TB W4-657SB Measures - Page 3 - diyAudio
Go Back   Home > Forums > Loudspeakers > Full Range

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 27th September 2006, 02:53 PM   #21
soongsc is offline soongsc  Taiwan
diyAudio Member
 
soongsc's Avatar
 
Join Date: Mar 2005
Location: Taiwan
MarkMck,

I take it that you think the roughness in the Jordan high frequency response is what you refer to as edge hole as well?

You also mentioned (similar to the second modification for the W4-657s), is this referring to a previous post?
__________________
Hear the real thing!
  Reply With Quote
Old 27th September 2006, 11:44 PM   #22
DDF is offline DDF  Canada
diyAudio Member
 
DDF's Avatar
 
Join Date: Sep 2004
Location: Ottawa
It may be that the need for the mod applies only to certain vintages. My units don't show signs of edge termination artefacts: the frequency response is very smooth on and off axis in this region, and there's no characteristic "blip" in the impedance trace that usually betray a badly terminated cone.

Could yours be the older "s" version? Mine are the SB models.

The great thing about these drivers is that they are remarkably stable in response over a wide off axis, all the way to 4 kHz, and they require really little to no baffle diffraction compensation.

They're prime candidates for a transient perfect first order two way. The D26NC55 is a perfect candidate to mate with them: great sound, tiny faceplate allowing close coupling, fairly excellent low distortion in the lower frequencies, and pricing in line with the TB's.

Mount the tweeter very close, tilt the cabinet back slightly, then cross using a first order and the transient response, distortion, off axis, tonal balance everything will be improved over the single driver, even eq'ed. The crossover will take a trap on the woofer, the series coil, and maybe an impedance balancing leg just to give a general up or down tilt as required to blend. The high pass will be "quasi" first order: first order until about 900Hz then it'll quickly move to third. This is really an overdamped third order, and it sounds great. The extra orders allow a bit of phase rotation to bring the tweeter more in line with the woofer, the slow roll off through crossover maintains transient response, and the greater roll off provides much better power handling. I've been using this topology on and off since the mid 90s. After 3 years of trial and error, George Short just discovered this for himself in his latest Seas based flag-ship. The topology's a killer.

Modding the driver is a worthwhile endeavour and I'm sure it'll reap great rewards. I've philosophically opted for a different path, having run some preliminary measures and a lot of modeling to convince myself. I plan to button them together this winter.
  Reply With Quote
Old 3rd October 2006, 02:16 PM   #23
MarkMcK is offline MarkMcK  United States
diyAudio Member
 
Join Date: Aug 2003
Location: Ohio
Skepticism is a good thing. When claims of "perfection" are made we don't have to operate on "trust."

We can test for ourselves. It does not require a lot of money. There is an instrumentation microphone available for around 50 US. Add another 50 for a microphone preamp by the same company.

If you already have a Windows based PC with a full duplex soundcard, then you can use Audua's free SpeakerWorkshop to complete an impulse capable LMS measuring system.

100 dollars US (plus tax and shipping) for a quasi-anechoic measurement system. Amazing really.

Last, the Hobby HIFI tests of the new Jordan show no evidence of edge holes. What they do show is most likely material vibration modes of the cone and dust cap. This does not mean that the transducer does not have a problem with edge termination, but nothing is shown. And just like the outer material vibration modes, edge termination problems rarely are severe enough to show up in an imped. plot. Indeed, why would they?
  Reply With Quote
Old 3rd October 2006, 06:37 PM   #24
MarkMcK is offline MarkMcK  United States
diyAudio Member
 
Join Date: Aug 2003
Location: Ohio
Just wanted to do a quick follow up to my last post.

Healthy skepticism is universally applied. That goes as much for what I say and what I claim as it does for anyone else.
  Reply With Quote
Old 4th October 2006, 02:38 AM   #25
MarkMcK is offline MarkMcK  United States
diyAudio Member
 
Join Date: Aug 2003
Location: Ohio
Default data revisited

I downloaded the data off the Dave's Web site. It was formatted as binary files instead of ASCII, but that was no problem.

In this and the next message I will attach impulse (time file), linear frequency response, waterfall, and step response.

There is a problem. I have an idea what it may be. For now, however, just look at the four interpretations of the time data and see if anything looks odd or out of place.
Attached Images
File Type: gif datarevisited.gif (9.0 KB, 647 views)
  Reply With Quote
Old 4th October 2006, 02:41 AM   #26
MarkMcK is offline MarkMcK  United States
diyAudio Member
 
Join Date: Aug 2003
Location: Ohio
Looking at the impulse (time data) attached to the last post, read the data from left to right. Each time the magnitude changes direction equals a "spike." Counting from the left, spikes four and six are the oddity or problem.

All other graphs are just interpretations of the same data. Still each one shows the problem.
Attached Images
File Type: gif 3dandstep.gif (14.6 KB, 495 views)
  Reply With Quote
Old 4th October 2006, 03:13 AM   #27
soongsc is offline soongsc  Taiwan
diyAudio Member
 
soongsc's Avatar
 
Join Date: Mar 2005
Location: Taiwan
MarkMck,

Are these graphs for the W4-657SB?
__________________
Hear the real thing!
  Reply With Quote
Old 4th October 2006, 04:36 PM   #28
DDF is offline DDF  Canada
diyAudio Member
 
DDF's Avatar
 
Join Date: Sep 2004
Location: Ottawa
Default LOL!

First test was to compare frequency responses, they both look the same (except for the window: I'm guessing Mark's isn't rectangular).
http://www3.sympatico.ca/dalfarra/W4657SB.gif

I haven't calculated step for this driver, but I caution not to take it out past 4 ms. My test set up is quasi anechoic to that time marker only. Anything past about 4 ms will include the very early room or floor contrinution too.

One thing to also be careful of is that the measures may have been taken with the furnace running. I always take 8 averages to get the SNR improvement, but this particular measure probably doesn't support the lower level granularity you're looking for in ETC type plots.

These were taken for frequency response, diffraction validation and to help chosse a tweeter.

I like your analysis though. What are you thinking may be happening?
  Reply With Quote
Old 4th October 2006, 05:00 PM   #29
DDF is offline DDF  Canada
diyAudio Member
 
DDF's Avatar
 
Join Date: Sep 2004
Location: Ottawa
Default Isn't that interesting!

I read your second post. Nice eye Mark!

MLS measurements (such as MLSSA, used in my test suite) are not robust when measuring non-linear systems. When MLS based measurement is applied to a non-linear test path, artificial "spikes" can appear in the impulse response.

For background:
http://ieeexplore.ieee.org/iel1/2190...onlinearity%22

Perhaps this driver has some serious linearity issues, not visible in the frequency response. I'm hoping these issues are at 8 kHz, near the break up, but now I plan on making near field THD and IM measurements, to double check this.

FWIW, I can guarantee the test equipment is linear: the MLSSA card, the mic, pre amp and measurement amp.

Let me also go back and double check the impulse, to make sure this is in the other off axis files.

All good stuff.
  Reply With Quote
Old 4th October 2006, 10:49 PM   #30
soongsc is offline soongsc  Taiwan
diyAudio Member
 
soongsc's Avatar
 
Join Date: Mar 2005
Location: Taiwan
Default Re: Isn't that interesting!

Quote:
Originally posted by DDF
I read your second post. Nice eye Mark!

MLS measurements (such as MLSSA, used in my test suite) are not robust when measuring non-linear systems. When MLS based measurement is applied to a non-linear test path, artificial "spikes" can appear in the impulse response.

For background:
http://ieeexplore.ieee.org/iel1/2190...onlinearity%22

Perhaps this driver has some serious linearity issues, not visible in the frequency response. I'm hoping these issues are at 8 kHz, near the break up, but now I plan on making near field THD and IM measurements, to double check this.

FWIW, I can guarantee the test equipment is linear: the MLSSA card, the mic, pre amp and measurement amp.

Let me also go back and double check the impulse, to make sure this is in the other off axis files.

All good stuff.
One thing to keep in mind is that what seems cone breakup may not be cone, but could also be the spider for light weight metal cone drivers like these. Slightly wet the spider with a wet cotton swab around parts of the spider to find out. If there is a change, then you know that it is.
__________________
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


Similar Threads
Thread Thread Starter Forum Replies Last Post
3875 sounds better, 3886 measures better? sendler Chip Amps 9 24th February 2009 06:30 AM
3-way around the TangBand W4-657SB: what drivers? tcpip Multi-Way 4 1st June 2006 07:07 AM
W4-657sb.... nerd of nerds Multi-Way 0 1st November 2005 01:36 PM
Charles Hansen measures audio interconnects? Steve Eddy Everything Else 1 21st June 2004 06:03 AM
pot feels broken but measures ok? travis Parts 3 12th May 2003 06:46 PM


New To Site? Need Help?

All times are GMT. The time now is 02:04 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