Go Back   Home > Forums > Loudspeakers > Multi-Way

Multi-Way Conventional loudspeakers with crossovers

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 9th August 2010, 08:48 PM   #7071
ScottG is offline ScottG  United States
diyAudio Member
 
Join Date: Feb 2003
Location: US
Quote:
Originally Posted by g3dahl View Post
Picture time!


Click the image to open in full size.

Very nice!
__________________
perspective is everything
  Reply With Quote
Old 9th August 2010, 09:46 PM   #7072
g3dahl is offline g3dahl  United States
diyAudio Member
 
g3dahl's Avatar
 
Join Date: Jun 2004
Location: Silverdale, WA
Quote:
Originally Posted by Robh3606 View Post
Do you know where you are going to place the ribbon tweeter or are you going to try it first without one??
I'll go without tweeters at first, and work on getting the crossover dialed in. I expect to keep the TD15M and Azurahorn as close together as possible, which means the tweeter will have to stay out of the way. A few possibilities:

1. Above the Azurahorns, time-aligned of course. Hard to imagine an attractive way of doing this.

2. Next to the Azurahorns, atop the TD15M cabinets.

3. Facing rearward, as Hiraga has done with his highly-regarded JH-MS15 Reference speakers. (Link: 6moons industry features: Jean Hiraga)

I would like to experiment with these placements before designing permanent mounting provisions.

Gary Dahl
__________________
"So many tubes, so little time..."
  Reply With Quote
Old 10th August 2010, 12:20 AM   #7073
diyAudio Member
 
dantheman's Avatar
 
Join Date: May 2008
Location: Mountain View, CA
Nice Gary! Any which way you mount them they'll look sweet to me. Can't wait to see the graphs of the horn.

Dan
__________________
My Blog
My Music Recordings
  Reply With Quote
Old 10th August 2010, 12:54 AM   #7074
diyAudio Member
 
Join Date: May 2009
Location: Wellington
Quote:
Originally Posted by Few View Post
Bottom line: Variation of amplitude over time does not imply variation in the frequency response over time.
If the amplitude of the signal at a given frequency differs with time, the "frequency response" differs identically. I measure the frequency response of a system and it is flat. Then I measure it again some time later and there is now a sharp resonant peak. It just means the resonance took time to build up (implying high Q).

I wonder if you might have intended to say,
"Variation of amplitude over time does not imply variation in the resonant frequency over time."

The variation in amplitude is a signal. It appears in the form of new frequencies. For example, a 1 KHz sine wave varying sinusoidally in amplitude at 100 Hz. When analysed in the frequency domain, it appears as a constant amplitude 1 Khz signal and two new signals, 0.9 KHz and 1.1 KHz.

Likewise, analysing the spectrum of a resonance building up due to excitation will show "sidebands" at frequencies related to the time constants (Q) of the resonance.
  Reply With Quote
Old 10th August 2010, 02:32 AM   #7075
Few is offline Few  United States
diyAudio Member
 
Join Date: Apr 2004
Location: Maine, USA
Hi Don,
Quote:
I wonder if you might have intended to say,
"Variation of amplitude over time does not imply variation in the resonant frequency over time."
No, I actually meant what I posted. The system (speaker, for example) has a frequency response. It doesn't have different frequency responses depending on which signal it's fed or on when you happen to measure them.

The frequency response can be expressed in the frequency domain as a magnitude and phase response as a function of frequency or as real and imaginary parts as a function of frequency. The system's response can also be characterized in the time domain by an impulse response (amplitude as a function of time). Any of these depictions are different ways of expressing the same information. If you have one, you have everything you need to produce the others.

Perhaps the confusion lies here: "Frequency response" and "function describing how much each frequency contributes to a given output signal within a particular time window" are not the same thing. To the extent that it's linear and time invariant, a system is uniquely characterized by its frequency response or, equivalently, by its impulse response (ignoring directivity issues). Given one of those, you have everything you need to compute what the system's response will be at any time to any input signal.

This is not the same as saying the system produces the same amount of sound at each frequency within each time interval no matter what signal it's fed. It's pretty clear all components in an audio system can produce different amounts of each frequency in different time intervals, otherwise they couldn't reproduce music.

Quote:
If the amplitude of the signal at a given frequency differs with time, the "frequency response" differs identically.
I think this may reflect a confusion between a frequency response and a snapshot of the spectrum of a system's response to some signal over some particular time interval. The former does not vary with time, and the latter definitely does. The display of a real time spectrum analyzer changes over time because it's showing how much each frequency contributes to the signal over some time period. It's varying primarily because the input signal is varying, and also because the response of a system with a given frequency response takes time to approach a steady state when it's driven by a signal. It's not varying because the system's frequency response has any time dependence.

I hope I'm not just muddying the waters with this post. I really think it comes down to the accepted definition of the frequency response of a system. This site is hardly authoritative, but perhaps it'll explain things more clearly than I have.

Few
  Reply With Quote
Old 10th August 2010, 02:37 AM   #7076
JLH is offline JLH  United States
diyAudio Member
 
JLH's Avatar
 
Join Date: Apr 2005
Location: Indianapolis, IN
Quote:
Originally Posted by g3dahl View Post
Picture time!

Click the image to open in full size.

Click the image to open in full size.

Click the image to open in full size.

Click the image to open in full size.
Okay, what exactly are we looking at here? What drivers were settled on and what are the loading configurations and the crossover points? I'd like to see a full system design description. I'm asking because I really haven't been following this thread much since it went over 3000 posts. I'm glad to see something actually got built.
__________________
Ah, how beautifully the orchestra sounds before a rain! In a dry sunny day there is no way for the instruments to sound this way!
  Reply With Quote
Old 10th August 2010, 03:17 AM   #7077
diyAudio Member
 
Join Date: May 2009
Location: Wellington
Quote:
Originally Posted by Few View Post
Hi Don,

No, I actually meant what I posted. The system (speaker, for example) has a frequency response. It doesn't have different frequency responses depending on which signal it's fed or on when you happen to measure them.

(...)

... It's varying primarily because the input signal is varying, and also because the response of a system with a given frequency response takes time to approach a steady state when it's driven by a signal. It's not varying because the system's frequency response has any time dependence.
... but if "the response of a system with a given frequency response takes time to approach a steady state when it's driven by a signal", then its frequency response has time dependence. But I see what you mean, and I think we're in agreement:

Quote:
Originally Posted by Few View Post
... I really think it comes down to the accepted definition of the frequency response of a system. ...
I originally wrote the same phrase in my post but edited it out before posting.

Thanks for the book pointer.
  Reply With Quote
Old 10th August 2010, 04:09 AM   #7078
g3dahl is offline g3dahl  United States
diyAudio Member
 
g3dahl's Avatar
 
Join Date: Jun 2004
Location: Silverdale, WA
Quote:
Originally Posted by JLH View Post
Okay, what exactly are we looking at here? What drivers were settled on and what are the loading configurations and the crossover points? I'd like to see a full system design description. I'm asking because I really haven't been following this thread much since it went over 3000 posts. I'm glad to see something actually got built.
Description was in post 6595:

Quote:
When the speakers are complete, each channel will consist of the following components:

Aurum Cantus G-3 ribbon (above ~8 kHz).
Azurahorn AH-425 with Radian 745PB driver (900 Hz to ~10 kHz).
AE TD15M in 3 cu ft sealed enclosure (75 Hz to 900 Hz). LF rolls off naturally.
AE TD15H and two PR15-700 passive radiators in 5 cu ft enclosure, actively powered. This will be allowed to overlap with the sealed cabinet in order to fill in below the baffle step, about 200 Hz.

The enclosures for both of the AE drivers are about 20" tall and 25" wide. The front vertical edges have a 4" radius. The lower (TD15H/PR15) cabinet depth is about 22", while the sealed (TD15M) cabinet is about 14" deep. The two cabinets will be stacked for a total height of about 40", and the Azurahorn and ribbon will be mounted on top. Definitely not a minimonitor!

At this point, all of the drivers are on hand, and all of the baltic birch and MDF parts have been cut. Today the first of the TD15M cabinets went through its first stage of glue-up. Assembly of remaining cabinets will continue through the coming week.

I chose the dual-woofer configuration because I wanted to have the ability to dial in the bass response shape. The TD15M is known for its excellent midrange, and I really liked the way the Unibox sims look for the the TD15H in the dual PR cabinet. Separate active power for the TD15H will allow me to adjust the lows for best in-room performance.
Here's a Unibox simulation of the TD15H-4 in the passive radiator enclosure:

Click the image to open in full size.

Here's a simulation of the TD15M-8A from the MH-Audio sealed-box calculator:

Click the image to open in full size.

Here's the measured response of the Radian 745 in the Azurahorn:

Click the image to open in full size.

Crossover development will commence this weekend, so these speakers haven't yet sung their first song.

Gary Dahl
__________________
"So many tubes, so little time..."
  Reply With Quote
Old 10th August 2010, 04:28 AM   #7079
mige0 is offline mige0  Austria
diyAudio Member
 
mige0's Avatar
 
Join Date: Jan 2005
Location: Austria, at a beautiful place right in the heart of the Alps.
Quote:
Originally Posted by Few View Post

I hope I'm not just muddying the waters with this post. I really think it comes down to the accepted definition of the frequency response of a system.

Few
Definitely you are not mudding the waters – quite in contrary – I'm actually happy you pointed out LTI such clearly, as I possibly wasn't so clear some times myself.

The question here is *if* we should look at CMP systems as to be LTI or not.

Seen as a whole – meaning from time zero to eternity, a CMP system definitely *is* both : time invariant and linear (and hence *can* be accessed for 100% correction in theory) - though I guess, the term „time invariant“ could also be interpretated differnently here. For sure a CMP system is „time invariant“ for each of the three time spans :
zero to delay time
delay time to end of input
end of input to end of input plus delay time


But however the math is defined – the net result with CMP is that FR varies with time we look at.
I could also state it different :

Looking only at a time window of „zero to delay time“ we *never* would get a clue that notches and bumps are happening in FR (which is not the case for „true“ LTI systems when we shorten the time window).

Also – conventionally EQ'ing out a „peak“ or „notch“ in FR of a CMP system will actually make performance in time domain worse (which also is not the case for „true“ LTI systems).

So we just as well could state that the concept of FR - meaning any FR plot actually - is quite useless (within the limits outlined) when it comes to CMP behaviour – but as a compromise its maybe more intuitively to speak of „FR changing in time“.


Michael

Last edited by mige0; 10th August 2010 at 04:41 AM.
  Reply With Quote
Old 10th August 2010, 05:03 AM   #7080
mige0 is offline mige0  Austria
diyAudio Member
 
mige0's Avatar
 
Join Date: Jan 2005
Location: Austria, at a beautiful place right in the heart of the Alps.
Quote:
Originally Posted by mige0 View Post
Seen as a whole – meaning from time zero to eternity, a CMP system definitely *is* both : time invariant and linear (and hence *can* be accessed for 100% correction in theory) - though I guess, the term „time invariant“ could also be interpretated differnently here. For sure a CMP system is „time invariant“ for each of the three time spans :
zero to delay time
delay time to end of input
end of input to end of input plus delay time
I should have added thats what appears to me as the "words meaning" of "LTI" and "time invariant" in the sense of "determined" (its definitely not me whose the math magician here).


Michael
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



New To Site? Need Help?

All times are GMT. The time now is 01:58 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2