ribbon crossover - 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 8th April 2009, 12:50 PM   #1
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Default ribbon crossover

Hi all,
I'm thinking about an actively driven ribbon treble driver.
25W of ClassA into 8ohms.
4pole 6kHz high pass to the ribbon.

Where should the crossover be?
1.) All 4poles before the amplifier, Bessel, or Butterworth or Linkwitz?
2.) 3pole active crossover in front followed by a Passive single pole before the power amp directly feeding the driver.
3.) 2pole active crossover in front followed by a Passive single pole before the power amplifier and a Passive single pole after the power amp to protect the treble from DC offset.

Can a passive single pole be Q corrected by altering the Q of the active filter to give any or some of the Bessel, Butterworth or Linkwitz characteristics?
Similarly can a pair of passive single poles be Q corrected?
__________________
regards Andrew T.
  Reply With Quote
Old 8th April 2009, 02:01 PM   #2
timpert is offline timpert  Netherlands
diyAudio Member
 
timpert's Avatar
 
Join Date: Nov 2004
Location: Netherlands
Hi,

First a practical consideration: real ribbons require an input capacitor, because the input transformer reduces to a dead short for DC and really low frequency AC. Most amplifiers don't like this behavior. DC will saturate the transformer's core quickly, or even fry it's primary. So the input cap is mandatory. This will give you a second order rolloff (two poles) with the input transformer's primary inductance.

You need to tackle some consequences: Use a small capacitor so that the Q of the resulting filter is low. This gives you real axis poles which are easy to compensate with corresponding real axis zeroes in the active filter. Plus, the low frequency pole has only negligible influence on the overall transfer function in your frequency range of interest, which means that the exact inductance of the transformer's primary is not so critical. You may even choose not to compensate for this low-frequency pole at all and only concentrate on the high frequency one.

Focus on the acoustic transfer function you want, the electronic transfer function to achieve this may turn out to be quite different!

Quote:
Can a passive single pole be Q corrected by altering the Q of the active filter
No, you need a zero in the transfer function at the same location as the pole you wish to compensate. This is easiest with real axis poles.

Quote:
can a pair of passive single poles be Q corrected?
Sure, see above, but now you need two zeroes.
  Reply With Quote
Old 8th April 2009, 02:10 PM   #3
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Quote:
Originally posted by timpert
real ribbons require an input capacitor, because the input transformer
Hi Tim,
this ribbon does not have a transformer. Re measures 7r4
Does this alter any of your advice?
__________________
regards Andrew T.
  Reply With Quote
Old 8th April 2009, 02:17 PM   #4
Dr_EM is offline Dr_EM  United Kingdom
diyAudio Member
 
Dr_EM's Avatar
 
Join Date: Oct 2006
Location: Swindon
I am doing something similar but with a Planar. I recently decided against using a capacitor which would give a rolloff below the crossover frequency and using a standard 4 pole active filter. What I will do is your opt.2.

I am not using the capacitor to protect the amp from low impedances but rather to protect the sensitive driver from low frequencies and over excursion in this case.

I see that you are using a ribbon with no coupling transfomer, so it is a planar? In any case, it has about the same Re as mine (mine is 7.6ohm).

See attached the response of a Linkwitz Riley 24db against a butterworth 18db plus simple capacitor filter. The response is identical. My crossover point is 4.7khz or so though.

http://margo.student.utwente.nl/stef...erworth3H.html
Attached Images
File Type: gif highpass filters.gif (70.5 KB, 231 views)
  Reply With Quote
Old 8th April 2009, 02:24 PM   #5
timpert is offline timpert  Netherlands
diyAudio Member
 
timpert's Avatar
 
Join Date: Nov 2004
Location: Netherlands
You probably don't have a true ribbon then. Still, using the cap is a good idea for the protection of your (probably expensive) driver against DCoffsets or amplifier malfunction.

In this situation, you get only a single real-axis pole which you may either leave in as it is if you can use it in the overall transfer function, or compensate for it if you don't want real axis poles. But please decide about your acoustical transfer function first!
  Reply With Quote
Old 8th April 2009, 03:08 PM   #6
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Quote:
Originally posted by timpert
..a single real-axis pole which you may either leave in as it is if you can use it in the overall transfer function, or compensate for it if you don't want real axis poles. But please decide about your acoustical transfer function first!
so a 3u1F film cap between the power amp and the 8ohm driver gives a Butterworth single pole electrical filter and the DC protection you are advising.
It still needs at least a further single pole to prevent damage and I'd prefer a 3pole for a steeper cut-off @ ~6400Hz.
The cheap ribbon which is described as a dynamic type as against static type, rolls off @ -12dB/octave below ~5kHz.
Foster ct2023 from Jaycar
Quote:
Specs:
Type: Regular-phase 100mm ribbon tweeter.
Power: 20 Watts RMS (cont), 50W max.
SPL: 92dB/Watt.
Frequency Response: 6,000 to 40,000Hz +/-2dB (SUBLIME!).
Impedance: 8ohms.
Crossover Frequency: 6,400Hz (12dB / Octave).
__________________
regards Andrew T.
  Reply With Quote
Old 8th April 2009, 03:21 PM   #7
timpert is offline timpert  Netherlands
diyAudio Member
 
timpert's Avatar
 
Join Date: Nov 2004
Location: Netherlands
Anything below second order (electrical, not acoustical) as a crossover is tweeter murder unless your tweeter is a very robust affair. Indeed you need at least another pole to be safe.

But think of how you want to cross over acoustically from midrange to tweeter before deciding on the tweeter filter's poles.
  Reply With Quote
Old 8th April 2009, 03:36 PM   #8
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Quote:
Originally posted by timpert
Anything below second order........ is tweeter murder. Indeed you need at least another pole to be safe.
agreed since the manufacturer specifies 12dB/octave @ 6400Hz.
Quote:
at least another pole to be safe
does this add up to -12dB/oct or -18dB/oct?
__________________
regards Andrew T.
  Reply With Quote
Old 8th April 2009, 03:59 PM   #9
sreten is offline sreten  United Kingdom
diyAudio Member
 
Join Date: Nov 2003
Location: Brighton UK
Hi,

You could build a 3-pole filter around the amplifier loop ( use
non-inverting input as the unity gain point ) and add the 4th
pole as the DC coupling capacitor.
(Strictly speaking for the 3-pole its 1 pole passive 2 pole active,
e.g. for 3rd order Butterworth its 1st order + Q=1 2nd order)

Single positive rail supply has some advantages in this case,
not least bypassing the worse negative rail PSSR of the
typical PNP LTP NPN VAS topology.

/sreten.
  Reply With Quote
Old 8th April 2009, 04:08 PM   #10
timpert is offline timpert  Netherlands
diyAudio Member
 
timpert's Avatar
 
Join Date: Nov 2004
Location: Netherlands
One pole = 6 dB/oct, two poles = 12 dB/oct and so on. Poles come together with zeroes in pole-zero pairs. For a lowpass filter, the corresponding zeroes are at infinite distance from the origin in the complex plane, and for a highpass they are on the origin (zero frequency). That's why usually only the pole locations are discussed. However, it is also possible to locate zeroes in arbitrary places, and with active circuitry, this is very easy to do. See for examples Siegfried Linkwitz's site.

A zero and a pole on top of each other in the complex plane cancel each other out. You can use an additional well aimed pole-zero pair to cancel another so effectively you reduce the order of the overall transfer function. Your active filter may contain six poles and six zeroes and thus be of sixth order. However, if two of these pole-zero pairs are set to cancel the natural pole-zero pairs that your tweeter's response has, you have basically four pole-zero pairs left that you can place freely, shaping the acoustic response to suit your needs.

However, if you don't know your tweeter's response, designing a really good filter for it becomes quite a challenge. If you insist on making a proper filter for them, it is a good idea to first have their responses measured.
  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
ribbon crossover question ricsmuts Planars & Exotics 6 1st July 2008 12:23 AM
ribbon planar tweeter with low crossover frequency LineArray Planars & Exotics 2 19th June 2008 09:13 PM
Ribbon crossover at low Xover Tenson Multi-Way 66 24th January 2007 07:23 PM
Tweeter Box Crossover with Ribbon and Bullet Dome Tweeters doccbst Planars & Exotics 2 5th August 2005 04:54 PM
Using a ribbon tweeter with an active crossover sardonx Planars & Exotics 19 18th June 2005 04:21 PM


New To Site? Need Help?

All times are GMT. The time now is 03:10 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