Is LCR trap necessary even for ferrofluid cooled tweeter? - 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 12th September 2007, 03:00 AM   #1
Jay_WJ is offline Jay_WJ  United States
diyAudio Member
 
Join Date: Mar 2006
Location: Indiana
Default Is LCR trap necessary even for ferrofluid cooled tweeter?

I know this issue depends on which particular tweeter is used and how it is used. For example, some tweeters exhibit relatively high resonance peak even with ferrofluid damping. Or they may have somewhat highish resonance frequency compared to crossover freq being used.

But I know (and perhaps you may also know) that some people use an LCR trap even for situations where other people think it is not necessary. Is this because it really gives audible improvement or simply because of their obsessions with perfectionism? They even use an expensive capacitor in the LCR circuit---this is usually large value.

I personally don't believe in this approach. Let's first consider a situation where a notch filter is required. The Vifa XT25TG-30 (non-ferrofluid) is a tweeter that needs an LCR trap especially with low order HPF application. Here's an impedance plot of this tweeter:

Click the image to open in full size.

John Krutke used this driver in his high-end ZD5 design. He used a notch filter in the LR2 tweeter net---of course, he tells people that a 100 uf cap in the trap can be electrolytic. In this case, using an LCR trap has a big effect. I simulated the tweeter net's transfer function with an LCR trap versus without it. Here's the result:

Click the image to open in full size.

As you can see, without the trap, the tweeter's roll off will be badly modulated by the resonace and the system's performance will be ruined.

Next, take a look at the impedance plot of the new Peerless HDS 810921:

Click the image to open in full size.

It's a ferrofluid cooled tweeter with a mild but somewhat wide resonance peak at 700 Hz. Do you think an LCR trap will be needed for an LR4 1.5 kHz XO? According to my experience, I do not think so. In my application (my Usher 2-way design), I used a 1.5 kHz XO without an LCR trap. Here's a simulation result that shows transfer functions with an LCR trap versus without it. I used L = .8 mH (w/ DCR .7 ohms), C = 63 uF, and R = 9.6 ohms for the simulation.

Click the image to open in full size.

What do you think? The notch filter does not affect the region around the resonance point at all but instead attenuates upper midrange a little bit. This difference can be easily mimicked by changing other component values of the network without a notch filter, if you want to. So, in this case an LCR trap is not necessary.

What's your opinion? Will you use a notch filter even in this case?
  Reply With Quote
Old 12th September 2007, 05:59 PM   #2
sreten is offline sreten  United Kingdom
diyAudio Member
 
Join Date: Nov 2003
Location: Brighton UK
Hi,

The answer to the question is obviously no. The less obvious is that
the need for a an LCR can be reduced or eliminated by the parallel
resistor of an l-pad.

The need for an LCR is not simply dependent on the impedance,
it also depends on effective Qts and the relative crossover, in
other words the target acoustic response.

Using an LCR regardless is the same as zobelling regardless, uniformed.

/sreten.
  Reply With Quote
Old 12th September 2007, 06:51 PM   #3
Jay_WJ is offline Jay_WJ  United States
diyAudio Member
 
Join Date: Mar 2006
Location: Indiana
Quote:
Originally posted by sreten
the need for a an LCR can be reduced or eliminated by the parallel resistor of an l-pad.
I agree. Actually the crossover of the above HDS tweeter has a parallel resistor in its l-pad. This eliminates the already little need for an LCR trap.

Quote:
The need for an LCR is not simply dependent on the impedance,
it also depends on effective Qts and the relative crossover, in
other words the target acoustic response.
I also agree with this. That's why I put a note in the beginning of my post. The reason why I used transfer functions in the above examples instead of actual acoustic curves was simply that I wanted to show clean curves. Repacing them with actual acoustic roll offs in my sim has no effect on the point I wanted to communicate.

Thanks for the reply, sreten, BTW!

P.S. what do you think about using an expensive high-end capacitor in an LCR notch?
  Reply With Quote
Old 13th September 2007, 02:28 AM   #4
diyAudio Member
 
Join Date: Aug 2006
Interesting topic. I'm curious about this too. sreten do beleive their is any validity to a XO point close to the tweeter Fs causing a reactance problem that is NOT apparent in the SPL model? In other words if the acoustic slopes look good in the model, I should be ok? I've heard this reasoning before so in a recent project I used an LCR on the 810921 tweeter, but could have just as well done with out, as far as the modeling was concerned, and kept some money in wallet!

I also designed another speaker using the SS 9800 which did require an LCR, getting a match to acoustic rolloff targets was near impossible without it.

Neither of the two designs used a Zobel, which I've also heard some critizism of, although from a powerhandling pov it's much better than a single series resistor.
__________________
~Brandon
DriverVault Soma Sonus Old Driver Tests
  Reply With Quote
Old 13th September 2007, 03:18 AM   #5
Jay_WJ is offline Jay_WJ  United States
diyAudio Member
 
Join Date: Mar 2006
Location: Indiana
I think what is important is the resultant acoustic curve. Any absolute form of guideline for using an LCR trap or a Zobel is for old practice when no CAD was available. That is, they were needed for a prescribed formula to be correctly used to compute component values. For example, a Zobel net for LPF is often redundant and can be replaced by a simple electrical 2nd or 3rd order filter to obtain a target curve. Even when the same shunted CR net as the Zobel is used to obtain some kind of effect between 1st and 2nd order filter, it should not be viewed as a Zobel but as one of many topologies to reach a target acoustic curve. I think the same thing applies to an LCR trap for a tweeter net. Whether the tweeter is ferrofluid cooled or not, an LCR notch is not necessary as long as the target acoustic curve can be obtained. Of course, a non-ferrofluid tweeter with a low order HPF is more likely to require a notch.

I found that, as sreten pointed out, a parallel resistor right before the tweeter helps to flaten tweeter's impedance. I also found that a parallel resistor before the tweeter combined with a series resistor before the entire crossover --- note that this is not a conventional l-pad configuration --- helps to obtain a flat impedance profile for the tweeter net; this is not only for the suppressed resonance but also for less steep top-end and low-end impedance rise. This helps to avoid a large system impedance swing.
  Reply With Quote
Old 13th September 2007, 03:44 AM   #6
diyAudio Member
 
Join Date: Jan 2002
Location: Wellington, New Zealand
Forgetting FR curves for a moment....

I suppose any LCR / CR circuit to help flatten the impedance profile of the driver and then overall system (and minimise phase swings) would present an easier load on an amplifier.

How much of an issue this really is for a decent amplifier is probably debatable.

David.
  Reply With Quote
Old 13th September 2007, 03:51 AM   #7
diyAudio Member
 
Join Date: Aug 2006
I should have said in my previous post that neither project use a Lpad, not a zobel.
__________________
~Brandon
DriverVault Soma Sonus Old Driver Tests
  Reply With Quote
Old 13th September 2007, 05:31 AM   #8
Jay_WJ is offline Jay_WJ  United States
diyAudio Member
 
Join Date: Mar 2006
Location: Indiana
Quote:
Originally posted by Dave Bullet
Forgetting FR curves for a moment....

I suppose any LCR / CR circuit to help flatten the impedance profile of the driver and then overall system (and minimise phase swings) would present an easier load on an amplifier.

How much of an issue this really is for a decent amplifier is probably debatable.

David.
An impedance compensation notch for flatening system impedance is always right after the input terminal. This notch filter does not affect the system's frequency response at all if the amp's damping factor is sufficiently high. It only flattens the system impedance. But an LCR or CR circuit is located between a crossover and a driver, it changes not only impedance but also frequency response.

There is a reason why an LCR or CR circuit was used with text crossover formulas in the past. Just as an amp's high output impedance can modulate a speaker's FR when the speaker has non-flat impedance swings, a crossover constructed by textbook formulas may not produce intended roll off when the driver has non-flat impedance.

But with CAD today, everything can be simulated, and a Zobel becomes redundant almost always, and an LCR trap can avoided in many cases.
  Reply With Quote
Old 13th September 2007, 06:58 AM   #9
Svante is offline Svante  Sweden
diyAudio Member
 
Svante's Avatar
 
Join Date: Feb 2004
Location: Stockholm
The answer is "no".

...but not always. An LCR trap can change the frequency response, as seen in the examples above. The change can be bad, but also good. It all depends on the application. Typically, if the trap is installed after the filter, it will decrease the amplitude at the resonance. This decrease might even be nessecary for a flat response, all depending on all of the other little quirks in the response of the system.

When it comes to presenting an "easy load" to the amplifier there are two ways of looking at this. First, if the amplifier has a high output impedance, such as seen in some tube amplifiers, the frequency response will be affected. It is questionable, however, whether this is a problem of the loudspeaker or the amplifier. Second, it is often stated that a frequency dependent load is in itself difficult for the amplifier due to a non-zero phase. I question this. It is true that non-zero phase combined with a low impedance is bad. However, the peak that occurs from a resonance always results in an increase of the impedance, and higher impedances are easier to drive.
__________________
Simulate loudspeakers: Basta!
Simulate the baffle step: The Edge
  Reply With Quote
Old 13th September 2007, 04:26 PM   #10
sreten is offline sreten  United Kingdom
diyAudio Member
 
Join Date: Nov 2003
Location: Brighton UK
Quote:
Originally posted by Jay_WJ

P.S. what do you think about using an expensive high-end capacitor in an LCR notch?

Hi,

It doesn't take a genius to predict my opinion of this.
Again regardlessly using expensive components is uniformed.
Especially low DCR inductors when not needed.

It does depend on the particular circumstances but for example

http://www.zaphaudio.com/ZD5.html

The LCR correction is centred at 500Hz at which point the
tweeter output is 30dB down. LCR = 100uF-4.7R-1mH.

$25 for this :

Click the image to open in full size.

or $2 for this :

Click the image to open in full size.

Fairly obvious which one looks the most impressive.

The 1st one is a complete waste of money. it never sees high
frequencies due to the inductor or low frequencies due to the
crossover, losses are irrelevant compared to the series 4.7R.
At 30dB down the distortion of the capacitor is near irrelevant.

/sreten.
  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
series crossover with resonance trap for tweeter? madinoz Multi-Way 0 8th June 2005 03:03 AM
The Onkyo buyout tweeter from PE (ferrofluid) thomas997 Multi-Way 0 18th December 2004 02:47 AM
Ferrofluid tweeter misshap! Tino Multi-Way 3 10th January 2004 07:27 PM
Water cooled plate on ebay (possible water cooled amp application) BrianGT Parts 1 27th May 2002 04:19 PM


New To Site? Need Help?

All times are GMT. The time now is 04:54 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