[brj]

Based on simple reliability metrics like parts count, etc., yes, the passive crossover is likely to be more reliable than the active.  That may not matter, however, as long as the active crossovers are reliable enough.  To make up numbers for the sake of example, if the MTBF (mean time between failure) for the passive is 1000 years and the active is 100 years.... well, I'm probably just fine with the active.  (Is the lifestyle of the individual with 10 billion dollars really any different than the individual with 5 billion? )

While not manufactured in the quantities of your favorite cell phone or hard drive, digital devices of an audiophile nature are mostly comprised of parts manufactured in quantities large enough that the product is still almost certainly going to have a similar type of "bathtub" reliability profile.  So if they don't fail within the warranty period, they're probably going to outlive your interest in owning them!

(Given that the neither the active nor passive crossover have moving parts, I wouldn't be surprised if the ultimate failure mode was the same for both... solder joints.)

[Rclark]

.... please define reliable in terms of crossovers, I'm not quite making the connection. 
 Do active crossovers often break down? Need an oil change?

[sts9fan]

What size cap are you guys putting on the tweeter? 20uF? So the crossover point is in the 900s?

[JohnR]

.... please define reliable in terms of crossovers, I'm not quite making the connection. 
 Do active crossovers often break down? Need an oil change?

It's more that they need scheduled rests, or they get CFS. Especially digital ones, just imagine if you had to execute an instruction millions and millions of times a second, I bet you would get tired pretty quickly too.

[cheap-Jack]

Hi.

Cheap-jack stop being so aggressive. This is not that sort of place. Your shit stinks like the rest of us. 
Everyone else, let's keep this civil. This has been a very nice thread.

Mr. Facilitator

Yes, this is a CIVIL venure for technical talks. Please stop using foul language, e.g. "shit" here AGAIN, may I suggest you speaking as a Lab facilitator.

If anybody wants to vent its spleen, please go somewhere else.

Let keep this venue a civil place.


c-J

[cheap-Jack]

Hi.

What size cap are you guys putting on the tweeter? 20uF? So the crossover point is in the 900s?

Let's talk civil.

It all depends on the corner frequency of roll off chosen for the tweeter.
The corner frequency will depends on the resonance frequency curves of the tweeter & the woodfer assuming this is a 2-way loudspeaker system.

c-J

[Trismos]

Hi

Hi.
Mr. Facilitator

Yes, this is a CIVIL venure for technical talks. Please stop using foul language, e.g. "shit" here AGAIN, may I suggest you speaking as a Lab facilitator.

If anybody wants to vent its spleen, please go somewhere else.

Let keep this venue a civil place.


c-J

Lol

[Davey]

Who has first hand knowledge of trying both on the same speaker?  Did they measure the same?  Sonically better, worse or the same?  What did you use for your active?  miniDSP, DEQX???

I have firsthand knowledge on my Magnepan MMG's....and a few others.  I configured them for identical transfer functions with both a passive (stock) and line-level (active analog that I built) crossover.

If you do it right there will be no measureable difference in SPL measurements....and indeed there wasn't.  I used the same amplifier in both cases.  A multi-channel type where I utilized more of the channels in the bi-amp configuration.

The system is audibly superior using the line-level crossover approach.  Superior dynamics, superior resolution at all SPL's, louder playback levels without strain, etc, etc.

In order for this type of comparison to be meaningful you MUST configure a line-level crossover and a speaker-level crossover with identical electrical slopes.  They don't necessarily have to be flat or optimized....just the same.
If the line-level crossover is a digital type then you need to adjust its slopes to be identical to a passive counterpart also.  But, in that case, keep in mind that you are adding a significant chunk of electronics into the mix.  ADC, DAC, DSP-processing, etc, etc.

I've also performed this same apples/apples experiment using my miniDSP, DCX2496, and a Rane RPM unit I have on hand.  In those cases I felt the reproduction was still superior to the passive speaker-level approach, but not equivalent to my active, analog crossover.  I have no interest in an expensive unit like the DEQX so haven't tried it.

I hope that is some meaningful information in this surreal thread. 

Cheers,

Dave.

[jtwrace]

I have firsthand knowledge on my Magnepan MMG's....and a few others.  I configured them for identical transfer functions with both a passive (stock) and line-level (active analog that I built) crossover.

If you do it right there will be no measureable difference in SPL measurements....and indeed there wasn't.  I used the same amplifier in both cases.  A multi-channel type where I utilized more of the channels in the bi-amp configuration.

The system is audibly superior using the line-level crossover approach.  Superior dynamics, superior resolution at all SPL's, louder playback levels without strain, etc, etc.

In order for this type of comparison to be meaningful you MUST configure a line-level crossover and a speaker-level crossover with identical electrical slopes.  They don't necessarily have to be flat or optimized....just the same.
If the line-level crossover is a digital type then you need to adjust its slopes to be identical to a passive counterpart also.  But, in that case, keep in mind that you are adding a significant chunk of electronics into the mix.  ADC, DAC, DSP-processing, etc, etc.

I've also performed this same apples/apples experiment using my miniDSP, DCX2496, and a Rane RPM unit I have on hand.  In those cases I felt the reproduction was still superior to the passive speaker-level approach, but not equivalent to my active, analog crossover.  I have no interest in an expensive unit like the DEQX so haven't tried it.

I hope that is some meaningful information in this surreal thread. 

Cheers,

Dave.

Thanks!

Do you think the electrical noise that has been talked about in other threads in the miniDSP is an issue?

[poseidonsvoice]

I hope that is some meaningful information in this surreal thread. 

Cheers,

Dave.

Indeed it is. Thank you for staying on topic. 

Anand.

[cheap-Jack]

.... please define reliable in terms of crossovers, I'm not quite making the connection. 
 Do active crossovers often break down? Need an oil change?

"Reliable" here defines as less chances of breaking down.

Yes, ALL active X-overs are built with active devices, e.g. silicon chips, bipolars or even vacuum tubes, which ALL need electricity to power them to work. These active devices can break down sooner or later. "Often" or not depends on the design/build of the X-overs.

Whereas passive X-overs are built with passive parts, e.g. resistors, capacitors &/or inductors. NO need any electric power to work. They are therefore much much more "reliable" to run.

As far as I am concerned, it is the sound quality of a X-over that matters me.
IMO, passives sounds better than actives MUSICALLY.

NO, they don't need "oil change".

c-J

[Davey]

Thanks!

Do you think the electrical noise that has been talked about in other threads in the miniDSP is an issue?

No, not really.

All DSP gadgets have higher electrical noise than well-designed active analog circuits.  Whether this noise is an issue depends upon how the gadget is used and the greater system is configured.  Gain structure, volume control position in the chain, power amp voltage gain, speaker efficiency, etc, etc.

The miniDSP 2x4 has somewhat lower noise on its outputs than other DSP gadgets.  The miniDSP 2x8 unit utilizes noise-shaping and exhibits some increased noise above the audio band, but it's not unusual compared to many DSP black boxes.  The miniDSP 2x4 has limited voltage capabilities (but still sufficient for most applications) that might be a concern to some users.

Hope that helps.

Cheers,

Dave.

[cheap-Jack]

Hi..

(1) Have you ever actually designed a crossover or gotten a speaker up and running from scratch?

(2) And you concern with speaker wire length is rather funny.  You save what, maybe 4 feet by using an external passive crossover?  The COIL in the crossover is DOZENS of feet long.  If you really want a "short path", then removing the coils from the signal path is far more important.  So, you up for removing some coils?  Course, then you'd have to use an active crossover.

(3) And, you've stated again, you hate the clinical sound of digital and love your vinyl.  That's exactly what I said you said.  I don't think that's putting words in your mouth at all.  And you still have not addressed the question of what the heck you are talking about when you say phase distortions get magnified by an amp.  That suggest to me that you really,

 truly don't understand how crossovers actually work....

(1) No, I never design/built any active X-overs. Why? Passive sounds better. Do I still want to waste my time & money to get somethng less good?

(2) I don't think you know how passive bi-wiring works at all, bud. Why don't you learn more about it before opening yr mouth please.

Read my posts again. I already stated clearly the alleged 'pitfall' of passive bi-wiring is HF & LF signals cross-talk along the same run of cable. Shortening the speaker cable run from the power amp will minimize the alleged cross-talk
btween LF & HF signals.

What's that anything to do with voice coils????????????

(3) Of course I know how a frequency filter or "line level X-over" works as I already posted clearly & explained how phase shift comes in. Either you did not read it properly or you just don't know how a filter works.

ALL electrical parts or components get phase shift or time delay for AC signals.
Even the best very best design/built active device, like an audio amp, also get time delay or phase shift itself & such phase delay will be frequency depending.

Let me quote you one very best brandname audio  power amp in the market, phase shift quoted; 1 degree at 20KHz.

So any audio signals passsing out of a line-level X-0ver with certain phase shift  go thru an amp will have more phase shift due to the phase shift generated by the amp itself. That's why I stated any active device, here an audio amp will "magnify" the phase shift of the incoming signals.

But signals passing thru a passive loudspeaker X-over don't need go thru any further phade shifting being at the end of the audio chain.

c-J

PS: may I suggest you to read more basic electronics to learn more like I did.
     "Radiotron Designer's Handbook, 4th edition".

[randytsuch]

Hi..
ALL electrical parts or components get phase shift or time delay for AC signals.
Even the best very best design/built active device, like an audio amp, also get time delay or phase shift itself & such phase delay will be frequency depending.

Let me quote you one very best brandname audio  power amp in the market, phase shift quoted; 1 degree at 20KHz.

So any audio signals passsing out of a line-level X-0ver with certain phase shift  go thru an amp will have more phase shift due to the phase shift generated by the amp itself. That's why I stated any active device, here an audio amp will "magnify" the phase shift of the incoming signals.

Speaking for myself, I agree that amps can introduce phase shift.  What I don't agree with is that they would "magnify", or amplify the phase shift.  I would expect the phase shift to be basically constant, at any given frequency.  So, amount of phase shift might change over frequency, but not change at any single frequency.  In other words, I see the phase shift is additive, and not multiplicative as you imply.  Amplifiers work in the voltage domain, not the time domain.

The concern here is if you used different amps for the low end, versus the mid or the tweeter, they could introduce different phase shifts, but I also don't know if it would be audible or not.  If you used the same amp for all drivers, this problem should not matter, at that point it would act the same as a single amp, in terms of phase shift.

Most likely, I think we just need to agree to disagree.

Randy

[brj]

cheap-Jack, since you've never designed or built (or presumably programmed) an active crossover, can you tell us which active speakers you've personally auditioned and which acoustic properties you specifically feel sounded better on the equivalent passive implementation?  Also, which amps were you using for the comparisons, since that could play a significant factor given your assertions regarding amplification of distortion.  Also, letting us know what your acoustic preferences are would be helpful.  (i.e. type of music, and do you consider tone and imaging more important than dynamics and soundstage, etc..)  You made what appear to be rather absolute statements regarding speaker performance in what is a very subjective hobby, so having some context in which to evaluate your statements would be helpful.

For reference, I know of only a few commercial manufacturers that have designed both passive and active versions of the exact same speaker, so I'm curious as to which ones you auditioned to base your opinions on.  PMC is the one that comes to mind first, though I know there are several professional monitor manufacturers that could be options.  Actually, even Audioengine could offer an interesting comparison between their A5 and P5 models, though that is a budget speaker and thus may have other compromises that might swamp the crossover implementation differences.

Thanks!

[sts9fan]

Paradigm also.

[BobRex]

Hi..
(1) No, I never design/built any active X-overs. Why? Passive sounds better. Do I still want to waste my time & money to get somethng less good?

(2) I don't think you know how passive bi-wiring works at all, bud. Why don't you learn more about it before opening yr mouth please.

Read my posts again. I already stated clearly the alleged 'pitfall' of passive bi-wiring is HF & LF signals cross-talk along the same run of cable. Shortening the speaker cable run from the power amp will minimize the alleged cross-talk
btween LF & HF signals.

What's that anything to do with voice coils????????????

(3) Of course I know how a frequency filter or "line level X-over" works as I already posted clearly & explained how phase shift comes in. Either you did not read it properly or you just don't know how a filter works.

ALL electrical parts or components get phase shift or time delay for AC signals.
Even the best very best design/built active device, like an audio amp, also get time delay or phase shift itself & such phase delay will be frequency depending.

Let me quote you one very best brandname audio  power amp in the market, phase shift quoted; 1 degree at 20KHz.

So any audio signals passsing out of a line-level X-0ver with certain phase shift  go thru an amp will have more phase shift due to the phase shift generated by the amp itself. That's why I stated any active device, here an audio amp will "magnify" the phase shift of the incoming signals.

But signals passing thru a passive loudspeaker X-over don't need go thru any further phade shifting being at the end of the audio chain.

c-J

PS: may I suggest you to read more basic electronics to learn more like I did.
     "Radiotron Designer's Handbook, 4th edition".

Regarding #1 - You've made this claim before, that according to your ears a passive sounds better.  What actives did you listen to in coming to that conclusion?

Regarding #2 - Didn't you just state something about staying civil?  That didn't last long.

Regarding  #2 - You have previously stated that you removed the crossover and placed it into a separate box in order to reduce crosstalk.  Can you site a reference that defines this crosstalk, defines the minimum length required to "avoid" it, and what frequencies are affected?

Everything I remember about removing crossovers is based upon eliminating the harsh vibratory environment of the speaker box.  This is based upon the theory that caps can become microphonic and then distort the signal.  I don't recall anybody (except you) claiming that the reason is to reduce / eliminate crosstalk.  Think about it this way - A simple 2 way, crossed over at, say, 2K, with a first order crossover will have a bass driver that responds to 30Hz and 3KHz (and beyond, but that's enough for this example.) Will those two extremes create crosstalk? If so, then why not run separate wires for those too?  Where would you draw the line.  Also, given the midrange wavelengths, just how short do you truly need?  Oh, and is that crosstalk readily audible?  Audible enough that a single full range speaker (say a Quad ESL) would readily display the artifacts?  If so, then why hasn't anybody else picked up on the crosstalk either with measurements or audible accounts?

Regarding #3 - No not all electrical componenets add phase shift in the audible range.  You can create bandwidth so far beyond audible that the phase shift is essentially non-existent.  Oh, resistors don't shift phase.  Besides, unless your loudspeakers are truly phase and time coherent, the tiny amount of phase shift you are concerning yourself with is inaudible.

PS:  Some of us have been to school and studied electrical theory beyond the radio handbook.  May I suggest you consider that.

[srb]

Audioengine could offer an interesting comparison between their A5 and P5 models, though that is a budget speaker and thus may have other compromises that might swamp the crossover implementation differences.

Just wanted to point out that although many "active" speakers have internal active electronic crossovers and separate amps for each driver, the Audioengine "powered" speakers simply use an internal stereo amplifier (to power both speakers) in one of the enclosures and employ a standard passive crossover between the midbass and tweeter drivers.
 
So the only comparison that could be made between the A5 and the P4 would be the difference between the internal 50W/channel amplifier and an external amplifier of your choosing, and even then there is the fact that the midbass driver is a 5" in the A5 and a 4" in the P4.
 
And now back to the "I know more than you" mean-spirited pissing contest that always ruins a good topic.
 
Steve

[Tyson]

I agree that the only meaningful comparisons are on the same speaker, using the same slopes between active and passive.  I've done this exact experiment with 3 different speakers, and it's consistently the active approach that sounds better.  By his own admission, cheap Jack has NOT done this ("Why would I waste my time?").  As I said before once he has done this type of work/experiment, maybe he can come back and make meaningful comments. 

Also, if you separate speaker wire to prevent high/low crosstalk, why does it only matter at the speaker?  What about all those other wires in the system carrying that terrible, crosstalk-prone full range signal?  Like the interconnect between the preamp and amp.  Per cheap-Jack's logic, we really should separate that signal into highs and lows and use 2 separate cables, to ensure there's no crosstalk.  And to be thorough, we should probably do the same from any/all sources.

I'm not trying to pick on anyone here, but I see this a lot - people in the audio world focus on minutia that makes a small difference (external passive crossovers, double wire runs), versus other stuff that makes a much bigger improvement, like room treatments, going active, using EQ to tame bass humps, etc....

[macrojack]

As I tried to suggest in my last post to this thread, theory is NOT what was requested. Jason is asking for experience from members who have actually changed from passive to active and vice versa. What did you learn? What did you prefer? And why? That sort of thing.

I've read his posts for a couple of years now and I've had a few offline conversations and business transactions with him. Therefore, I feel sure in what I'm saying. He's a practical, pragmatic guy so I'm sure he's not just asking for speculation. He wants to know which approach you preferred and why you chose it over the alternative.

If you have never tried both active and passive crossover systems in the same pair of speakers, he is not talking to you.