What is even the point of passive crossovers in 2021?

[danadam]

I presume that you meant the acronym: FUBAR (FU BAR)? (the acronym SNAFU [SNA FU]) might also be fitting.

I meant it as https://en.wikipedia.org/wiki/Metasyntactic_variable, but yes, it is possible that this particular one comes from FUBAR: https://en.wikipedia.org/wiki/Foobar

 

[EJ3]

I learned the term before computers had anything to do directly for most people (SHURE CARTRIDGE'S & PONTIAC HIGH PERFORMANCE CAMSHAFT DESIGN WAS HAPPENING) but few knew. FUBAR: Fowled (substitute another "F" word if you like) Up Beyond All Repair (or "Reality").
SNAFU: Situation Normal All Fowled (substitute another "F" word if you like) Up. Those are what I was taught. Thanks for the enlightenment about things that I had never heard of. Interesting how pervasive some acronyms become.

 

[antcollinet]

I learned the term before computers had anything to do directly for most people (SHURE CARTRIDGE'S & PONTIAC HIGH PERFORMANCE CAMSHAFT DESIGN WAS HAPPENING) but few knew. FUBAR: Fowled (substitute another "F" word if you like) Up Beyond All Repair (or "Reality").
SNAFU: Situation Normal All Fowled (substitute another "F" word if you like) Up. Those are what I was taught. Thanks for the enlightenment about things that I had never heard of. Interesting how pervasive some acronyms become.

Perhaps there is not a hard and fast defintition, but my understanding of the R in Fubar is "recognition". Also the F has to be the proper F word, else what's the point. (But I'm British, and I think we use it with more reckless abandon than over the pond)

 

[antcollinet]

I understood it as you state here. Perhaps because I have a wife who has a third-eighth grade level of English & have to interpolate a lot of things.

So what is my excuse (engineering** degree educated ENGLISH person)

**Actually that's the problem right there. "I used not to be able to spell engineer - now I are one"

 

[EJ3]

Perhaps there is not a hard and fast defintition, but my understanding of the R in Fubar is "recognition". Also the F has to be the proper F word, else what's the point. (But I'm British, and I think we use it with more reckless abandon than over the pond)

I was being polite due to I don't like my stuff being canceled (and due the nature of the communications on this site being a bit less emotional & [hopefully] more intellectual than ] My mother learned the King's English there shortly after WWII. She is Austrian from Salzburg (I was concieved in Charleston, South Carolina USA but born in Salzburg) & a survivor of Hitler's power grab. My father was a US Merchant marine in the Atlantic at the age of 17 during that WWII era.

 

[Tangband]

If using active filters and amps, I think the ideal is to use impedance matching between amp and drivers, current drive etc. You can win in distortion.

Thats very true, but not common in active loudspeakers that one can buy in a store.
sometimes a tweeter may sound less compressed if driven with high impedance (with sound levels compensated ) .
Tweeters with metal domes are often much less prone to this problem, such as the ones used in, for example, Genelecs.

faktiskt.io • Visa tråd - effekt av uppvärmning talspole-tråd i diskant?

 

[Lambda]

Impedance matching the speaker to the source is a bad idea.
You don't want to maximize the power transfer like in a RF system.

it is conventionally a voltage driven device.
And the ideal amplifier impedance is zero.

tweeter may sound less compressed if driven with high impedance

What dose sound less compressed even mean to you?
more dynamic?

 

[Tangband]

Impedance matching the speaker to the source is a bad idea.
You don't want to maximize the power transfer like in a RF system.

it is conventionally a voltage driven device.
And the ideal amplifier impedance is zero.


What dose sound less compressed even mean to you?
more dynamic?

Below the EMF area ( subwoofers ) a zero impedance is ideal, true.
But above the EMF area one has to try , measure and listen to see if the distortion eventually goes down with a somewhat higher impedance . What Im talking about is to optimise the loudspeaker system.
Some tweeters benefit from driven with a somewhat higher impedance ( 2-3 Ohm ) soundwise, ( with the level compensated ) , other tweeters dont .

 

[Thomas_A]

Impedance matching the speaker to the source is a bad idea.
You don't want to maximize the power transfer like in a RF system.

it is conventionally a voltage driven device.
And the ideal amplifier impedance is zero.


What dose sound less compressed even mean to you?
more dynamic?

AES E-Library » Synthesis of Loudspeaker Mechanical Parameters by Electrical Means: A new Method for Controlling Low-Frequency Loudspeaker Behaviour

A method for extending bass response and lowering distortion in louspeaker systems, which differs from both equalized and servo or feedback systems, is described. It can be seen as an extension of the technique to increase loudspeaker damping by giving the driving amplifier a negative output...

www.aes.org

And

Distortion example: Scan-Speak | Current-Drive - The Natural Way of Loudspeaker Operation

How distortion is reduced by current-drive in a dome tweeter

www.current-drive.info

 

[Reed]

I think the industry and people needs a new kind of thing that can make every passive speaker into an Active speaker...

But a cloud data of each speaker its very important and the device must auto configure itself for the speaker

Hifiberry makes kits for vintage B&O speakers. Active crossover. Can easily me used in different speakers. You’d need to BYO crossover slopes. https://www.hifiberry.com/wp-content/uploads/2018/06/s35-guide.pdf

 

[Tangband]

AES E-Library » Synthesis of Loudspeaker Mechanical Parameters by Electrical Means: A new Method for Controlling Low-Frequency Loudspeaker Behaviour

A method for extending bass response and lowering distortion in louspeaker systems, which differs from both equalized and servo or feedback systems, is described. It can be seen as an extension of the technique to increase loudspeaker damping by giving the driving amplifier a negative output...

www.aes.org

And

Distortion example: Scan-Speak | Current-Drive - The Natural Way of Loudspeaker Operation

How distortion is reduced by current-drive in a dome tweeter

www.current-drive.info

Thank you Thomas for the links.
Even with a midbass or midrange driver, the distortion can sometimes be significantly lowered with current drive.

Distortion example: Infinity | Current-Drive - The Natural Way of Loudspeaker Operation

How distortion is drastically reduced by current-drive in a typical 3" midrange driver

www.current-drive.info

 

[Lambda]

AES E-Library » Synthesis of Loudspeaker Mechanical Parameters by Electrical Means: A new Method for Controlling Low-Frequency Loudspeaker Behaviour

A method for extending bass response and lowering distortion in louspeaker systems, which differs from both equalized and servo or feedback systems, is described. It can be seen as an extension of the technique to increase loudspeaker damping by giving the driving amplifier a negative output...

www.aes.org

And

Distortion example: Scan-Speak | Current-Drive - The Natural Way of Loudspeaker Operation

How distortion is reduced by current-drive in a dome tweeter

www.current-drive.info

Even with a midbass or midrange driver, the distortion can sometimes be significantly lowered with current drive.

Negative impedance drive is the opposite of Current drive. Current drive is basically Infinite impedance drive.
Both is not the same as Impedance Matching.
Both can be done in DSP assuming the parameters of the speaker are well Known or the Amplifier has a current sens Like Powersoft M-Force.

A Speaker can be desigend to be current driven or Voltage driven. later one is the standard but in the end the speaker don’t know if its current or voltage driven.
At the speaker terminals you can measure a voltage. If this voltage is played back to the terminals it will Result in the same speaker moment, current and sound as before.
If you Record the Current going in the Speaker and play it back to the terminals it will Result in the same speaker moment, Voltage and sound as before.

(this is assuming the speaker is not heating up significantly)

 

[Thomas_A]

Negative impedance drive is the opposite of Current drive. Current drive is basically Infinite impedance drive.
Both is not the same as Impedance Matching.
Both can be done in DSP assuming the parameters of the speaker are well Known or the Amplifier has a current sens Like Powersoft M-Force.

A Speaker can be desigend to be current driven or Voltage driven. later one is the standard but in the end the speaker don’t know if its current or voltage driven.
At the speaker terminals you can measure a voltage. If this voltage is played back to the terminals it will Result in the same speaker moment, current and sound as before.
If you Record the Current going in the Speaker and play it back to the terminals it will Result in the same speaker moment, Voltage and sound as before.

(this is assuming the speaker is not heating up significantly)

Yes that is why I separate impedance match and current drive. Point is whether you can decrease distortion with amp-driver optimization rather than just using traditional amps. True or not? It is also described in several books e g Martin Collom.

 

[Lambda]

True or not

Sure you can but you also changing frequency and phase response at the same time
you can do the same with a High Efficacy low impedance Class D amp, sensors and software as demonstrated by Powersot.

Driving with Current (and measuring voltage) has the advantage that it can cancel out the effects of increasing resistance by the temperature rise in the voice coil or cables.

 

[Thomas_A]

Sure you can but you also changing frequency and phase response at the same time
you can do the same with a High Efficacy low impedance Class D amp, sensors and software as demonstrated by Powersot.

Driving with Current (and measuring voltage) has the advantage that it can cancel out the effects of increasing resistance by the temperature rise in the voice coil or cables.

Ok thanks. Frequency response and phase could however be dealt with by other means. My original point was when going active why not use all benefits that are possible? More complicated for most DIYers but for speaker companies it should certainly possible to design.

 

[Wolf]

Passives forever!!!! (Except in subwoofers that are not bandpass types.)

I also design passives, and just love the process.

 

[KSTR]

A Speaker can be desigend to be current driven or Voltage driven. later one is the standard but in the end the speaker don’t know if its current or voltage driven.
At the speaker terminals you can measure a voltage. If this voltage is played back to the terminals it will Result in the same speaker moment, current and sound as before.
If you Record the Current going in the Speaker and play it back to the terminals it will Result in the same speaker moment, Voltage and sound as before

Very true and I've been stressing this point many times as well. But there's a "got'cha":
The damping of the cone to external excitation (including its own motor distortion products) only depends on the terminal impedance it sees. With current drive, the driver only undergoes mechanical damping. If it's a multi-way the terminal impedances for the drivers will be significantly higher with current drive, resulting in a likewise lack of damping.

I've written about it before, the sum of the driver's DC coil resistance (Rdc) and any external impedance defines the mode of operation of the driver which is what actually happens when you change electrical damping.

Some interesting points in the continuous mode space are:
- Zout (amp output impedance) = -Rdc. This is full motional feedback, the driver operating in full velocity control. The VC's "back-EMF" voltage is the sensed quantity.

- Zout = infinite (current source) : This is zero motional feedback, the driver operating in fully force-steered mode. Note the difference between steering and controlling, only controlling means you monitor the output in some way and correct the signal to expected output.

- Zout = 0 (standard voltage drive) : Reduced motional feedback, hence a mix of a reduced motion control and some plain force injection.

Zout = 0 is a arbitrary choice from the driver's view. It's more meaningful to look at the required terminal impedance to achieve a Bessel or Butterworth target of the cone excursion function, a low-pass. We want a step response that is as fast as possible but does not overshoot or even ring, this will give us the cleanest recovery behavior from excursion errors. Depending on the driver and cabinet size, this "best candidate" damping resistance might result in a high positive Zout (for Qts=0.2 variety of drivers) or a negative one (for very weak magnets), and when you're just lucky with zero even.

Negative output impedances are problematic in practice for number of a reasons I could explain, bottom line is you should no go below -Rdc/2 and best avoid it altogether (exceptions nonwithstanding).

Higher up in frequency high drive impedance almost always gives better results so the basic output impedance profile is pretty much set: Make it as high as you can afford by your enclosure type and design, without under-damping the driver. For many drivers in reasonable cabinets, Zout=0 around resonance is already close to ideal as that is what the drivers have been designed for in the first place.

 

[Schollaudio]

Yes!
I design passive xo speakers, self powered speakers and speakers with one or more separate rack mount amplifiers.

If you were a speaker designer, one immediate question would be "is my aim for people that have existing power amplifiers?"
Does my audience care about latency or FIR or DSP in general and if so is that philosophical or based on an actual problem in use? Some are kinda phobic about data bit's and such.

An observation with efficient horns is that often the amount of correction to get to flat is less with a passive xo. A Voltage source loaded by a changing R produces a larger amplitude variation than the matched impedance source.

A misconception about passive crossovers is that they are inefficient / lossy.

They can be if there is a lot of current passing through resistors but the reactive parts normally have small losses.

The problem / reason it seems like it is lets say you had a 300W woofer, a 100W mid range and 25W tweeter with an all passive xover and wanted to reach that power with an imaginary signal peak.
To do this, you need an amplifier that can swing the full Voltage needed for the 300W woofer plus the full Voltage needed to reach 100W to the mid-range plus the Voltage needed to produce the 25W for the HF.

Since that's a tall order often when one changes to full active they hear a more dynamic sound, not having instantaneous amplifier Voltage clipping anymore (which a cycle or two at a time is normally inaudible as clipping) .

It has been said "what we need is a perfect 10W amplifier that can put out peaks of 1000W"
Best
Tom

Listening to CDs, DVD, Bluray, SACD and steaming I'm not bit or opamp phobic but find DSP hasn't arrived yet for me.
sure they have tremendous functional advantages but going from HDMI to DAC to volume IC to 5532 opamp sounds better (to me) than adding the DSP to the chain. Direct mode.

Switching over to a stupid old Pioneer universal to a simple two stage preamp sounds better still. (all with the same poweramp and speakers)

Also, using passive crossover parts at 400hz sounded better to me than a prosound crossover and that's with using the same 250watt and both lows and mids. of course the midrange\high frequency amp needs to produce much less power.

In home environments maximizing mid and tweeter wattage isn't needed, in fact its best to operate those well under rated power. Based on power distribution by spectrum with crossovers above 400hz 300, 100 and 25 watt power distribution is fine with one 200watt amp. And that's based on music with 10DB dynamic range which would run at 20watts with 200 watt peaks and in at least my systems the neighbors would enjoy some really good music. LOL Low dynamic range music might cause some tweeter failures. but active, vs passive isn't going to help with that. In prosound this is all different the designer wants to consider power handling across the spectrum and a single typical line array box will have 2 woofers, 2-4 or more mids and 4 or more tweeters all powered with properly rated amplifiers with safe overhead.

In the home, sat\sub systems are very useful and biamping a must. A well designed closed box with 2nd order high pass to 4th order lowpass can save the high pass amp lots of power. I don't know of any commercial analog systems that match the crossover slope to the midbass alignment slope to do this. It's either DIY or DSP.

 

[Digital_Thor]

Passives are just nice and simple. Some old off the shelf Dali or B&W is easy to hook up to some fine integrated like a PM7000 from Marantz or NAD C3xx. Very cheap and with a bit of simple EQ they also play nicely in many rooms.

Actives simply lift everything to a higher level. Suddenly we have tons of options and manipulations in the signal path, which can transform an "ok" sounding system to be one that really sound great.

Still - nothing can compensate for poorly designed speakers. Well, they can be mended a bit with some dsp-trickery - but basic guidelines for cabinet constructions and driver choice, still needs to be followed, to achieve good sound.