Dual Sealed SI HS24 Build - Page 5 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 11Likes
Reply
 
Thread Tools
post #121 of 233 Old 01-17-2014, 07:14 PM
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 15,491
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 5777
Quote:
Originally Posted by WereWolf84 View Post

Bill, do you mean 20 cubic feet is the ideal box volume for this driver?
33 cu ft, actually.
Quote:
I guess you mean stuffing until the box is full & tight, right? anyways to get the Qtc down to 0.7?
I don't think you can get it to 0.7, maybe 0.8, which isn't bad.
Quote:
I prefer a sealed cab for the extreme single digit low freq.
That's OK, but to tell the truth this probably wasn't the best driver choice to do it with, unless it was going into an IB, where the 'enclosure' size wouldn't be an issue.
Bill Fitzmaurice is offline  
Sponsored Links
Advertisement
 
post #122 of 233 Old 01-17-2014, 07:20 PM - Thread Starter
AVS Forum Special Member
 
WereWolf84's Avatar
 
Join Date: May 2011
Posts: 1,150
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 416 Post(s)
Liked: 175
Quote:
Originally Posted by Bill Fitzmaurice View Post

33 cu ft, actually.
I don't think you can get it to 0.7, maybe 0.8, which isn't bad.
That's OK, but to tell the truth this probably wasn't the best driver choice to do it with, unless it was going into an IB, where the 'enclosure' size wouldn't be an issue.
33 cf eek.gif I'm going to stuff with pillows until the box is full tongue.gif I might going to re-build & put the HS24 drivers in IB application in future since I have about 2 feet space behind my AT screen biggrin.gif but right now as my first ever DIY, I'm going to try the simplest sealed enclosure build first smile.gif meanwhile studying & learning how to build an much more complex IB properly with the HS24
WereWolf84 is online now  
post #123 of 233 Old 01-17-2014, 08:41 PM
Advanced Member
 
Electrodynamic's Avatar
 
Join Date: Oct 2009
Posts: 900
Mentioned: 115 Post(s)
Tagged: 0 Thread(s)
Quoted: 536 Post(s)
Liked: 806
Quote:
Originally Posted by Bill Fitzmaurice View Post

I don't have to hear a sub to pass judgement. That's one of the perks of being an acoustical engineer. wink.gif
I wouldn't call these drivers a huge value, but they're not bad. In 10 cu ft sealed they have a maximum SPL of 124dB at 50Hz, 112dB at 20Hz. By comparison the Dayton Ultimax 15 in 5 cu ft comes in about 8dB lower, so one of the SI is equal to almost three of the Daytons. For equal SPL I'd go with the option of more smaller cabs spread around the room, but that's a different question entirely. A more appropriate question relates to room size. If it's more than 20 feet in any dimension it's going to take a lot of eq, and therefore power, to get these flat to 20Hz.
BTW, OP, in 10 cu ft your Qtc is .95, and that's not so good. Even heavily stuffed it will only go down to perhaps .8, which is better, but with the high f3 of these .7 would be better.
IMO the manufacturer recommendation of 10cu ft isn't a good one, that driver really wants twice that. The 10 cu ft recommendation is probably to keep potential buyers from being scared away.

You are correct, it takes over three Dayton's to equal the displacement of one HS 24. How it performs when not measured in an anechoic chamber is another story...yet that seems to be ignored here. What kind of acoustical engineer are you? Who have you worked for and what systems have you designed? I'm not being negative but rather genuinely curious because I haven't seen any engineers use, and quote, WinISD plots so biblically before. If you look through Josh Ricci's test of the SI HT and the Dayton HO you'll see that the SI HT was over 2 dB louder across the board...but WinISD won't tell you that. So in reality it will take five Dayton HO's to equal the output of one HS 24.
Quote:
Originally Posted by kramskoi View Post

And this...this, is the rather confounding thing about this driver. Even if I wanted to afford it and build a new cabinet, the qts is in the stratosphere. I suppose that you could beat on the suspension in free air for a duration sufficient enough to mechanically lower qts, but that is hit or miss. The suspension may give a lot or it may give little under such an exercise, but the results would not be dramatic enough to significantly lower qts. in this case.

Stratosphere? I disagree. Qts is 0.55 on small signal analysis, not 1.3. I consider a "stratospheric Qts" to be above a 1.0. If you over-motor a driver and drive the Qts down too far you diminish low frequency extension in sealed alignments. With treatments within the motor to linearize and lower inductance a moderate Qts woofer will yield more output over the listening bandwidth than that of a low Qts woofer. Output across a specified bandwidth is what you are after [and is also what dictates transient response as described in the Fourier Transform]. Do we recommend a 0.707 alignment? No we do not. We recommend an alignment that provides a lot of output from 80 Hz down. So far the reviews of our alignment have been stellar even when compared to lower "Q" alignment systems. We are not going to change our recommendation.
Electrodynamic is offline  
Sponsored Links
Advertisement
 
post #124 of 233 Old 01-18-2014, 04:49 AM
Senior Member
 
WVSyd's Avatar
 
Join Date: Feb 2013
Posts: 333
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 47
Quote:
Originally Posted by Electrodynamic View Post

.... What kind of acoustical engineer are you? Who have you worked for and what systems have you designed? ....

F.Y.I Speaker Builder magazine chronicles his builds and Biography

"Beware of Salesmen: They are the modern Svengali, immune to Science and Reality"
WVSyd is offline  
post #125 of 233 Old 01-18-2014, 04:49 AM
Advanced Member
 
kramskoi's Avatar
 
Join Date: Dec 2005
Location: south-central Louisiana
Posts: 502
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 28
Quote:
Originally Posted by Electrodynamic View Post

You are correct, it takes over three Dayton's to equal the displacement of one HS 24. How it performs when not measured in an anechoic chamber is another story...yet that seems to be ignored here. What kind of acoustical engineer are you? Who have you worked for and what systems have you designed? I'm not being negative but rather genuinely curious because I haven't seen any engineers use, and quote, WinISD plots so biblically before. If you look through Josh Ricci's test of the SI HT and the Dayton HO you'll see that the SI HT was over 2 dB louder across the board...but WinISD won't tell you that. So in reality it will take five Dayton HO's to equal the output of one HS 24.
Stratosphere? I disagree. Qts is 0.55 on small signal analysis, not 1.3. I consider a "stratospheric Qts" to be above a 1.0. If you over-motor a driver and drive the Qts down too far you diminish low frequency extension in sealed alignments. With treatments within the motor to linearize and lower inductance a moderate Qts woofer will yield more output over the listening bandwidth than that of a low Qts woofer. Output across a specified bandwidth is what you are after [and is also what dictates transient response as described in the Fourier Transform]. Do we recommend a 0.707 alignment? No we do not. We recommend an alignment that provides a lot of output from 80 Hz down. So far the reviews of our alignment have been stellar even when compared to lower "Q" alignment systems. We are not going to change our recommendation.
Low frequency extension and output are relative. The qts is in the stratosphere when I personally enjoy a qts around .2 or even less. But that is just what I prefer for MANY reasons. Most here would regard using a low qts/high BL driver with its consequent high Fb/F3 to be lunacy and a waste of the drivers available bandwidth. However, it actually depends on the driver in question. How much of the bandwidth is actually used with a 60 Hz crossover? The thing that matters most is Sd*Xmax below about 40 - 50Hz. As such, I believe two HS-24 drivers, with the same Sd and Xmax, will deliver similar displacement-limited output whether the BL is high or low. The "native" small-signal/enclosure bandwidth is simply shifted between the two but the mechanical potential still exists equally for both. Equalization and power will "mostly" solve that issue. How many manufacturers or DIY'ers actually leave it to the native FR of a sealed enclosure to determine its final output characteristics in this age of almost infinitesimally adjustable digital signal processors?

Yes, you do indeed limit "native" low frequency extension/output with low qts. But once again, that is relative because Sd*Xmax still remains for equalization. You can get infrasound from a low qts. driver, even a pro-sound one with enough Sd*Xmax. Will it be the most efficient use of the driver without equalization? Probably not, but there is more to be examined in personal choice than just the native bandwidth.

That said, I really like this driver. I just find its qts much too high on a personal level, and it perturbs me every time a driver like this hits the market with parameters that just don't work for me. That is not to impugn the design, the sound and/or build quality. It looks top-notch. Big time output in a single driver, even if you have to shape up the FR to get it all, but that is just the nature of the business when you go sealed.

"Frequency response is NOT efficiency response."

 

Klipsch RB-35's main

Klipsch RB-35's surround

Acoustech HT-65 center

Cerwin-Vega CV-2800

 

Subwoofer

B&C 21SW152-4  [21" critical-Q]

 

kramskoi is offline  
post #126 of 233 Old 01-18-2014, 05:15 AM
AVS Forum Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 21,216
Mentioned: 862 Post(s)
Tagged: 1 Thread(s)
Quoted: 2955 Post(s)
Liked: 3685
"I'm not being negative but rather genuinely curious because I haven't seen any engineers use, and quote, WinISD plots so biblically before."

must be completely out of the loop then. the underlying thiele small equations are amazingly accurate and have been at the core of speaker enclosure design since they were published 40 years ago.

Thiele, A.N., "Loudspeakers in Vented Boxes, Parts I and II", J. Audio Eng. Soc., vol. 19, pp. 382–392 (May 1971); pp. 471–483 (June 1971).
Small, R.H., "Direct-Radiator Loudspeaker System Analysis", J. Audio Eng. Soc., vol. 20, pp. 383–395 (June 1972).
Small, R.H., "Closed-Box Loudspeaker Systems", J. Audio Eng. Soc., vol. 20, pp. 798–808 (Dec. 1972); vol. 21, pp. 11–18 (Jan./Feb. 1973).
Small, R.H., "Vented-Box Loudspeaker Systems", J. Audio Eng. Soc., vol. 21, pp. 363–372 (June 1973); pp. 438–444 (July/Aug. 1973); pp. 549–554 (Sept. 1973); pp. 635–639 (Oct. 1973).
WVSyd likes this.

Listen. It's All Good.
LTD02 is offline  
post #127 of 233 Old 01-18-2014, 05:20 AM
AVS Forum Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 21,216
Mentioned: 862 Post(s)
Tagged: 1 Thread(s)
Quoted: 2955 Post(s)
Liked: 3685
"the ultimate party speaker! frat party all day... everyday."

not sure that would be the best use.

would make a good ht subwoofer. that's for sure.

Listen. It's All Good.
LTD02 is offline  
post #128 of 233 Old 01-18-2014, 05:36 AM
Senior Member
 
WVSyd's Avatar
 
Join Date: Feb 2013
Posts: 333
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 47
Quote:
Originally Posted by LTD02 View Post

... the underlying thiele small equations are amazingly accurate and have been at the core of speaker enclosure design since they were published 40 years ago.

Thiele, A.N., "Loudspeakers in Vented Boxes, Parts I and II", J. Audio Eng. Soc., vol. 19, pp. 382–392 (May 1971); pp. 471–483 (June 1971).
Small, R.H., "Direct-Radiator Loudspeaker System Analysis", J. Audio Eng. Soc., vol. 20, pp. 383–395 (June 1972).
Small, R.H., "Closed-Box Loudspeaker Systems", J. Audio Eng. Soc., vol. 20, pp. 798–808 (Dec. 1972); vol. 21, pp. 11–18 (Jan./Feb. 1973).
Small, R.H., "Vented-Box Loudspeaker Systems", J. Audio Eng. Soc., vol. 21, pp. 363–372 (June 1973); pp. 438–444 (July/Aug. 1973); pp. 549–554 (Sept. 1973); pp. 635–639 (Oct. 1973).

Indeed.
Even before the work of T/S: Basic physical and electrical characteristics of loudspeakers ( circa 1927 ) are documented in the work of individuals such as Leo Beranek, Harry Olson, W.B. Snow and others.

"Beware of Salesmen: They are the modern Svengali, immune to Science and Reality"
WVSyd is offline  
post #129 of 233 Old 01-18-2014, 06:43 AM
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 15,491
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 5777
Quote:
Originally Posted by Electrodynamic View Post

What kind of acoustical engineer are you?
Very successful. rolleyes.gif
Quote:
If you look through Josh Ricci's test of the SI HT and the Dayton HO you'll see that the SI HT was over 2 dB louder across the board...
I don't see any testing of the Ultimax 15 there. What is the 'HO'?
Quote:
must be completely out of the loop then. the underlying thiele small equations are amazingly accurate and have been at the core of speaker enclosure design since they were published 40 years ago.
+1. If the data and measured results don't agree then somebody messed up with the data. For actual design work I'll go through the time and effort involved using AkAbak, but otherwise WinISD and HornResp give sufficiently accurate results.
Bill Fitzmaurice is offline  
post #130 of 233 Old 01-18-2014, 12:01 PM
Senior Member
 
WVSyd's Avatar
 
Join Date: Feb 2013
Posts: 333
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 47
I Think, therefore I am ( René Descartes )
I SAY, therefore I am ( Yoda )

"Beware of Salesmen: They are the modern Svengali, immune to Science and Reality"
WVSyd is offline  
post #131 of 233 Old 01-18-2014, 07:44 PM
Advanced Member
 
Electrodynamic's Avatar
 
Join Date: Oct 2009
Posts: 900
Mentioned: 115 Post(s)
Tagged: 0 Thread(s)
Quoted: 536 Post(s)
Liked: 806
Quote:
Originally Posted by Bill Fitzmaurice View Post

Very successful. rolleyes.gif

Besides the blanket response of "Very successful" what have you positively given the loudspeaker industry aside of DIY projects? Can you elaborate for me?

Quote:
Originally Posted by LTD02 View Post

"I'm not being negative but rather genuinely curious because I haven't seen any engineers use, and quote, WinISD plots so biblically before."

must be completely out of the loop then. the underlying thiele small equations are amazingly accurate and have been at the core of speaker enclosure design since they were published 40 years ago.

This is not 40 years ago. Back then low frequency transducers were doing good to have 6mm of linear travel. Things have changed since then. People can focus on a specific T/S measurement and base their projections on one parameter all they want. The reality of the systems interaction with the environment in which it is placed in has a huge factor. Modeling only takes you so far. And even then those models are limited. Freeware does does not project unlimited displacement/SPL figures accurately. We did a measurement set a few years back where we took a model-perfect woofer against a woofer we already had in production. The "model-perfect" woofer was louder [on paper] than the woofer we had out on the market. WinISD, T/S measurements, etc, were all leaning towards the "model-perfect" woofer. When we applied power [clamped/measured] to the two systems the woofer we had in production was over 3 dB louder across the board compared the WinISD perfect woofer. Was the WinISD perfect woofer louder at all? Yes, it was louder at 70 Hz. The production woofer [not perfect in WinISD] was louder from 60 Hz down to 20 Hz.
Electrodynamic is offline  
post #132 of 233 Old 01-18-2014, 11:05 PM
AVS Forum Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 21,216
Mentioned: 862 Post(s)
Tagged: 1 Thread(s)
Quoted: 2955 Post(s)
Liked: 3685
"When we applied power [clamped/measured] to the two systems the woofer we had in production was over 3 dB louder across the board compared the WinISD perfect woofer."

what is a winisd perfect woofer?

everybody who knows knows that t/s are small signal parameters and that they change with use. not at all meant for the purpose that you describe.

plop your big sub in a 10 cubic foot box, play it for an hour at its rated power, measure the qtc and get back to me. if qtc doesn't matter, then all drivers would have half the motor that they do, would cost half as much as they do, would be plopped in boxes half as large as they are...i.e., everybody would be selling drivers with infinite baffle t/s parameters and running them in bose boxes.

i'm not saying that your driver in anything less than some given q alignment would be a fail, just that it would be pretty far out from what is typically considered "within reason". 10 cubic footer gives a qtc of 0.94 and 8 cubic feet gives a qtc of 1. high q alignments resonate so give higher spl btw.

Listen. It's All Good.
LTD02 is offline  
post #133 of 233 Old 01-19-2014, 06:43 AM
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 15,491
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 5777
Quote:
Originally Posted by Electrodynamic View Post

Besides the blanket response of "Very successful" what have you positively given the loudspeaker industry aside of DIY projects? Can you elaborate for me?
Non-disclosure agreements prevent me from revealing the many designs I've done for commercial enterprises. Manufacturers do not want it known when their 'exclusive technologies' are created by independent designers like myself, just as many driver companies do not want it known that their 'exclusive technologies' actually came out of places like Guangdong and Jinan. rolleyes.gif
Not that I care, I gladly sign non-disclosure agreements, as that always leads to signing the back of a big fat check. biggrin.gif

Much interesting reading to be found here:
http://www.alibaba.com/showroom/speaker-parts.html
Bill Fitzmaurice is offline  
post #134 of 233 Old 01-19-2014, 08:25 AM
Senior Member
 
WVSyd's Avatar
 
Join Date: Feb 2013
Posts: 333
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 47
Quote:
M: Utter one more syllable and I'll have you killed 007

If secrets were revealed - he'd have to dispatch you with extreme prejudice...tongue.gif

Makes one appreciate the work of Engineers who share their papers.
Linkwitz and Don Keele are 2 prime examples

"Beware of Salesmen: They are the modern Svengali, immune to Science and Reality"
WVSyd is offline  
post #135 of 233 Old 01-19-2014, 04:15 PM
AVS Forum Special Member
 
Ricci's Avatar
 
Join Date: Mar 2007
Location: Louisville KY
Posts: 5,502
Mentioned: 167 Post(s)
Tagged: 0 Thread(s)
Quoted: 389 Post(s)
Liked: 910
WinISD and many other programs cannot accurately model many of the modern high inductance long stroke bass drivers. If interested go ahead and attempt to make a model match the actual measured response of the FI Q18 at Data-Bass. It requires complex inductance effects to be considered and very few programs do and then there is the problem that you are very unlikely to have this data to input in the first place unless you measure it yourself.

TSP's can be very accurate if the bass driver does not have large inductance related effects on the response and if only considering low strength signals where the driver is still linear. Anyone who understands Klippel testing or high power measurements has seen how perfectly driver behavior scales with more enthusiastic input levels. wink.gif Invariably when discussion revolves around bass systems and drivers the talk is always about theoretical output maximums, big amplification and big, heavy coiled, high inductance bass drivers with much emphasis on bench racing...None of which necessarily lead to accurate simplified models or correspond well with real world results when key parameters are missing from the simulation and the signals are anything but small.
Ricci is offline  
post #136 of 233 Old 01-19-2014, 05:10 PM
Advanced Member
 
kramskoi's Avatar
 
Join Date: Dec 2005
Location: south-central Louisiana
Posts: 502
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 28
Quote:
Originally Posted by Ricci View Post

WinISD and many other programs cannot accurately model many of the modern high inductance long stroke bass drivers. If interested go ahead and attempt to make a model match the actual measured response of the FI Q18 at Data-Bass. It requires complex inductance effects to be considered and very few programs do and then there is the problem that you are very unlikely to have this data to input in the first place unless you measure it yourself.

TSP's can be very accurate if the bass driver does not have large inductance related effects on the response and if only considering low strength signals where the driver is still linear. Anyone who understands Klippel testing or high power measurements has seen how perfectly driver behavior scales with more enthusiastic input levels. wink.gif Invariably when discussion revolves around bass systems and drivers the talk is always about theoretical output maximums, big amplification and big, heavy coiled, high inductance bass drivers with much emphasis on bench racing...None of which necessarily lead to accurate simplified models or correspond well with real world results when key parameters are missing from the simulation and the signals are anything but small.
Heat, and it all goes to hell, literally. It is amazing just how inefficient a driver actually is under power!wink.gif

"Frequency response is NOT efficiency response."

 

Klipsch RB-35's main

Klipsch RB-35's surround

Acoustech HT-65 center

Cerwin-Vega CV-2800

 

Subwoofer

B&C 21SW152-4  [21" critical-Q]

 

kramskoi is offline  
post #137 of 233 Old 01-20-2014, 06:33 AM
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 15,491
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 5777
Quote:
Originally Posted by Ricci View Post

WinISD...cannot accurately model many of the modern high inductance long stroke bass drivers.
You have to go into the Advanced tab and activate the 'Simulate voice coil inductance' option. Doing so in this case results in a 1.8dB hump centered at 55Hz, and a shift of the upper f3 to 145Hz.
The main shortcoming of basic programs lies in not being able to account for thermal and mechanical power compression to give an accurate Max SPL figure, but where comparing the relative capabilities of two systems is concerned the combination of Sensitivity, Pe and Vd give an accurate enough result. It won't be perfect in terms of absolute SPL, but the average user is less concerned about that than they are the simple question of which option will go louder.
Bill Fitzmaurice is offline  
post #138 of 233 Old 01-20-2014, 06:56 AM
AVS Forum Special Member
 
Ricci's Avatar
 
Join Date: Mar 2007
Location: Louisville KY
Posts: 5,502
Mentioned: 167 Post(s)
Tagged: 0 Thread(s)
Quoted: 389 Post(s)
Liked: 910
Sorry Bill but it doesn't work. The response still will be nothing like the real response of the FI driver or any number of others. It cannot do it. I have tried it. Hornresp while a great program also cannot accurately model that type of driver. The FR shape is way off as is the sensitivity and resultant impedance curve, applied current, excursion, etc. Manufacturer SD, power handling and Xmax specs are all over the map as well.
Ricci is offline  
post #139 of 233 Old 01-20-2014, 07:10 AM
AVS Forum Special Member
 
blah450's Avatar
 
Join Date: Jan 2010
Posts: 1,176
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 142 Post(s)
Liked: 173
Bill and Josh...I am learning a lot (and a lot still going over my head) reading your preceding posts. Very much appreciate it being posted.

I look forward to reading/seeing the conclusion of the debate above...seems as if something is not adding up here with one or the other.
blah450 is offline  
post #140 of 233 Old 01-20-2014, 07:18 AM
Advanced Member
 
kramskoi's Avatar
 
Join Date: Dec 2005
Location: south-central Louisiana
Posts: 502
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 28
Quote:
Originally Posted by blah450 View Post

Bill and Josh...I am learning a lot (and a lot still going over my head) reading your preceding posts. Very much appreciate it being posted.

I look forward to reading/seeing the conclusion of the debate above...seems as if something is not adding up here with one or the other.
The dual B&C 21SW152 build thread from 2010 is a real hoot as well!

"Frequency response is NOT efficiency response."

 

Klipsch RB-35's main

Klipsch RB-35's surround

Acoustech HT-65 center

Cerwin-Vega CV-2800

 

Subwoofer

B&C 21SW152-4  [21" critical-Q]

 

kramskoi is offline  
post #141 of 233 Old 01-20-2014, 07:30 AM
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 15,491
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 5777
Quote:
Originally Posted by Ricci View Post

Sorry Bill but it doesn't work. The response still will be nothing like the real response of the FI driver or any number of others. It cannot do it. I have tried it. Hornresp while a great program also cannot accurately model that type of driver. The FR shape is way off as is the sensitivity and resultant impedance curve, applied current, excursion, etc. Manufacturer SD, power handling and Xmax specs are all over the map as well.
That would imply you have a measured response chart for the SI 24, but I don't see it on your site.
Bill Fitzmaurice is offline  
post #142 of 233 Old 01-20-2014, 07:52 AM
AVS Forum Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 21,216
Mentioned: 862 Post(s)
Tagged: 1 Thread(s)
Quoted: 2955 Post(s)
Liked: 3685
"I look forward to reading/seeing the conclusion of the debate above...seems as if something is not adding up here with one or the other."

there really isn't anything to debate. winisd just doesn't model a driver such as the fi q inductance monster on the upper end of the frequency range with very much accuracy. most drivers don't have as bad of inductance though. that one is normalized le/re over 2, which is quite high.

Listen. It's All Good.
LTD02 is offline  
post #143 of 233 Old 01-20-2014, 07:54 AM
AVS Forum Special Member
 
Ricci's Avatar
 
Join Date: Mar 2007
Location: Louisville KY
Posts: 5,502
Mentioned: 167 Post(s)
Tagged: 0 Thread(s)
Quoted: 389 Post(s)
Liked: 910
Quote:
Originally Posted by Bill Fitzmaurice View Post

That would imply you have a measured response chart for the SI 24, but I don't see it on your site.

My comments have nothing to do with the SI24" driver specifically but instead focus on the bench racing that often goes on around here and the comments that WinISD and small signal TSP's are accurate for simulation of the very large signals and the types of drivers often used in these bench races which they are not.

Again...The measurements of the FI Q18d2 are at Data-Bass. I could not make WinISD simulate a curve that matches the real world behavior of this driver even using inductance and a fair amount of trickery with the available parameters. I would be very surprised if anyone else could either (Pepsi challenge anyone?). Also the Q driver I only picked as an example there are a number of other very high inductance drivers that I have tested and a multitude more I have not, that also cannot be accurately simulated in WinISD or similar programs. In the last copy of WinISD I messed around with the complex inductance components were not even functional, perhaps I missed the newer version where this functionality was operational? Even with Unibox and using 3 part inductance from measured parameters of the driver it is much closer but still not that great of a match.



John,
It isn't just about the top end it is the entire bandwidth of the device that is affected. The response hump isn't even in the right place. It's off by more than 1/3rd octave. You can force it to become closer by messing with a few parameters that aren't even under the driver tab but then the voltage sensitivity will be even further off.
Scott Simonian likes this.
Ricci is offline  
post #144 of 233 Old 01-20-2014, 08:26 AM
AVS Forum Special Member
 
blah450's Avatar
 
Join Date: Jan 2010
Posts: 1,176
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 142 Post(s)
Liked: 173
"'I look forward to reading/seeing the conclusion of the debate above...seems as if something is not adding up here with one or the other.'

there really isn't anything to debate. winisd just doesn't model a driver such as the fi q inductance monster on the upper end of the frequency range with very much accuracy. most drivers don't have as bad of inductance though. that one is normalized le/re over 2, which is quite high."


debate...discussion...dialogue...biggrin.gif...all good.

I have learned a lot from the discussions on the board (and you, LTD!).

To me, I was wondering if Bill's post #137 "relative capabilities of two systems" is valid given Josh's further dialogue following that post.

Of course, maybe I misunderstood something Bill posted...again, these fellas, are heads and shoulders (nay, many whole bodies) above me on this learning curve. I'm just trying to learn more.
blah450 is offline  
post #145 of 233 Old 01-20-2014, 08:39 AM
Senior Member
 
WVSyd's Avatar
 
Join Date: Feb 2013
Posts: 333
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 47
Quote:
Originally Posted by blah450 View Post

...
debate...discussion...dialogue...biggrin.gif...all good.....
Klippel and valid real world measurement - Even better
( The S stands for Science - rather than Speculation and Sales Pitch )

"Beware of Salesmen: They are the modern Svengali, immune to Science and Reality"
WVSyd is offline  
post #146 of 233 Old 01-20-2014, 08:45 AM
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 15,491
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 5777
Quote:
Originally Posted by Ricci View Post

The measurements of the FI Q18d2 are at Data-Bass. I could not make WinISD simulate a curve that matches the real world behavior of this driver even using inductance and a fair amount of trickery with the available parameters. I would be very surprised if anyone else could either (Pepsi challenge anyone?)
A challenge for someone, maybe Juha or Dave McBean, not for me. I don't write code, I just use it. As I already said if the measured results and data don't agree there's something wrong with the data. In this case what that is remains to be identified. It would make a nice project for someone who wants to make an AES presentation.

Edit: I went back and did what I always do when creating a design, which is to compare the predicted versus measured response and then see where the major differences lie. I always start with impedance, because it to a great extent reveals the reasons for FR. The difference between what Ricci measured and HornResp predicted was significant. When I changed a couple of box details to get the impedance to match up the FR did as well. The same happened using HornResp, though there the initial results were closer to begin with, as it allows for a lot more initial data entry than WinISD, especially with regards to specific box details. I didn't do a check with AkAbak, as that takes ten times as long, and I'm not getting paid to do this particular gig. wink.gif
This is SOP when creating a design, as you never can assume that a program is going to get it perfect, there being details that it can't consider, including but not limited to the shape of the box. That's why we still measure the results. cool.gif
Bill Fitzmaurice is offline  
post #147 of 233 Old 01-20-2014, 10:06 AM
AVS Forum Special Member
 
Ricci's Avatar
 
Join Date: Mar 2007
Location: Louisville KY
Posts: 5,502
Mentioned: 167 Post(s)
Tagged: 0 Thread(s)
Quoted: 389 Post(s)
Liked: 910
I already asked David about it and he said nay and it really matters little as no one publishes the data required anyway. Anyway my point is simply that with a lot of the high inductance bass drivers the real world resembles the simplified simulation very little. The worse the inductive performance of a bass driver the more important those parameters become in order to get a simulation that matches reality. The funny part is that those are exactly the drivers that often offer the least amount of information.

Anyway that's all I was trying to say. People act like WinISD is infallible when the truth is often times there are missing parameters needed to get a true result that are unknown and even the small signal simulation is way off so you can definitely forget about the high power stuff. People buy drivers based
on this stuff and which one appears more powerful.
Ricci is offline  
post #148 of 233 Old 01-20-2014, 10:10 AM
AVS Forum Special Member
 
Join Date: Feb 2013
Posts: 1,539
Mentioned: 35 Post(s)
Tagged: 0 Thread(s)
Quoted: 17 Post(s)
Liked: 690
Quote:
Originally Posted by Ricci View Post

Again...The measurements of the FI Q18d2 are at Data-Bass. I could not make WinISD simulate a curve that matches the real world behavior of this driver even using inductance and a fair amount of trickery with the available parameters. I would be very surprised if anyone else could either (Pepsi challenge anyone?).

Leach's paper on voice coil inductance solves this very problem, but I don't know of any loudspeaker simulation programs that use this model other than the Geddes Speak program. It can be implemented in SPICE though.
kramskoi likes this.
andyc56 is offline  
post #149 of 233 Old 01-20-2014, 01:26 PM
Advanced Member
 
kramskoi's Avatar
 
Join Date: Dec 2005
Location: south-central Louisiana
Posts: 502
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 28
Quote:
Originally Posted by Ricci View Post

I already asked David about it and he said nay and it really matters little as no one publishes the data required anyway. Anyway my point is simply that with a lot of the high inductance bass drivers the real world resembles the simplified simulation very little. The worse the inductive performance of a bass driver the more important those parameters become in order to get a simulation that matches reality. The funny part is that those are exactly the drivers that often offer the least amount of information.

Anyway that's all I was trying to say. People act like WinISD is infallible when the truth is often times there are missing parameters needed to get a true result that are unknown and even the small signal simulation is way off so you can definitely forget about the high power stuff. People buy drivers based
on this stuff and which one appears more powerful.
It is not infallible but it is not invaluable either. One can still measure the finished results. I prefer to design against as many negatives as possible in a sealed alignment. I remember the subject of inductance from when Dan Wiggins was championing the XBL^2 drivers like the Tumult. I didn't get it until I measured the thing and then it was clear to see why it sounded the way it sounded without equalization, and yes, the pristine little Win graph bared little resemblance to the measured nearfield response. Dan's woofer-speed paper extolled the virtues of reduced inductance as a facilitator of faster current changes through the coil and thus better transient behavior. However, his Tumult driver ends up with a pronounced hump in its FR signature? I mean the nearfield response had a mountain profile. So is high Le not worth the paper it is printed on?

The trappings of using larger and/or longer voice coils contribute to the Le problem in all but the most well-built, expensive drivers, and even those can miss the mark. I confess to knowing little as to where the hump will appear in the FR [Win won't do it either] and that is why high BL pro-sound drivers have become more and more attractive for me personally. Perhaps if you can get the potential Le hump out of the subwoofers operating range [below the crossover frequency] you can mitigate or subdue the negative effects.

I feel like this is the dark horse of subwoofer design because it is hard to find definitive answers. Is it good, bad, neither, either/or, etc. Why can't inductance get a proper discussion with some real answers? Just saying.

"Frequency response is NOT efficiency response."

 

Klipsch RB-35's main

Klipsch RB-35's surround

Acoustech HT-65 center

Cerwin-Vega CV-2800

 

Subwoofer

B&C 21SW152-4  [21" critical-Q]

 

kramskoi is offline  
post #150 of 233 Old 01-20-2014, 01:52 PM
 
MD28's Avatar
 
Join Date: Nov 2013
Posts: 179
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 17

i think i stirred up a giant dick swinging contest lol. i will say this... bill knows audio. excellent designer. in the near future i plan a design of his.

MD28 is offline  
Sponsored Links
Advertisement
 
Reply DIY Speakers and Subs

Tags
Svs Pb12 Plus Black Oak 12 Inch 800 Watt Powered Subwoofer , Rythmik Audio F12 Direct Servo Subwoofer , Fv15hp Subwoofer

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

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