about 2 octaves past the crossover points to that the fixed L-R 24 slopes integrate properly between the drivers in the real world. In pro audio, this issue of perfectly integrating drivers is not nearly as big a deal as it is in a home hi-fi application.

The DCX's AD-DA process, well, it's excellent, and there is no reason that it will color any sound audibly for music reproduction. That digital input is very picky on the DCX though, you need to be sure to feed it only 96kHz/24 bit PCM. Any other will enable the re-sampling chip that produces substantial aliasing artifacts.

I am not hung up on 'analog' only as you seem to be, but I can understand how the mere concept of putting your signal through an AD-DA process could bother you. In fact, it is probable that the mere concern for this can cause 'perceived' SQ issues in the case you focus on this; perceptions that may be not possible to actually detect if the comparison was under controlled/blinded conditions - but that is another subject - and pertains to so many issues.
 
On the face of it, one would need both a digital crossover and something to deal with timing, phase and parametric equalization. Maybe a DCX plus a miniDSP? or a DCX and DEQ?

I may be confused. The DCX allows precise adjustment of time delay, phase and it has many powerful EQ functions. Perhaps you specifically want something like convolution?

You mention adjusting for room issue? Beyond BSC control for a specific room/setting, LF peak reduction for major room modes or general broad tonal customization, I am not sure what you desire here.
 
The deqx hdp-express is about 2K. You'd need 3 for the scenario I described plus the cost of a dcx crossover (separate from the amps and speakers). I guess you could get by with just 2 units. Using 2 dcx units and 2 Behringer DEQ units would cost about 1000 dollars. The DEQ is a digital equalizer and it can do some interesting things but you provide the intelligence - the DEQX has smart software that does a lot of this and other things like phase correction.

The miniDSP is an interesting solution. A 2x4 is $125 assembled, plus 2 dcx crossovers at 225 or so each, that's a very attractive package. But you are on your own for algorithms and corrections.
 
Using 2 dcx units and 2 Behringer DEQ units would cost about 1000 dollars. The DEQ is a digital equalizer and it can do some interesting things but you provide the intelligence -

I just wanted to know: what EQ shaping is added to capability by adding a DEQ to a DCX?
 
Still relatively new down this road. My understanding is that the DEQ has an FFT so you can do sophisticated functions, not just fancy tone controls or parametric equalization though it can do that too.

http://www.enjoythemusic.com/magazine/equipment/0705/behringerultracurve2496.htm

The DCX on the other hand, works more like a fancy crossover. By working in the digital domain you avoid some phase problems that affect LC crossovers. You are limited to 3 filter functions if I recall correctly.

http://www.avsforum.com/t/1268335/dcx2496-dynamic-eq-how-to-configure
http://www.avsforum.com/t/1087360/deq2496-or-dcx2496

Some things can be accomplished on either unit but the DEQ is capable of a lot more and also has the FFT and computer interface.

Edit: This is an old but insightful link from a couple of years back about using the DEQ for room correction. Mind you I don't believe you can solve everything in DSP, you need to start with a room that is able to be corrected without horrible reverb, early hard reflections or problems absorbing sound energy.
http://forums.audioreview.com/gener...-behringer-deq2496-room-correction-16568.html
 
Last edited:
Nyal Mellor at Acoustics Frontiers is an acoustics specialist that has written about these digital solutions online for some time (among other authors).

http://www.computeraudiophile.com/f11-software/help-all-digital-deqx-solution-6695/
http://blog.acousticfrontiers.com/

Apparently these solutions are not as easy to implement as they appear at first glance but the results can be magic. One has to do the basics like room treatment and speaker placement but this can be the "icing" and one does not have to be a millionaire to do it if one is determined enough and has the math background.
 
Still relatively new down this road. My understanding is that the DEQ has an FFT so you can do sophisticated functions, not just fancy tone controls or parametric equalization though it can do that too.

I am not aware of the DEQ performing any other than standard EQ shaping filters such as parametric, shelving and graphic(which is simply a series of precision designated notch/peak filters).

However, the DEQ has a flashy FFT display/analyzer function for room measurements when you attach a measurement mic.

The DCX on the other hand, works more like a fancy crossover. By working in the digital domain you avoid some phase problems that affect LC crossovers. You are limited to 3 filter functions if I recall correctly.

It should be cautioned that standard filters are usually preferred in practical application. If say you correct a specified minimum phase response peak of the actual loudspeaker(not a non-minimum phase error), which of course, most speakers have some characteristics that could be improved, then the counteracting filter such as that an analog EQ(and the DCX is designed to mimic, as well as the DEQ) produces nulls the reverse amplitude and phase error. The DCX and DEQ use IIR digital filter types. Now, the other kind of digital filter is the FIR filter; functions that can produce amplitude changes while not affecting phase for example, or alter phase in ways not normally possible otherwise. However, the reality of this is that it would not then correct the phase error along with the amplitude in the example given above of a minimum phase speaker response error. And if you wanted FIR for the crossover functions themselves(say to use extreme rate rolloff slopes without phase shift), realize that the desired phase perfect response only works closely on design axis of multiple integrated drivers on a speaker. As you move off axis, the filters do not any longer have the predicted properties, since the drivers change their FR and phase response as you move off axis, and severe pre-ringing artifacts can contaminate the off axis response of the speaker. Refer to this link and scroll down to 'digital crossovers': http://www.linkwitzlab.com/frontiers.htm#I

For say LF correction of a room; this is more or less minimum phase situation, and the analog-like phase change with a standard corrective parametric band is desired as it nulls the phase change in the minimum phase error.

Now, there may be some cases where FIR is beneficial, but it would probably be most so in an environment that has unique problems that need to be addressed in other ways for optimal effect(such as having a decent/good acoustic environment to begin with instead of trying to use extreme correction methods on the signal). One could for example improve SQ on a very narrow listening window by counteracting some extreme room artifacts in the full bandwidth; but better results would be had by addressing the actual room itself rather than depend entirely on signal manipulation.

The DCX also has dynamic EQs, etc. Now, the DEQ's features not found on the DCX are the stereo expander, the graphic EQ(but since this is just a collection of notch/peak filters - they can be manually entered into the DCX's parametric filter.

If the DEQ does have FIR filter types in some capacity, I am not aware of it.
 
Last edited:
Back
Top Bottom