tweeter suggestion

tweeter suggestion

I have a pair of Altec 620s (604-8G speakers) with a McIntosh C28 preamp and
a McIntosh 240 tube amp. I'm using a pair of JansZen Model 65 electrostatic
tweeters that sound excellent with anything acoustic or light but suck for
loud rock. Any suggestions? It doesn't have to be vintage and I'm not
interested in looking at anything worth more than the rest of the system but
I'd like to get some ideas and opinions before spending money. Thanks.

Bob Walker

Re: tweeter suggestion

"Walker" <lbfoh@yahoo,com > wrote in message
news:fmao3c01bqt@news1.newsguy,com

> I have a pair of Altec 620s (604-8G speakers) with a
> McIntosh C28 preamp and a McIntosh 240 tube amp. I'm
> using a pair of JansZen Model 65 electrostatic tweeters
> that sound excellent with anything acoustic or light but
> suck for loud rock.

The 604-8G series of loudspeakers has been around for a long time. Compared
to modern true coaxial speakers, they have two major problems, a woofer with
relatively small Xmax, and a very under developed crossover between the
woofer and the tweeter.

You can address the woofer Xmax problem with a really good subwoofer, and
you can address the crossover problems with a good modern digtial speaker
management unit, AKA electronic crossover.

Re: tweeter suggestion

"Arny Krueger" <arnyk@hotpop,com > wrote in message
news:fmb7o301sel@news1.newsguy,com ...
> "Walker" <lbfoh@yahoo,com > wrote in message
> news:fmao3c01bqt@news1.newsguy,com
>
>> I have a pair of Altec 620s (604-8G speakers) with a
>> McIntosh C28 preamp and a McIntosh 240 tube amp. I'm
>> using a pair of JansZen Model 65 electrostatic tweeters
>> that sound excellent with anything acoustic or light but
>> suck for loud rock.
>
> The 604-8G series of loudspeakers has been around for a long time.
> Compared
> to modern true coaxial speakers, they have two major problems, a woofer
> with
> relatively small Xmax, and a very under developed crossover between the
> woofer and the tweeter.
>
> You can address the woofer Xmax problem with a really good subwoofer, and
> you can address the crossover problems with a good modern digtial speaker
> management unit, AKA electronic crossover.
>

I never considered an electronic crossover and want to pursue that route.
Any suggestions where to start?

Bob Walker

Re: tweeter suggestion

"Walker" <walkerent@cox,net > wrote in message
news:fmdbt202fof@news4.newsguy,com
> "Arny Krueger" <arnyk@hotpop,com > wrote in message
> news:fmb7o301sel@news1.newsguy,com ...
>> "Walker" <lbfoh@yahoo,com > wrote in message
>> news:fmao3c01bqt@news1.newsguy,com
>>
>>> I have a pair of Altec 620s (604-8G speakers) with a
>>> McIntosh C28 preamp and a McIntosh 240 tube amp. I'm
>>> using a pair of JansZen Model 65 electrostatic tweeters
>>> that sound excellent with anything acoustic or light but
>>> suck for loud rock.
>>
>> The 604-8G series of loudspeakers has been around for a
>> long time. Compared
>> to modern true coaxial speakers, they have two major
>> problems, a woofer with
>> relatively small Xmax, and a very under developed
>> crossover between the woofer and the tweeter.

>> You can address the woofer Xmax problem with a really
>> good subwoofer, and you can address the crossover
>> problems with a good modern digtial speaker management
>> unit, AKA electronic crossover.

> I never considered an electronic crossover and want to
> pursue that route. Any suggestions where to start?

Behringer DCX 2496. It has every feature that you'll ever need to
implement - filters, limiters, time delays, parametric equalization. Just
remember to provide it with a good healthy signal, and use it with power
amps that require it to provide a strong signal.

Re: tweeter suggestion

On Sun, 13 Jan 2008 20:50:44 -0800, Arny Krueger wrote
(in article <fmepn40d00@news2.newsguy,com >):

> "Walker" <walkerent@cox,net > wrote in message
> news:fmdbt202fof@news4.newsguy,com
>> "Arny Krueger" <arnyk@hotpop,com > wrote in message
>> news:fmb7o301sel@news1.newsguy,com ...
>>> "Walker" <lbfoh@yahoo,com > wrote in message
>>> news:fmao3c01bqt@news1.newsguy,com
>>>
>>>> I have a pair of Altec 620s (604-8G speakers) with a
>>>> McIntosh C28 preamp and a McIntosh 240 tube amp. I'm
>>>> using a pair of JansZen Model 65 electrostatic tweeters
>>>> that sound excellent with anything acoustic or light but
>>>> suck for loud rock.
>>>
>>> The 604-8G series of loudspeakers has been around for a
>>> long time. Compared
>>> to modern true coaxial speakers, they have two major
>>> problems, a woofer with
>>> relatively small Xmax, and a very under developed
>>> crossover between the woofer and the tweeter.
>
>>> You can address the woofer Xmax problem with a really
>>> good subwoofer, and you can address the crossover
>>> problems with a good modern digtial speaker management
>>> unit, AKA electronic crossover.
>
>> I never considered an electronic crossover and want to
>> pursue that route. Any suggestions where to start?
>
> Behringer DCX 2496. It has every feature that you'll ever need to
> implement - filters, limiters, time delays, parametric equalization. Just
> remember to provide it with a good healthy signal, and use it with power
> amps that require it to provide a strong signal.

Thanks. That unit looks super! I'd never seen it before. I have a couple of
Behringer mike mixers and a Behringer A/D converter and they make good stuff,
so, I have no doubt that this unit is everything you (and Behringer) says it
is.

Re: tweeter suggestion

Arny Krueger wrote:
> "Walker" <lbfoh@yahoo,com > wrote in message
> news:fmao3c01bqt@news1.newsguy,com
>
>> I have a pair of Altec 620s (604-8G speakers) with a
>> McIntosh C28 preamp and a McIntosh 240 tube amp. I'm
>> using a pair of JansZen Model 65 electrostatic tweeters
>> that sound excellent with anything acoustic or light but
>> suck for loud rock.
>
> The 604-8G series of loudspeakers has been around for a long time. Compared
> to modern true coaxial speakers, they have two major problems, a woofer with
> relatively small Xmax, and a very under developed crossover between the
> woofer and the tweeter.

What do you mean by an under developed crossover?

Edmund

Re: tweeter suggestion

"Edmund" <sigm,fr eud2@freenet.de> wrote in message
news:fn610f012eg@news1.newsguy,com
> Arny Krueger wrote:
>> "Walker" <lbfoh@yahoo,com > wrote in message
>> news:fmao3c01bqt@news1.newsguy,com
>>
>>> I have a pair of Altec 620s (604-8G speakers) with a
>>> McIntosh C28 preamp and a McIntosh 240 tube amp. I'm
>>> using a pair of JansZen Model 65 electrostatic tweeters
>>> that sound excellent with anything acoustic or light but
>>> suck for loud rock.
>>
>> The 604-8G series of loudspeakers has been around for a
>> long time. Compared to modern true coaxial speakers,
>> they have two major problems, a woofer with relatively
>> small Xmax, and a very under developed crossover between
>> the woofer and the tweeter.
>
> What do you mean by an under developed crossover?

One that does not optimally blend the upper and lower range drivers.

Re: tweeter suggestion

Arny Krueger wrote:
> "Edmund" <sigm,fr eud2@freenet.de> wrote in message
> news:fn610f012eg@news1.newsguy,com
>> Arny Krueger wrote:
>>> "Walker" <lbfoh@yahoo,com > wrote in message
>>> news:fmao3c01bqt@news1.newsguy,com
>>>
>>>> I have a pair of Altec 620s (604-8G speakers) with a
>>>> McIntosh C28 preamp and a McIntosh 240 tube amp. I'm
>>>> using a pair of JansZen Model 65 electrostatic tweeters
>>>> that sound excellent with anything acoustic or light but
>>>> suck for loud rock.
>>> The 604-8G series of loudspeakers has been around for a
>>> long time. Compared to modern true coaxial speakers,
>>> they have two major problems, a woofer with relatively
>>> small Xmax, and a very under developed crossover between
>>> the woofer and the tweeter.
>> What do you mean by an under developed crossover?
>
> One that does not optimally blend the upper and lower range drivers.

Hmm years ago I bought a few active crossovers from a Dutch company
Temporal Coherence, I really do think these are the best in the
world. And unlike some other things in high end audio, it is very
well possible to measure crossovers to see how they will sound.
If I feed my crossovers with a sawtooth and put the low,
middle and high together again, guess what, I get the exact same
sawtooth back again. Show me the first other brand crossover
which does the same!
As I understand things, this is the only right way, if one get
anything else then the exact same input, it is simply wrong.
The downside is, this is only possible with 1st and 2nd order filters
so the system need to be able to handle these modest slopes.
It is a problem I have heard often with all kind of filtering,
very aggressive filtering and it simply does not sound right.
Measure it and you will see why it doesn't sound right, with these
steep filters you will never be able to get the input signal back.

BTW I don't know about digital filtering with time shift and
stuff. Maybe it is possible to make such a filters temporal
coherent too. But I like to see that first :-)

Edmund

Re: tweeter suggestion

"Edmund" <sigm,fr eud2@freenet.de> wrote in message
news:fnb5pt0go2@news2.newsguy,com
> Arny Krueger wrote:
>> "Edmund" <sigm,fr eud2@freenet.de> wrote in message
>> news:fn610f012eg@news1.newsguy,com
>>> Arny Krueger wrote:
>>>> "Walker" <lbfoh@yahoo,com > wrote in message
>>>> news:fmao3c01bqt@news1.newsguy,com
>>>>
>>>>> I have a pair of Altec 620s (604-8G speakers) with a
>>>>> McIntosh C28 preamp and a McIntosh 240 tube amp. I'm
>>>>> using a pair of JansZen Model 65 electrostatic
>>>>> tweeters that sound excellent with anything acoustic
>>>>> or light but suck for loud rock.
>>>> The 604-8G series of loudspeakers has been around for a
>>>> long time. Compared to modern true coaxial speakers,
>>>> they have two major problems, a woofer with relatively
>>>> small Xmax, and a very under developed crossover
>>>> between the woofer and the tweeter.
>>> What do you mean by an under developed crossover?
>>
>> One that does not optimally blend the upper and lower
>> range drivers.

> Hmm years ago I bought a few active crossovers from a
> Dutch company Temporal Coherence, I really do think these
> are the best in the world.

Can't find any references to them.

> And unlike some other things
> in high end audio, it is very well possible to measure
> crossovers to see how they will sound.

Absolutely not. Crossovers and speakers must work together very closely to
produce the desired acoustical results. If you are going to measure the
results, you do it in the acoustical domain where accurate reliable
measurements and equating measured results to sound quality is very
difficult.

> If I feed my crossovers with a sawtooth and put the low,
> middle and high together again, guess what, I get the
> exact same sawtooth back again. Show me the first other
> brand crossover which does the same!

I believe that two 6 dB/octave crossovers of the simplest design will do
this.

> As I understand things, this is the only right way, if
> one get anything else then the exact same input, it is
> simply wrong.

You have to look at the output of the speakers, not the output of the
crossover.

Re: tweeter suggestion

On Thu, 24 Jan 2008 17:10:51 -0800, Arny Krueger wrote
(in article <fnbcur01bhf@news3.newsguy,com >):

> "Edmund" <sigm,fr eud2@freenet.de> wrote in message
> news:fnb5pt0go2@news2.newsguy,com
>> Arny Krueger wrote:
>>> "Edmund" <sigm,fr eud2@freenet.de> wrote in message
>>> news:fn610f012eg@news1.newsguy,com
>>>> Arny Krueger wrote:
>>>>> "Walker" <lbfoh@yahoo,com > wrote in message
>>>>> news:fmao3c01bqt@news1.newsguy,com
>>>>>
>>>>>> I have a pair of Altec 620s (604-8G speakers) with a
>>>>>> McIntosh C28 preamp and a McIntosh 240 tube amp. I'm
>>>>>> using a pair of JansZen Model 65 electrostatic
>>>>>> tweeters that sound excellent with anything acoustic
>>>>>> or light but suck for loud rock.
>>>>> The 604-8G series of loudspeakers has been around for a
>>>>> long time. Compared to modern true coaxial speakers,
>>>>> they have two major problems, a woofer with relatively
>>>>> small Xmax, and a very under developed crossover
>>>>> between the woofer and the tweeter.
>>>> What do you mean by an under developed crossover?
>>>
>>> One that does not optimally blend the upper and lower
>>> range drivers.
>
>> Hmm years ago I bought a few active crossovers from a
>> Dutch company Temporal Coherence, I really do think these
>> are the best in the world.
>
> Can't find any references to them.
>
>> And unlike some other things
>> in high end audio, it is very well possible to measure
>> crossovers to see how they will sound.
>
> Absolutely not. Crossovers and speakers must work together very closely to
> produce the desired acoustical results. If you are going to measure the
> results, you do it in the acoustical domain where accurate reliable
> measurements and equating measured results to sound quality is very
> difficult.

Arny, did you miss the point that these are ACTIVE crossovers? That means
that they go between the pre-amp and the power amps. One does not need a
speaker connected or even the amplifiers connected in order to measure them.
You would be correct if these crossovers in question were the LRC kind that
come between the amp(s) and the speaker(s). But with active crossovers, you
can perform frequency response tests, plot the crossover/rollover points,
check distortion and even perform the sawtooth test that the OP mentions,
below.

>
>> If I feed my crossovers with a sawtooth and put the low,
>> middle and high together again, guess what, I get the
>> exact same sawtooth back again. Show me the first other
>> brand crossover which does the same!
>
> I believe that two 6 dB/octave crossovers of the simplest design will do
> this.
>
>> As I understand things, this is the only right way, if
>> one get anything else then the exact same input, it is
>> simply wrong.
>
> You have to look at the output of the speakers, not the output of the
> crossover.

What does the output of the speaker have to do with the electronic
characteristics of an adjustable, active filter like an active crossover?
Sure, the desired results will be best heard/measured at the speakers
themselves, but I believe that what the OP was looking at was the phase shift
of the active crossover itself.

Re: tweeter suggestion

"Sonnova" <sonnova@audiosanatorium,com > wrote in message
news:fnds3002b2q@news3.newsguy,com
> On Thu, 24 Jan 2008 17:10:51 -0800, Arny Krueger wrote
> (in article <fnbcur01bhf@news3.newsguy,com >):
>
>> "Edmund" <sigm,fr eud2@freenet.de> wrote in message
>> news:fnb5pt0go2@news2.newsguy,com
>>> Arny Krueger wrote:
>>>> "Edmund" <sigm,fr eud2@freenet.de> wrote in message
>>>> news:fn610f012eg@news1.newsguy,com
>>>>> Arny Krueger wrote:
>>>>>> "Walker" <lbfoh@yahoo,com > wrote in message
>>>>>> news:fmao3c01bqt@news1.newsguy,com
>>>>>>
>>>>>>> I have a pair of Altec 620s (604-8G speakers) with a
>>>>>>> McIntosh C28 preamp and a McIntosh 240 tube amp. I'm
>>>>>>> using a pair of JansZen Model 65 electrostatic
>>>>>>> tweeters that sound excellent with anything acoustic
>>>>>>> or light but suck for loud rock.
>>>>>> The 604-8G series of loudspeakers has been around
>>>>>> for a long time. Compared to modern true coaxial
>>>>>> speakers, they have two major problems, a woofer
>>>>>> with relatively small Xmax, and a very under
>>>>>> developed crossover between the woofer and the
>>>>>> tweeter.
>>>>> What do you mean by an under developed crossover?
>>>>
>>>> One that does not optimally blend the upper and lower
>>>> range drivers.
>>
>>> Hmm years ago I bought a few active crossovers from a
>>> Dutch company Temporal Coherence, I really do think
>>> these are the best in the world.
>>
>> Can't find any references to them.
>>
>>> And unlike some other things
>>> in high end audio, it is very well possible to measure
>>> crossovers to see how they will sound.
>>
>> Absolutely not. Crossovers and speakers must work
>> together very closely to produce the desired acoustical
>> results. If you are going to measure the results, you do
>> it in the acoustical domain where accurate reliable
>> measurements and equating measured results to sound
>> quality is very difficult.
>
> Arny, did you miss the point that these are ACTIVE
> crossovers?

Not at all, but...

My comment applies equally to active or passive crossovers.

>That means that they go between the pre-amp
> and the power amps. One does not need a speaker connected
> or even the amplifiers connected in order to measure
> them.

You've missed the point - what crossovers sum to electrically is totally
irrelevant, other than as an intermediate result that is only interesting
when the response of the loudspeaker drivers is considered.

> You would be correct if these crossovers in
> question were the LRC kind that come between the amp(s)
> and the speaker(s).

Not at all.

Both active and passive crossovers of a given kind sum to the same response,
electrically.

Measuring the voltages across the drivers in a system with passive
crossovers is analogous to measuring the voltages across the drivers in a
system with active crossovers. Since the power amplifiers should have
essentially flat phase and frequency response through the crossover regions,
measuring the voltages at the output of an active crossover is analogous to
measuring the voltages at the output of the amplifiers or the voltages
across the drivers of a system with a passive crossover.

> But with active crossovers, you can
> perform frequency response tests, plot the
> crossover/rollover points, check distortion and even
> perform the sawtooth test that the OP mentions, below.

The point is that having a crossover that sums to flat response and $2.00
will get you a pretty fair cup of coffee at the Caribou across the street
from my house. Other than that, its a useless measurement. The desired
output of a crossover is whatever it takes to optimize the response of the
speaker system, nothing more and nothing less.

I can pretty well guarantee that if done right, the output of an ideal
crossover for a speaker system with real world drivers will sum to
*anything* but flat frequency and phase response.

Doesn't matter whether the crossover is active or passive.

Re: tweeter suggestion

Arny Krueger wrote:
> "Edmund" <sigm,fr eud2@freenet.de> wrote in message
> news:fnb5pt0go2@news2.newsguy,com
>> Arny Krueger wrote:
>>> "Edmund" <sigm,fr eud2@freenet.de> wrote in message
>>> news:fn610f012eg@news1.newsguy,com
>>>> Arny Krueger wrote:
>>>>> "Walker" <lbfoh@yahoo,com > wrote in message
>>>>> news:fmao3c01bqt@news1.newsguy,com
>>>>>
>> And unlike some other things
>> in high end audio, it is very well possible to measure
>> crossovers to see how they will sound.
>
> Absolutely not. Crossovers and speakers must work together very
> closely to produce the desired acoustical results.

Oh yes you are absolutely right about that, I assumed
perfect speaker units with perfect acoustic coupling.
So we can look at the effects of the crossovers only.
See what 4th order crossovers do to the signal and and it
is clear that that is not the way to go.

> If you are going to measure the
> results, you do it in the acoustical domain where accurate reliable
> measurements and equating measured results to sound quality is very
> difficult.

True
>
>> If I feed my crossovers with a sawtooth and put the low,
>> middle and high together again, guess what, I get the
>> exact same sawtooth back again. Show me the first other
>> brand crossover which does the same!
>
> I believe that two 6 dB/octave crossovers of the simplest
> design will do this.

In theory, yes, I have 1 and 2nd order and they do that IRL.
>
>> As I understand things, this is the only right way, if
>> one get anything else then the exact same input, it is
>> simply wrong.
>
> You have to look at the output of the speakers, not the output of the
> crossover.

I think one should look at every step in the proces and
see where things go wrong.
When the output of the crossovers is already disturbed
there is no way the speakers will make that any better,
only worse.

Edmund

Re: tweeter suggestion

On Jan 25, 6:42 pm, Edmund <sigm,fr e...@freenet.de> wrote:
> Arny Krueger wrote:
> > You have to look at the output of the speakers, not the output of the
> > crossover.
>
> I think one should look at every step in the proces and
> see where things go wrong.
> When the output of the crossovers is already disturbed
> there is no way the speakers will make that any better,
> only worse.

Simply not true. The output of a crossover that is properly
to the drivers will look anything BUT perfect, in fact, the
output may be quite "disturbed." But if the job of the crossover
is to correct for minimum-phase errors in the drivers, than
that crossovers output will be "disturbed" in a way that is
the precise complement of the "disturbance" of the driver.

> See what 4th order crossovers do to the signal
> and it is clear that that is not the way to go.

Again, simply not true as a generality. If you limit your
exploration to conventionally implemented crossovers
based on normal transfer functions that can be
implemented with passive ladder networks or their
active equivalents, then it is problematical. But there
is a huge specturm of responses and implementations
that are not conventionally followed that do not exhibit
these problems.

Take for example, the work done at B&O in the '70's.
The problem, once you understand the domain, is quite
straightforward. Take a first order network. The low-pass
transfer function (simplified for brevity) is:

Gl = 1/(s + 1)

while the high-pass function is

Gh = s/(s + 1)

Sum them together and we get:

Gl+h = [s/(s + 1)] + [1/(s + 1)]

Gl+h = (s + 1) / (s + 1)

Gl +h = 1

which indicates no "disturbance."

For the second order case, it's not so neat:

Gl = 1/(s^2 + ks + 1)

and

Gh = s^2/(s^2 + ks + 1)

and the sum

Gl+h = (s^2 + 1)/(s^2 + ks + 1)

which is most definitely not 1, i.e., the output is disturbed.

But that doesn't mean that a 2nd order network with
"undisturbed" output is not possible. All we need to do
is supply the missing portion of the output, which is:

G? = ks/(^s^2 + ks + 1)

And that turns out to be a 2nd order bandpass function.
We simply implement a third output of the crossover to
a third driver with the requisite bandpass function with
this transfer function, and we end up with a 2nd order
system whose complex transfer function is 1.

And we haven't even begun to explore the implementation
in the signal processing realm, where many of a
continuous signal domain implementation are removed
(like we can build non-causal filters, for example).

Re: tweeter suggestion

dpierce.cartchunk.org@gmail,com wrote:
> On Jan 25, 6:42 pm, Edmund <sigm,fr e...@freenet.de> wrote:
>> Arny Krueger wrote:
>> > You have to look at the output of the speakers, not the output of the
>> > crossover.
>>
>> I think one should look at every step in the proces and
>> see where things go wrong.
>> When the output of the crossovers is already disturbed
>> there is no way the speakers will make that any better,
>> only worse.
>
> Simply not true. The output of a crossover that is properly
> to the drivers will look anything BUT perfect, in fact, the
> output may be quite "disturbed." But if the job of the crossover
> is to correct for minimum-phase errors in the drivers, than
> that crossovers output will be "disturbed" in a way that is
> the precise complement of the "disturbance" of the driver.

Well that is kind of a long shot isn't it?
Phase errors in drivers occur due to impedance
changes in frequency, Changes you find any driver that
exactly needs to be compensated by the
errors of a let say a 4th order butterworth crossover
are significantly less then winning the major price in
the lottery four times in a row.
That is not a realistic scenario at all.
>
>> See what 4th order crossovers do to the signal
>> and it is clear that that is not the way to go.
>
> Again, simply not true as a generality. If you limit your
> exploration to conventionally implemented crossovers
> based on normal transfer functions that can be
> implemented with passive ladder networks or their
> active equivalents, then it is problematical. But there
> is a huge specturm of responses and implementations
> that are not conventionally followed that do not exhibit
> these problems.

As I said before I don't know what is possible with digital
crossovers but for analoge stuff, IFAIK it is a given they
introduce temporal problems. ( no doubt digital
crossovers introduce other problems )
I have a simulation here from a 4th oder butterworth which
gives a combined output with three “tops” by an input
of a sawtooth with only TWO tops.
That cannot by right by any standard.

>
> Take for example, the work done at B&O in the '70's.
> The problem, once you understand the domain, is quite
> straightforward. Take a first order network. The low-pass
> transfer function (simplified for brevity) is:

Sorry I am not the developer and I do not know nearly
enough to show you any temporal effect of this.
( not to mention my limitations with the English language )
I do believe Temporal Coherence does, hence there name.
I have a computer simulation of there crossover and I have seen
a photograph from a oscilloscope. For me it is magic but both are
exactly the same, so I think these guys know what they are doing.
Not the least, it sounds really good too :-)

Edmund

Re: tweeter suggestion

"Edmund" <sigm,fr eud2@freenet.de> wrote in message
news:fnfkal02oe0@news1.newsguy,com
> dpierce.cartchunk.org@gmail,com wrote:
>> On Jan 25, 6:42 pm, Edmund <sigm,fr e...@freenet.de>
>> wrote:
>>> Arny Krueger wrote:
>>> > You have to look at the output of the speakers, not
>>> the output of the > crossover.
>>>
>>> I think one should look at every step in the proces and
>>> see where things go wrong.
>>> When the output of the crossovers is already disturbed
>>> there is no way the speakers will make that any better,
>>> only worse.

>> Simply not true. The output of a crossover that is
>> properly to the drivers will look anything BUT perfect, in fact,
>> the output may be quite "disturbed." But if the job of
>> the crossover is to correct for minimum-phase errors in
>> the drivers, than that crossovers output will be
>> "disturbed" in a way that is the precise complement of the "disturbance"
>> of the
>> driver.

> Well that is kind of a long shot isn't it?

Yes.

> Phase errors in drivers occur due to impedance
> changes in frequency,

The acoustical transfer function of the speaker as enclosed is the major
problem.

Using an electronic crossover minimizes the effects of the impedance curve
of the driver.

> Changes you find any driver that
> exactly needs to be compensated by the
> errors of a let say a 4th order butterworth crossover
> are significantly less then winning the major price in
> the lottery four times in a row.
> That is not a realistic scenario at all.

The point is that the crossover needs to have response characterstics that
are closely aligned with the characteristics of the driver.

>>> See what 4th order crossovers do to the signal
>>> and it is clear that that is not the way to go.

>> Again, simply not true as a generality. If you limit your
>> exploration to conventionally implemented crossovers
>> based on normal transfer functions that can be
>> implemented with passive ladder networks or their
>> active equivalents, then it is problematical. But there
>> is a huge specturm of responses and implementations
>> that are not conventionally followed that do not exhibit
>> these problems.

> As I said before I don't know what is possible with
> digital crossovers

Digital crossvers can easily and economically do things that are very
inconvenient and expensive in the analog domain.

> but for analoge stuff, IFAIK it is a
> given they introduce temporal problems. ( no doubt digital
> crossovers introduce other problems )
> I have a simulation here from a 4th oder butterworth which
> gives a combined output with three “tops” by an input
> of a sawtooth with only TWO tops.
> That cannot by right by any standard.

It's not necessarily more right or wrong that any number of possible
waveforms, except for a facsimile of the origional.