Go Back   Home > Forums > >

Multi-Way Conventional loudspeakers with crossovers

Passive Loudspeaker Delay
Passive Loudspeaker Delay
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
Old 5th October 2015, 05:08 PM   #1
Patrick Bateman is offline Patrick Bateman  United States
diyAudio Member
 
Patrick Bateman's Avatar
 
Join Date: Oct 2006
Location: San Diego
Default Passive Loudspeaker Delay

Click the image to open in full size.
The Keele CBT uses a curved cabinet.

Passive Loudspeaker Delay

The JBL CBT does not.

I have another thread on a car audio forum where I'm doing a build on my car. And a member raised the idea of doing the delay passively, like JBL does. (his post is here : Car Audio | DiyMobileAudio.com | Car Stereo Forum - View Single Post - 28 Weeks Later )

So I wanted to explore ways to do this but wanted to do it in a separate thread, since the CBT has been discussed more here than there
  Reply With Quote
Old 5th October 2015, 05:26 PM   #2
Patrick Bateman is offline Patrick Bateman  United States
diyAudio Member
 
Patrick Bateman's Avatar
 
Join Date: Oct 2006
Location: San Diego
First, a bit of editorializing:

Another forum member sold his Gedlee Abbeys and purchased CBTs. I was curious about the CBTs also, as Don Keele knows a lot about horns, and for him to switch from horns to arrays, it must be something special.

Based on that, I made the effort to audition the CBTs. I heard them at the Parts Express booth at CES and was underwhelmed. To me, they had that Bose Wave Radio effect, where it's very spacious but the treble is AWOL.

But I started to think about arrays again due to a few conversations I've had. First, Jason Winslow over at Diyma mentioned that Mark Eldridge switched from horns to arrays, and that the arrays sounded better. Second, I talked to Bill Waslo and he spoke very highly about the Keele CBTs that he heard.

So I started to think about giving it another go.

If you look at the horizontal polar response of the Keele CBTs, you can see that there are some issues. This is no different than any horizontal array. I think there's a real argument to be made for flipping the CBT on it's side, similar to what Monte Kay did here:
Click the image to open in full size.

(more info here : mfk-projects CBT center speaker)

And in the March 2014 issue of Voice Coil magazine, the CBT inventors point out that the 'straight' version measures better than the curved version:

The early CBT versions were implemented with complex curved structures or discrete amplifier channels and DSP to create the time delays and attenuation—while effective, it was also complex and expensive. HARMAN /JBL engineer Doug Button worked with Don Keele on the earlier implementations of CBT and also co- authored AES papers. Button is listed as co-inventor on at least one other CBT patent, set out to explore the possibility of a simpler straight line array, passive filter expression of the CBT formulation. The patent application of this review discloses the outcome of that exploration. Basically, Button developed a method to emulate an electronically time-delayed CBT line array's performance with the use of all-passive components. The principal of operation is based on the notion that group delay with flat amplitude vs. frequency response is no different than an electronically derived time delay. Capacitive and inductive reactive com - ponents create degrees of phase shift expressed as group delay in time. The group delay of a single passive component is not flat with frequency, but combinations of inductors and capacitors can be configured to achieve flat group delay over a wide bandwidth. For many applications, the drawback is the small group delay, per L/C network section. Achieving the appropriate delay at each transducer node in the CBT array is advantageous to develop the ideal delay to provide the prescribed curvature in a straight-line array. The desired delay between each successive transducer is quite small particularly when the transducer diameters and center-to- center spacing is small, which also creates more seamless, wideband pattern control. In the passive CBT system, the desired group delay is derived by tapping points along an L/C/R network with the optimal delay and shading attenuation serially accumulating down the line. Besides the reduced cost and complexity, the passive system is claimed to also improve performance by eliminating disruptions in the smoothness of the dispersion pattern due to quantization issues that are often a side effect of standard digital implementations. While some have claimed that the purest implementation of this type of array, is by way of actual physical curvature instead of electrically derived delay, the "straight array," delay-based systems, tend to maintain the most consistent vertical polar pattern over the full 360 of horizontal angles around the device. For an assessment of CBT vs prior art approaches to directivity control see Keele's most recent AES paper, "A Performance Ranking of Seven Different Types of Loudspeaker Line Arrays". Additionally, to explore the rest of the HARMAN intellectual property on CBT technology, (see US Patent 7684574, US Patent 7826622, and US Patent 8170223). In conclusion, the patent application discloses what appears to be a more practical, simplified, and cost-reduced implementation of the very effective, high performance CBT architecture. Kudos to Don Keele and Doug Button for intro - ducing this interesting technology to the audio industry, and evolving it into an effective and practical system.

Click the image to open in full size.
Here is the passive xover for a twenty element CBT array. (via patent Patent EP2247120A2 - Passive group delay beam forming - Google Patents )
  Reply With Quote
Old 5th October 2015, 05:53 PM   #3
Patrick Bateman is offline Patrick Bateman  United States
diyAudio Member
 
Patrick Bateman's Avatar
 
Join Date: Oct 2006
Location: San Diego
I don't think that hornresp can model that xover network.

There are some spreadsheets that can do it, but I'm not certain that they can plot group delay. (Hornresp can plot the delay, but can't deal with the complexity of that network.)

Eight years ago I used Speaker Workshop to model xover networks, so let's give that a go.

The audua website is long dead, but the program is still available via the wayback machine here:

https://web.archive.org/web/20130306...s/Spkrwork.zip
  Reply With Quote
Old 5th October 2015, 07:53 PM   #4
Patrick Bateman is offline Patrick Bateman  United States
diyAudio Member
 
Patrick Bateman's Avatar
 
Join Date: Oct 2006
Location: San Diego
I did some research, and found that the crossover in the speaker that JBL sells is quite a bit different than the crossover in the patent.

This is good news actually! To me, the most obvious way of curving the wavefront is to use DSP or a curved baffle. But if you don't go that route, the most obvious solution is to use a low pass filter to delay the drivers on the top and bottom of the array. Basically you run the center drivers with a full range signal, and then you low pass the outer drivers. The low pass filter introduces a delay. Off the top of my head, a second order low pass will introduce one half wavelength of delay.

For instance, if you need to delay the outside drivers by seventeen centimeters, you could use a second order low pass filter at 1000hz to achieve that delay. (1000hz is thirty four centimeters long, and the second order delay introduces one half wavelength of delay.)

Click the image to open in full size.
Here's the actual crossover in the loudspeaker. Methinks that's just what they did - note that the four drivers in the center are running full range, while the drivers above and below are progressively delayed via a second order low pass.

Or at least I THINK that's what's going on
  Reply With Quote
Old 5th October 2015, 07:57 PM   #5
Patrick Bateman is offline Patrick Bateman  United States
diyAudio Member
 
Patrick Bateman's Avatar
 
Join Date: Oct 2006
Location: San Diego
A couple of other random observations about the xover:

1) there are a bunch of jumpers in the xover. I think those are there so that the user can choose the beamwidth. For instance, you could change the wavefront shape by changing the xover point on the drivers. Basically you can *narrow* the beamwidth by moving the xover point UP, and vice versa. That's what the jumpers are there for - they modify the capacitance value in the second order low pass network.

2) The manual mentions that there's a transformer in the circuit. Again, I think this has to do with the switches that set the beamwidth. By using a transformer you can have multiple taps, which would allow for multiple inductance levels for the inductors in the circuit.

Both of these items are ways for JBL to offer the end user a variety of beamwidths, without resorting to the expense of having two crossovers. Basically a clever way of giving the end user the flexibility of a DSP xover without the expense of ten amp channels and DSP channels.
  Reply With Quote
Old 5th October 2015, 10:25 PM   #6
Don Hills is offline Don Hills  New Zealand
diyAudio Member
 
Join Date: May 2009
Location: New Zealand
I think you could model that in AkAbak.
  Reply With Quote
Old 5th October 2015, 10:56 PM   #7
Patrick Bateman is offline Patrick Bateman  United States
diyAudio Member
 
Patrick Bateman's Avatar
 
Join Date: Oct 2006
Location: San Diego
Good point!

I'll try in speaker workshop first, but akabak is an option.
  Reply With Quote
Old 6th October 2015, 02:18 AM   #8
Patrick Bateman is offline Patrick Bateman  United States
diyAudio Member
 
Patrick Bateman's Avatar
 
Join Date: Oct 2006
Location: San Diego
Click the image to open in full size.

Looks like you could build one of these fairly easily. The drivers and the crossovers are both available. The former is JBL part number 442845.
  Reply With Quote
Old 6th October 2015, 05:23 AM   #9
bwaslo is offline bwaslo  United States
Old guy with soldering iron
diyAudio Member
 
bwaslo's Avatar
 
Join Date: May 2006
Location: Portland, Oregon!
Passive Loudspeaker Delay
You could model that pretty easily in Xsim
__________________
[W9MJE] Horn spreadsheet SynergyCalc/; SmallSyns SmallSyns;
Crossover design Xsim; Depot diffusor super-easy diffusors
  Reply With Quote
Old 6th October 2015, 05:29 AM   #10
goldyrathore is offline goldyrathore  India
diyAudio Member
 
Join Date: Nov 2002
Linking a related thread
Straight Line CBT array
  Reply With Quote

Reply


Passive Loudspeaker DelayHide this!Advertise here!
Thread Tools

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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Is passive delay circuit possible? MXM Car Audio 9 4th September 2015 11:46 PM
active vs passive time delay skogs Analog Line Level 0 22nd September 2009 09:34 PM
Time delay in passive crossover martinkruit Parts 2 2nd May 2007 08:28 PM
Passive delay/phase adjustment circuit Tenson Parts 5 10th March 2006 09:32 PM
Delay on loudspeaker switch on. lykkedk Solid State 24 18th December 2002 04:38 AM


New To Site? Need Help?

All times are GMT. The time now is 07:17 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2021 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2021 DragonByte Technologies Ltd.
Copyright ©1999-2021 diyAudio
Wiki