[Porcupine]

I've read that pulse rise time is limited and made many times longer in a speaker driver if it is filtered through a low-pass crossover. Does this have a significant effect on transient response, or is the transient response of the driver itself the dominant factor in all cases?

I've recently been looking at VMPS and their ribbon midrange - ribbon tweeter designs used in most of VMPS speakers. I've noticed that the ribbon tweeters tend to be crossed very high, around 6 kHz to 10 kHz depending on the model. I worry that this would produce inferior transient response in the 3 kHz to 6 kHz range, compared to other companies' speaker designs that cross their ribbon tweeters to conventional drivers around a more conventional 3 kHz point. (Though of course, VMPS transient response would be superior in the 150 Hz to 3 kHz range, compared to any conventional midrange).

Does VMPS pass their ribbon mids through a passive low-pass crossover filter?

I was thinking maybe a 1/2-way design would be ideal, where the ribbon mid rolls off naturally where the tweeter comes in, and doesn't use a low-pass filter.

[warnerwh]

Actually I believe all the Vmps speakers with the FST tweeters cross about 6,900hz.  When it comes to transient response it's no less than awesome, period.  You can't compete with these drivers with any cones made I'm aware of.  A line array may get you part way there but I'm not so sure there either.  You obviously haven't heard any Vmps speakers.  The transient response is one major contributor of making them sound so very real.

A couple of reasons are the midrange panels I think weigh about 1.25 grams, a magnitude less than any cone.  They are also driven over their entire surface by neodymium magnets. The neodymium magnets also have the advantage of being able to focus their magnet energy in a small area where strontium ferrite radiates it's energy all around.  The neodymium is also about 8 times stronger.

On the FST ribbon where the diaphram only weighs fifteen thousands of one gram you also have the neodymium magnets.  15/1000
s of a gram is so light as to make some feathers appear heavy. A dome tweeter typically weighs 100 times or more as much.  You tell me which one is likely to have the better transient response.

 I can say the FST tweeter absolute trounces any dome I've heard for realism. I also read that one of the reviewers felt it was the best treble he's ever heard in any speaker.  At minimum Vmps speakers are a must hear in their respective price categories because they are a true bargain.

The last thing you want to do is try to think you know about speaker design. I was going to build a world class speaker and learned enough to know it's not for beginners or even people with some years of experience.  Go over to the madisound.com or partsexpress.com websites where there's many diy speaker builders and ask. There's more to it than straight engineering and theory.

[_scotty_]

In general terms the risetime of the driver,in this case a tweeter,is determined by its high frequency extension which for the purpose of this explanation will be its 3dB down point;[-3dB at 50kHz for example].
The filter used to keep unwanted lower frequencies that are outside the physical capabilities of the tweeter should not affect the drivers HF 3dB down point or its risetime. In the simplest passive network a series capacitor
of the correct value is inserted between the amplifier and the tweeter.
The capacitor will have a much higher 3dB down point than the tweeter perhaps as much 3 to 5 times higher or more.  The physical parameters
of the driver will dominate its risetime in this situation.  In the case of a
midrange driver,once again its HF 3dB down point comes into play as does its physical dimensions compared to the wavelength of the highest frequencies
we wish to reproduce.  As the frequency increases the wavelength decreases
until the wavelength reproduced equals the horizontal dimension of the ribbon, past this point the driver becomes increasingly beamier and exhibits poor horizontal dispersion.  If the midrange, due to its physical makeup, has a lower HF 3dB downpoint than a tweeter it follows that its transient response and risetime are also slower than a tweeter.
If this is the case than it makes very little sense to allow it to reproduce
frequencies that are more accurately reproduced by a tweeter and  a good case can be made to limit its operational bandwidth via a crossover to frequencies that it is better suited to.  Ideally we would listen to a single
massless driver that would reproduce all of the frequencies associated with the music we are interested in. In the real world we have to divide the signal
among drivers that are physically capable of reproducing that portion of the music spectrum given to them. It is in the physical reassembly of the original waveform that the key to accurate impulse response resides.
  You have to decide by listening to the VMPS speakers whether or not
their pulse accuracy or transient response satisfies your requirements.
The crossover is at 6kHZ or 10kHZ in some models to keep the high pass
network as simple as possible.
This also has the advantage of keeping lowpass network on the midrange
simple.  
  John Casler can probably answer other aspects of your question in greater depth. I hope this information is helpful,Scotty

[Porcupine]

warner, duh I know all those things already. You don't need to spout all that stuff about why ribbons are great. I already know all those things you said. And regardless of whether I did or didn't, I have heard a fair share of ribbon drivers in real life and have always liked them. Perhaps I should mention that EVERY speaker I own has a ribbon tweeter? If anything this is one of my weaknesses because I am not well-rounded in my hearing experience of different types of hifi speaker. Too early on in my life I was introduced to ribbon tweeters and have never ever departed. I haven't had a chance to hear any top-of-the-line metal domes in my home, etc.

I have no doubt about the transient response capabilities of the current true ribbon tweeter used by VMPS. If you read my original post more carefully you would see I am worried about how the VMPS Neodymium 8" mag planar *midrange* will compare against, for example, Bohlender-Graebener's 3" mag planar Neo3PDR tweeter in the 3 kHz to 7 kHz range or so. I'm not comparing VMPS to conventional drivers. I'm comparing it to other ribbons so that it has some real competition.

> The last thing you want to do is try to think you know about speaker design. I was going to build a world class speaker and learned enough to know it's not for beginners or even people with some years of experience.

I totally agree. Way too much is involved. Why do you think I came to talk on this forum? I learned of VMPS speakers and thought Wow maybe that's almost exactly what I want! Maybe I should come and ask around and try to find out more info on their speakers. Maybe if I make some suggestions about what I'd like to see in a speaker, someone else like VMPS would take it into consideration and make something similar. Because there is no way I can do it myself.

> In general terms the risetime of the driver,in this case a tweeter,is determined by its high frequency extension which for the purpose of this explanation will be its 3dB down point;[-3dB at 50kHz for example].

This totally makes sense to me. This is what I was thinking from before, thank you for confirming.

> The filter used to keep unwanted lower frequencies that are outside the physical capabilities of the tweeter should not affect the drivers HF 3dB down point or its risetime. In the simplest passive network a series capacitor of the correct value is inserted between the amplifier and the tweeter.

Agree totally. But what you just described is a highpass filter. A highpass filter indeed does not affect transient response. I am not worried about the transient response of the tweeter.

I am worried about the LOWPASS filter used on the midrange planar element. Probably an inductor, or something more complicated. As you have said, that would limit the pulse risetime of any signal coming from the midrange roughly according to the crossover frequency.

> As the frequency increases the wavelength decreases until the wavelength reproduced equals the horizontal dimension of the ribbon, past this point the driver becomes increasingly beamier and exhibits poor horizontal dispersion.

A separate topic, but this is good information thanks. I think it may partially answer some questions I posed in another thread.

> If the midrange, due to its physical makeup, has a lower HF 3dB downpoint than a tweeter it follows that its transient response and risetime are also slower than a tweeter.

True, but the current 8" mag planar mid on the VMPS probably extends past 7 kHz right? I worry that by cutting it off at 7 kHz that worsens the pulse risetime and transient response of the mid driver from what it *could* have been if left unfiltered at the high end.

> If this is the case than it makes very little sense to allow it to reproduce
frequencies that are more accurately reproduced by a tweeter and a good case can be made to limit its operational bandwidth via a crossover to frequencies that it is better suited to.

Hmm. Possibly. But crossovers make me paranoid in general they do all sorts of weird things. Plus it might still be better to crossover the high-freq cutoff of the midrange even higher to get quicker risetimes out of it, right? I think if a way could be found to leave the midrange uncrossed at the high-end that would be best. (Maybe the VMPS speakers are already this way?)

> Ideally we would listen to a single massless driver that would reproduce all of the frequencies associated with the music we are interested in. In the real world we have to divide the signal among drivers that are physically capable of reproducing that portion of the music spectrum given to them.

Speaker basics. Yes I know these things already.

> It is in the physical reassembly of the original waveform that the key to accurate impulse response resides.

Exactly. If the high frequency crossover (both high-pass and low-pass) worked idealistically the midrange and tweeter would reassemble into a single driver in perfect unison. Thus even if the midrange has a limited pulse risetime due to the lowpass filter on it, the tweeter's output should contain the necessary waveform data to "add back" to the midrange to recreate the original pulse shape and risetime exactly. But does this actually happen? I don't think so, not even with the best crossovers. That's why I think it *may* be important to leave the midrange element uncrossed at the high freq end if possible. But maybe it doesn't matter that much in reality. That's why I'm asking, to see what others have to say.

> You have to decide by listening to the VMPS speakers whether or not
their pulse accuracy or transient response satisfies your requirements.
The crossover is at 6kHZ or 10kHZ in some models to keep the high pass
network as simple as possible.

Yeah, if I had a VMPS here of course I would listen to it.

> I hope this information is helpful,Scotty

Your information was indeed quite helpful thank you very much.