What do you think of passive crossovers?

[john k...]

I would like to add one additional comment regarding active vs passive x-os. The second major argument that actives are superior to passives is based on the idea that voltage transfer function of the passive crossover will be affected by changes in the VC impedance due to temperature changes, and with an active crossover it will not be. This is true, providing the amplifier's output Z is small compared to the driver's Z. But again the implication is that the unchanging transfer function of the active crossover is better. Maybe, maybe not. The point is that the force accelerating the drivers radiating surface is a function of the current through the VC, F = Bl x I = Bl x Vapplied/Zdriver. Thus the actually transfer function of the radiating surface's motion is a function of Vapp/Zd, not just Vapp. Thus holding Vapp constant as the VC Z changes due to temperature does not imply that the acoustic transfer function of the filtered driver will remain constant.

There are lots of other issues but I think enough worms have gotten out of the can for now.

[john k...]

Originally Posted by abraxalito

I'm not saying its a bad thing. I'm just avoiding it in my own designs because I've no reason to believe it helps.



Nope, I don't see. Where is it in the question please?



Just cut and paste this 'assertion' that you're asserting I've made. I predict you will fail to do so because I can see none. So this is a red herring, a distraction.



Who's making the observation here? Certainly not you - you're inferring something that's in reality absent. I'm observing you doing so



The first part of your straw man is indeed correct. I am drawing attention to an effect I prefer myself to avoid, by going active. Presumably you have good reason not to avoid it, to even advocate that others don't avoid it themselves. Its this reason that I'm curious about. If it turns out to be good (by which I mean to say helpful, an improvement), when I've digested your arguments then I can introduce it into my designs without any difficulty.

<snip>



So then why are you making it out to be a 'devil', even though its a devil that you choose not to avoid?

As I said earlier, I have no preference. If you look at my speakers you will see I offer an active (analog or digital)/passive hybrid as well as fully active (active between woofer and mid is pretty much required for a dipole if you want to maintain the midrange efficiency).

Re the other comments, I was not pointing at you (which I why I said I'm not asking you to provide proof, sorry if you took it wrongly), but the argument about damping is one that active crossover proponents commonly use as an example of all that is bad about passives and it just really isn't the case. It is those proponents that imply that lack of damping is the devil. I'm personally petty neutral on active vs passive.

[KSTR]

Originally Posted by john k...

[...] The electromotive damping is anything but linear to start with. We usually see this simplified as F = (Bl)^2 /Re x U because, as I said earlier, the effect of damping is only significant around the driver's resonance were wLe is small compared to Re. [Note: Even if you want to make the argument that the damping force is present even at high frequency, you can see that the associated nonlinearity in that force can only contribute to nonlinear distortion. Thus, with a 2nd order passive LP crossover it could even be argued that the rising Zeff as the x-o frequency is approached actually reduces the effect of motor generated nonlinear distortion, due to reduced (nonlinear) electromotive damping force compared to an active crossover.]

Exactly my view, too. To make active really any better than passive one must make full use of optimized damping. I found there is often a "best" drive impedance profile that is sort of a (graphical) inverse of driver impedance. Amp and driver must team-up in a very literal sense. "Best" is of course open to discussion, it's again preferences and compromises as usual.

"Driver impedance" in its common usage itself is a problematic terminus IMHO, because we lump in two mechanisms (somebody mentioned already), the true electrical impedance of the coil at the given DC excursion position in the gap (clamped cone) plus the microphonic voltage from the coil (gives a voltage proportional to cone velocity) which pickes up any motion no matter what caused it. Both are distorted, but when we do feed as little as possible of this sensor voltage back into the system, chances are better to get lower overall distortion, except at unstable / ill-defined points of mechanical driver resonance. Geometrical resonances (standing waves) tend to be stable and can be handled with signal pre-correction.

- Klaus

[speaker dave]

Originally Posted by john k...

Another point is that direct connection to the amplifier maximizes the effect of the back emf since it maximizes the reverse current. Again, that is always touted as a benefit by those who subscribe to the active crossover school. However, maximizing the reverse current also maximizes the distortion associated with the motor/generator nonlinearity. Hawksford has written about this and how it and be countered by current drive. Current drive, of course, required a high output impedance which then basically reduced the electromotive damping to zero everywhere.

I did a series of test some years ago to look at distortion vs. amplfier output impedance. I saw no connection between output Z and LF distortion.

Of course changing drive impedance changes Lf response considerably and so changes excursion vs. frequency. To get around this and insure an apples and apples comparison I used one of the old B&K oscillators with the compressor (feedback loop) and connected to keep the output SPL flat (showing my age here). In fact the compressor was tracking the fundamental so that the non distorted "1st harmonic" component was dead level.

I achieved high and low Z with one of the old Fisher "Z Matic" tube amps. Second and third (I might have plotted to 5th) were identical as long as the compressor held the fundamental output curve flat. This was true all through resonance +- and Octave or so.

The only place where output Z made any difference to distortion was in the midrange area where there was high 2nd harmonic due to flux modulation. High Z lowered the distortion.

I don't know if that agrees with Hawksford but I still have the measurements.

David S.

[wolf_teeth]

Originally Posted by tinitus

hey Wolf, impressive and tricky looking xo

had to count my xo components
only 16, and about half of them 'just' resistors

well, there's a couple free hanging 'non active' resistors, from experiments
but I hesitate to completely remove them, yet
because I know from many other experiences that something changes when they are removed
even if they are only connect at one end
sound silly, eh

A lot of what is pictured is superfluous, being that this 'case' was an experiment in swapping capacitors. The 'cap-banks' to one side of the switchboard is only applying 2 different positions to the circuit.

Later,
Wolf

[wolf_teeth]

Originally Posted by vacuphile

Wolf, Tinitus, please think about our Planet and where all that copper went!

Vac.

You should think about your copper EI-core/toroids, aluminum amp-cases, and all the metals involved in your amplifiers.

Same difference, bud.

Later,
Wolf

[wolf_teeth]

Originally Posted by bobodioulasso

In a passive crossover, without a zobel, the corner frequency is shifted by impedance variations.

Zobels in the true sense of the word are not always required. You can actually have a lower net impedance of the system using them to minimize Le of the multiple drivers in 3-ways.

You optimize the xover from the impedance/FR/phase data, so you can't just say, "it's this impedance at this frequency, so I need this value L and C for this xover Fc". The Fc (corner) is selected by the components used in reference to the impedance and the dB level of the FR at that point. The Fc is where you put it with that in mind, not "shifting" if you don't do something else.

Later,
Wolf

[DBMandrake]

Originally Posted by wolf_teeth

Zobels in the true sense of the word are not always required. You can actually have a lower net impedance of the system using them to minimize Le of the multiple drivers in 3-ways.

Sometimes the rise in impedance of a driver at high frequencies due to its inductance will be transformed by the crossover into a decrease in input impedance - in this case adding a zobel network to flatten the impedance of the driver will actually increase the input impedance of the crossover, as in the example I cited earlier with the 3rd order butterworth driving a tweeter.

Whether adding zobels increases or decreases the net input impedance presented to the amplifier is highly dependant on the topology of the rest of the crossover network, and how it interacts with the drivers impedance.

[DBMandrake]

Originally Posted by speaker dave

The only place where output Z made any difference to distortion was in the midrange area where there was high 2nd harmonic due to flux modulation. High Z lowered the distortion.

I don't know if that agrees with Hawksford but I still have the measurements.

What level of improvement were you measuring in midrange distortion ? I presume the tests were done with a driver without a copper shorting ring or other means of achieving very low inductance, that would tend to offset reduction in distortion achieved through current drive ?

I was recently thinking about the possibility of a pseudo-current source network for a midrange driver, and whether there would be a worthwhile improvement in distortion.

If we assume a driver that is quite sensitive and needs more than 6dB of attenuation relative to the woofer, the idea is to provide all the attenuation only via series resistance to present as high as possible source impedance to the driver.

The usual shunt resistor of an L-Pad would be placed before the series resistor instead of after it, and the zobel (if needed) would also be placed before the series resistor and scaled the correct amount to still present a flat impedance to the output of the crossover at the high frequency end.

Of course that imposes two issues -

One would be a rise in response of the driver at higher frequencies relative to pure voltage drive, but that could be corrected for in the filter section before the series resistor, or depending on the driver a gradual rise in response may actually be desirable and beneficial if the driver already has some natural roll off. (The latter is likely with the driver I have in mind)

The second issue is controlling the fundamental resonance of the driver, which will be greatly increased in amplitude with the higher source impedance. This will almost certainly require an RLC compensator connected directly to the driver to accurately correct this, but chances are depending on the driver and midrange enclosure size an RLC compensator already would have been required to flatten the fundamental resonance anyway - all that changes is that the required values will be different.

So essentially we have the same number of components (assuming both zobel and resonance compensator were wanted in the more conventional configuration) but the question is, is it worth the trouble ? I guess it depends entirely on the individual driver, both in terms of how much attenuation it requires (and therefore how closely you can approximate current feed) and also in terms of how much L(e) modulation it has.

I'm also assuming it would be fairly pointless for any driver with a copper shorting ring.

[RickDangerous]

Originally Posted by Charles Darwin

I on the other hand have compared and converted a few passives and actives, from Dynaudio (DynaudioAcoustics BM5 passive and active plus the equivalent consumer model, the differences were two different passives and one active. Drivers and cabs remained the same.) to my own Tannoys.
To my ears passives don't get close to actives, with regards to bass reproduction particularly.
And for the Tannoys the improvement gained from going active is far greater than in passive mode going from a crappy Arcam Alpha 8 amp to a rather decent MC2 Audio MC450.

Is that the same debate as active crossover vs passive crossover, though? Do active speakers necessarily have active crossovers ? I don't think so but i could be wrong. An active speaker has the amp inside it but it could be built with exactly the same crossover (ie passive) as in the passive version. The active version usually has one amp per driver but that equates to bi-amping / tri-amping, not to having an active crossover.

As for the Dynaudio BM5a, they are absolute crap compared to the passive version with a good amp. They use car audio electronics and the passive version on a cheap used Bryston sounded waaaayyyy better.

I could be wrong but i think you're confusing active speakers with active crossovers.