Beyond the Ariel - Page 361 - diyAudio
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 17th May 2008, 07:53 AM   #3601
diyAudio Member
 
Graham Maynard's Avatar
 
Join Date: Dec 2003
Location: UK
Hi Michael,

You wrote >>
here an annoying resonance follows a DIP in the frequency response <<

But is that rear 0ms trace the same frequency response we are so used to observing when a driver is energised via a steady sine ?

Where there is a resonant characteristic then there is a loss of energy during a first half cycle as a portion of drive becomes stored, then integrated within cone motion, and then released when the drive is removed.

In Posts#3546>3549 above, Lynn made significant comments relating to effects happening in 'time'.

Delayed effects arising within a waveform time period can modify dynamic waveform reproduction responses and propagation patterns in time, but cannot be electronically compensated for before they arise, as with this peak/dip.
A reproduction peak, or room positioning LS peak, might be mitigated by reducing drive at the problem frequency, but that can lead to further degradation of the first cycle of the dynamic responses, and to waveform induced positional shifting of generated imagery in time.
Hence poor driver/room characteristics cannot be EQed out (only optimised) and it is driver choice/mounting/positioning which counts.

We can have a good 'sine' response without having a good 'first cycle' (dynamic articulation) response, whereas if we take trouble to get the first cycle response correct, then the measured sine response is likely to be reasonable anyway !

Cheers ........... Graham.
  Reply With Quote
Old 17th May 2008, 10:45 AM   #3602
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.
Good point, Graham.

But frankly - I havent explored that any deeper –maybe someone else can comment on that more profund.


Taking some old datasheets I would expect that the "first half cycle behaviour" shouldn't be considered as a constant. My half educated guess would be that it's related to the Qts of the speaker. Qts past 0.7 should result in overshots if I am right.

The envelope of that first oscillations is kind of same thing like seen at (op)amps where it is usually called setting time and related to the phase margin.

To give a idea of what we are talking here, the comparison of two speakers may be helpful
D52 with a Qts of 0.53
30W54 with a Qts of 0.36


Click the image to open in full size.


Click the image to open in full size.


The comments on that data sheets about compression and the 120dB plots should fit perfect to that thread.

Greetings
Michael
  Reply With Quote
Old 17th May 2008, 01:31 PM   #3603
KSTR is offline KSTR  Germany
diyAudio Member
 
KSTR's Avatar
 
Join Date: Jul 2007
Location: Central Berlin, Germany
Quote:
Originally posted by mige0
But frankly - I havent explored that any deeper –maybe someone else can comment on that more profund.

Taking some old datasheets I would expect that the "first half cycle behaviour" shouldn't be considered as a constant. My half educated guess would be that it's related to the Qts of the speaker. Qts past 0.7 should result in overshots if I am right.
Yes, this "wrong" lead-in part of the waveform is a consequence of the highpass function (and, with a real driver, also of it's lowpass behaviour) applied to a non-sinusodial input (the sudden start of sine is not sine anymore, spectrally). The higher the Q of the 2nd order highpass, the longer it takes for the wave to settle. Unfortunately any highpass shows a step response overshoot, the magic Q value only applies for the lowpass function and there it isn't 0.71 (this only yields maximum flat amplitude frequency response, it's 0.50 (aperiodic response). Therefore, at the sudden start of sine running through a highpass we have two events, an intitial ringing/overshoot and a more or less slow build-up of amplitude. The overshoot/ringing only is visible when we are below the corner frequency, otherwise is masked by the amplitude. To test effects that are really a function of driver nonlinearities etc, one might better use sine bursts with raised cosine envelope which have a known and benign spectrum.

Linkwitz has some graphs showing these relationships:
http://linkwitzlab.com/filters.htm#9
And textbook highpass filter simulations confirm this perfectly.

- Klaus
  Reply With Quote
Old 17th May 2008, 05:34 PM   #3604
diyAudio Member
 
Graham Maynard's Avatar
 
Join Date: Dec 2003
Location: UK
Hi KSTR,

Thanks for your Linkwitzlab link.

The last set of drawings in section 9 show EQ controlling Fp=55Hz with Qp=1.2, however the first cycle dynamics are cannot be improved due to the higher driver Qp figure.

If a LF driver of lower Qes, lower Fs and low moving mass is used, especially on an open baffle, then the first cycle is much less degraded, any subsequent resonance less troublesome, and the dynamic response more accurate.

Cheers ...... Graham.
  Reply With Quote
Old 17th May 2008, 05:57 PM   #3605
soongsc is offline soongsc  Taiwan
diyAudio Member
 
soongsc's Avatar
 
Join Date: Mar 2005
Location: Taiwan
There are a few things to take note of when looking at CSDs:

1. When two drivers have pretty much the same effeciency and frequency response, the driver that decays slower will sound louder, the driver that decays faster will sound cleaner.

2. Since not all frequencies will have the same rate of decay, for frequency ranges that decay slower, you will notice non-percussion instruments in these ranges will sound more pronounced. In some cases this may create the illusion of having more detail which may not be correct representation of the recording but may be preferred by some listeners.
__________________
Hear the real thing!
  Reply With Quote
Old 17th May 2008, 06:40 PM   #3606
KSTR is offline KSTR  Germany
diyAudio Member
 
KSTR's Avatar
 
Join Date: Jul 2007
Location: Central Berlin, Germany
Quote:
Originally posted by Graham Maynard
Thanks for your Linkwitzlab link.

The last set of drawings in section 9 show EQ controlling Fp=55Hz with Qp=1.2, however the first cycle dynamics are cannot be improved due to the higher driver Qp figure.
Maybe I misread you, but do you say the composite response, driver+correction, doesn't follow the product of the individual transfer functions? Assuming linear time-invariant systems (which is to be taken with a grain of salt wrt drivers) this initial "error" is only a matter of the combined transfer function.

I've attached a 28Hz Critical/Aperiodic (Q=0.50) 2nd order highpass time response, with 100Hz test frequency, which gives the same frequency ratio as in the Linkwitz example. All in all a 1:1 copy of the Linkwitz graph, isn't it? The perfect theoretical response for HP2 with Q=0.5 and a 3.5:1 freq ratio. If you have that conditions with a real system, this will be the response, regardless of system type (OB/CB/BR/...), provided that the system is well executed.

- Klaus
Attached Images
File Type: gif linkwitz.gif (17.6 KB, 626 views)
  Reply With Quote
Old 17th May 2008, 06:54 PM   #3607
KSTR is offline KSTR  Germany
diyAudio Member
 
KSTR's Avatar
 
Join Date: Jul 2007
Location: Central Berlin, Germany
Addendum, even one decade away from the filter frequency the ideal HP response still isn't a "copy" of the input waveform. We have to say goodbye to that, unless we go way up in frequency. This "error" is, BTW, a common source of confusion in distortion measurements, be in in reality or under simulation, you have to wait until things have settled.

- Klaus
Attached Images
File Type: gif linkwitz2.gif (17.0 KB, 618 views)
  Reply With Quote
Old 17th May 2008, 07:49 PM   #3608
diyAudio Member
 
Robh3606's Avatar
 
Join Date: Nov 2002
Location: Destiny
"1. When two drivers have pretty much the same effeciency and frequency response, the driver that decays slower will sound louder, the driver that decays faster will sound cleaner."

How much of a difference in the decay rate is audible as a change in loudness?? 2 ms? 10ms? 30ms??

Rob
  Reply With Quote
Old 17th May 2008, 07:50 PM   #3609
diyAudio Member
 
Join Date: Aug 2006
Quote:
Originally posted by soongsc
2. Since not all frequencies will have the same rate of decay, for frequency ranges that decay slower, you will notice non-percussion instruments in these ranges will sound more pronounced. In some cases this may create the illusion of having more detail which may not be correct representation of the recording but may be preferred by some listeners.

Maybe I'm just thick, but isn't exactly this what is referred to as "coloration" ?

For example smth like this ?
  Reply With Quote
Old 17th May 2008, 11:41 PM   #3610
soongsc is offline soongsc  Taiwan
diyAudio Member
 
soongsc's Avatar
 
Join Date: Mar 2005
Location: Taiwan
Quote:
Originally posted by Robh3606
"1. When two drivers have pretty much the same effeciency and frequency response, the driver that decays slower will sound louder, the driver that decays faster will sound cleaner."

How much of a difference in the decay rate is audible as a change in loudness?? 2 ms? 10ms? 30ms??

Rob
I cannot quantify exactly how much difference is audible because not enough listener sampling is done, the if you look at the CSDs below, the one on the bottom will sound the loudest.
Click the image to open in full size.
Note that normally I compare with different drivers in each of thw two channels.
__________________
Hear the real thing!
  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 11:05 AM.


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