What quality of active crossover will surpass passive?

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I recently tried active crossover with LE85 compression drive + Foxtex H420 as high frequency, and 15 inch pro woofer fitted in one of Parts Express flat pack box as low frequency.
I used Rolls SX45 active crossover crossed at 1.5kHz. Although it was very convenient (and cheap), its power supply has some noise.
It sounded great anyway.

Today I tried 2nd order butterworth crossover, and honestly I thought it sounded equal if not better.

So I would like to ask a question to those who has been using active crossover. Everything being equal (amp, source, speaker, etc), what circuit and its components (such as op amps, capacitors, etc) do I have to use for an active crossover to achieve equal or better sound quality (except the phase issue) of the equal order passive crossover?
The quality you'll get with a passive crossover is most likely limited by the source's drive ability as PLLXOs tend not to be very high impedance loads (inductors would need to be very high values). The quality of an active crossover depends on the loading on the opamps and the quality of their power supply.
You should try DSP based active XO's - I think they sound very nice. Even a cheap 2x4 miniDSP is probably better than an op-amp based XO which will not be as flexible in freq and slopes etc. The higher end all digital miniDSP like nanoSHARC allow you to use your own DAC's for better sound quality.

Some people will always say that passives made with super expensive components will sound better than same in active DSP. It would be good to have a blind test of same XO's in DSP vs speaker level.
Hi Doug Kim,
I built a 3 way active crossover from audio ammature magazine and not used a passive crossover since,sometimes I wish I had not like them so well because now I have to build a amp for each channel,the biggest thing was they were driven by op amp 1 was 6 bd x over and 2 op amp's gave you 12 bd x overs and now a days Linkwitz- Riley have developed any thing you want .
I don't have tone controls I get the best op amp for the channel and EQ it into place,
I currently have a DEQ2496 and a DCX2496 fed through digital inputs and on to my amps,
My wife :mad:calls it the plate buster as it knocks plates out of the cabinets and breaks them on the floor,LOL ,then I'm in trouble,I use bungee cords now,lol.:spin:
Build a 2 way crossover active and passive and give it a try best way to find out!
Happy days are here again!
Last edited:
None of the digital ones will quite go from a to d and back to a without *some* slightly objectionable and audible degredation I have noted, remember there is an excellent chance that a previous adda has already occured. However with digital in to analog out on the xover you will have to search for one (given internals at 24/96 or better) that will not give you some benifits over big junk inside the speakerThe Behringer unit certainly will as will many others.

IMO we are fast approaching the day when a crossover adda chain will also be better. I am speaking as a mastering engineer now, not as a speaker design enthusiast. In that domain there are a lot of reasons to go active with the finest analog or digital units. FWIW, too many addaaddaddada transitions are *surely* audible, and to my ears analog noise additive is prefferable. Good analog actives can get pretty expensive however. Now I've done, spoken about something I actually do know a bit about. My turn to pay back.
Last edited:
Joined 2002
Paid Member
The active solution is not able to exploit the driver's natural rolloff and impedance characteristics. For example, in a passive xo the baffle step is achieved by simply sizing the inductor to roll off the woofer down from the baffle step till the crossover point, then padding the tweeter accordingly, this is usually not possible with active solutions in a simple way.

Otoh active solutions will always be 'blameless'when done right. Also with DSP in the mix you can correct for anomalies in driver behaviour.

My personal take is that with high quality drivers with wide bandwidth and few FR anomalies, a good passive XO will usually sound better, and for average drivers active XOs can be used to extract some extra performance. I use both kinds of solutions, even combining DSP and active XOs. Both have their merits and it depends on your use case in the end.
The active solution is not able to exploit the driver's natural rolloff and impedance characteristics. For example, in a passive xo the baffle step is achieved by simply sizing the inductor to roll off the woofer down from the baffle step till the crossover point, then padding the tweeter accordingly, this is usually not possible with active solutions in a simple way.
Yeah takes several mouse clicks. :)

the reason many still prefer passive originals against tweaked to active systems is imho due to that active xovers designed after textbook don't suffice to deal with the driver requirements.
Sangram pointed towards the reason.
Almost any driver requires some kind of equalization to achieve the desired acoustic filter response.
Now most simple active xovers don't feature EQs at all and only a few allow for very basic equing ... and those EQs are then again designed after textbook.
If You design a xover plus EQs in textbook fashion the parts number count may rise considerably.
Out of conviniance and the rising requirements in board space and overall effort most designers then use integrated OPAmps as active devices.
Don't be astonished if such a graveyard of parts doesn't sound natural or authentic at all.
Passive xovers on the other hand are typically designed specifically to the driver's requirements.
Additionally they tend to combine xover and EQ functionality in the aim to reduce component number count.
I'm not overly surprised if some listeners prefer a passive box over a active system designed after above mentioned rules.
DSP xovers on the other hand share similarities with passive xovers in the way the signal is handled.
Once the signal is in the digital domain all filtering and equing is done in one step/device before the final DA conversion.
The result can be astonishingly good even with the cheaper models, but in the end the cheap codecs used for adda limit those devices.
I'm quite sure that the DSP part, the filtering is acoustically fully transparent and any sonic fingerprint due to the sound character of the associated ADCs and DACs.
If You want to stay with analog xovers I'd suggest to deviate from textbook filters and to try to design as You'd do with a passive xover.
Filter and EQ functionality can be combined and parts number count -and active stages number count- can be reduced considerably.
I prefer to use unity gain filter structures and to replace the common OPAmp -connected as follower for a gain of 1- by a discrete Buffer, for example a simple JFET-CCS loaded JFET-follower.
Such designed xovers sounded always more authentic and natural to my ears ... Music instead of HiFi.

It would probably help to define "active" as there are many ways. Lozek makes a good point. There are still way to many that go for text book named x-overs in active without consideration of driver behaviour. Both methods deserve more care to tailor the x-over to the drivers used. Some forms of active can do things you simply cannot do with passive components.
Look at actual acoustical slopes and forget about "named" active slopes and you're on your way.
Linkwitz's site is contains great examples of methodically generating active crossovers hand in hand with the realisation of the speakers themselves.

But there is not really 'one true way' for this - which is why just throwing the textbook modules at certain problems may or may not work.
all very good points,A active crossover to me is one that goes in after your input be it digital or analog,with digital you can alter almost any signal being input to the processor, I use digital outputs (spdif or AES) into a DEQ2496 it has alot of sound processors in it,Digital signal goes into the DCX and is split into 6 channels or 2 3way analog outputs it also has EQ and other processors in it,The dac's are AK4393's getting updated by newer ones like AK4395 and 96's not on the bottom of the list,
YEP ANYTHING you put in the signal chain is going to be noticed,unless ALL is perfectly flat ,but that's the reason to be able to modify what ever you want,Now a good built passive system might be built to play classical music very well ,try some AC/DC and maybe not sound too well,I am not knocking anyone's system just a example of how you can change the sound,But if your digital move some clicks and now that's rock and roll,and vice versa.
Speakers that were available in 1973 , I would buy Full range and tweeters and actively cross them to needed ranges and eq the rest ,all analog at that time ,got comments like man that sound's live,
so enough of my rambles,build one and hear for your self,But it is addicting,
I do have volume controls on each amp,once they are set I use the front end to control overall volume,works for me,One more Q people ask does balanced sound better,Does papa Pass build balanced amps,almost all of them have both jacks available,but of most upgrades can you go balanced,
Last edited:
your right

This all depends if you understand the definition of balanced. Many do not.
:) Explain your definition ,I think I have it down,signals are out of phase so they cancel noise ,you can run long distances and not degrade the signal,This is due to 2 signal wires careering the signal ,it also has a shield to block emi/rfi ,and doesn't change with load currents like a single conductor shield return does,I know I missed something fill me in.
I don't know alot about the pin 1 problem can you elaborate ?
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.