Building understanding of crossover network construction - AVS Forum
Forum Jump: 
 
Thread Tools
post #1 of 26 Old 12-07-2012, 11:00 AM - Thread Starter
Advanced Member
 
Dr. Spankenstein's Avatar
 
Join Date: Dec 2005
Posts: 700
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Hello all,

I've got my sights set on purchasing a trio of SEOS12/2512 flatpacks and wanted to get my head wrapped around the crossover construction. Poking around the interwebz, I came across this nice instructional page by Danny Richie (GR Research) showing a step-by-step build of a XOver.

Taking the scehmatic from the DIYSG forum, I went about trying to correlate this to the XOver that Mr. Richie was building. I'm wondering what the function of the area that I've encircled with purple would serve:


Is this what is referred to as a shunt leg? It seems to be completely independent of the other components. Maybe I'm just confused as to how the signal travels within such a network.

Can anyone explain to a simpleton, like me?


Cheers,

Bryan
Dr. Spankenstein is offline  
Sponsored Links
Advertisement
 
post #2 of 26 Old 12-07-2012, 11:08 AM
AVS Special Member
 
tuxedocivic's Avatar
 
Join Date: Feb 2010
Location: Ladysmith, BC
Posts: 5,080
Mentioned: 11 Post(s)
Tagged: 1 Thread(s)
Quoted: 554 Post(s)
Liked: 657
What you have there doesn't look like the SEOS/2512 circuit. Regardless, what you have circled in purple looks like an impedance compensation circuit.

Often (It happens to me often because I like small cheap caps on my tweeters) the low pass and high pass combine to form an impedance resonance, which is a spike in the impedance near the cross over. I never bother, but some designers don't want that there, so they use an LCR (not left, right, center - rather inductor, capacitor, resistor) across the circuit to flatten the impedance at the resonance point.
tuxedocivic is online now  
post #3 of 26 Old 12-07-2012, 11:25 AM - Thread Starter
Advanced Member
 
Dr. Spankenstein's Avatar
 
Join Date: Dec 2005
Posts: 700
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
You are correct, Tux. That photo is for a GR Research Venue OB speaker. I was merely trying to correlate how the SEOS12/2512 and this circuit was laid-out and connected.

Does impedence compensation have any benefit? It doesn't seem to have any effect on the frequency response?

Thanks for identifying what I was looking at.
Dr. Spankenstein is offline  
post #4 of 26 Old 12-07-2012, 11:39 AM
AVS Special Member
 
tuxedocivic's Avatar
 
Join Date: Feb 2010
Location: Ladysmith, BC
Posts: 5,080
Mentioned: 11 Post(s)
Tagged: 1 Thread(s)
Quoted: 554 Post(s)
Liked: 657
IMO, the benefit is very small, and even has some disadvantages. Actually, for 99% of amps, it's just a disadvantage... And a waste of parts. If I'm not mistaken, GR Research is really into wiring and botique cross overs. Things not so popular around here.

But to better answer your question, it makes the impedance profile more flat, thus making the load easier for the amplifier to drive. At least that's the argument. In reality, my feeling would be it's better to keep the impedance higher, so the amplifier doesn't have to provide as much power. It's a really small difference either way.
tuxedocivic is online now  
post #5 of 26 Old 12-07-2012, 11:47 AM
AVS Special Member
 
tuxedocivic's Avatar
 
Join Date: Feb 2010
Location: Ladysmith, BC
Posts: 5,080
Mentioned: 11 Post(s)
Tagged: 1 Thread(s)
Quoted: 554 Post(s)
Liked: 657




See the peak in the above impedance. The below is the exact same thing (cross over, drivers, frequency response) but with a quick/rough impedance compensation across the terminals. The impedance is much more flat and true to 8ohms.
tuxedocivic is online now  
post #6 of 26 Old 12-07-2012, 12:47 PM
AVS Addicted Member
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 10,129
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 1650
Quote:
Originally Posted by Dr. Spankenstein View Post

Does impedence compensation have any benefit?
Yes and no. If you aren't making a purpose built crossover that takes the actual measured impedance of the drivers into consideration then an impedance compensation circuit is necessary to provide a relatively constant impedance load to the filter, as in if you're using a simple component calculator that assumes the actual driver impedance at the crossover frequency is equal to the nominal driver impedance. If you're using the actual measured impedance at the crossover frequency a flattener is not used.

Bill Fitzmaurice Loudspeaker Design

The Laws of Physics aren't swayed by opinion.
Bill Fitzmaurice is online now  
post #7 of 26 Old 12-07-2012, 12:56 PM
AVS Special Member
 
tuxedocivic's Avatar
 
Join Date: Feb 2010
Location: Ladysmith, BC
Posts: 5,080
Mentioned: 11 Post(s)
Tagged: 1 Thread(s)
Quoted: 554 Post(s)
Liked: 657
Bill, you're talking about impedance compensation such as a zobel or an LCR around a tweeter or mid's Fs. The one in the picture is for the whole system.
tuxedocivic is online now  
post #8 of 26 Old 12-07-2012, 01:25 PM - Thread Starter
Advanced Member
 
Dr. Spankenstein's Avatar
 
Join Date: Dec 2005
Posts: 700
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
I can sort of see why it might be beneficial for the impedence curve of a speaker an/or system to be flatter. Since it impacts the load, could this detract from the systems ability to handle loud transients in the crossover region. Would this matter with solid state amplifiers?
Dr. Spankenstein is offline  
post #9 of 26 Old 12-07-2012, 02:33 PM
AVS Special Member
 
tuxedocivic's Avatar
 
Join Date: Feb 2010
Location: Ladysmith, BC
Posts: 5,080
Mentioned: 11 Post(s)
Tagged: 1 Thread(s)
Quoted: 554 Post(s)
Liked: 657
Well, when impedance is high, less current is required. So with that logic, the higher impedance is preferred rather than the compensated version. But then there's the phase angle of the curve and other issues that need looking at. IMO, it's not a big deal with solid state. I pay attention to it. I try and keep it flat. But I've never used an impedance compensation network. Even with the parts on hand. Just not even worth the solder, to me. Others who know more about amps may step in and correct me.
tuxedocivic is online now  
post #10 of 26 Old 12-07-2012, 03:04 PM
AVS Addicted Member
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 10,129
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 1650
Quote:
Originally Posted by tuxedocivic View Post

Bill, you're talking about impedance compensation such as a zobel or an LCR around a tweeter or mid's Fs. The one in the picture is for the whole system.
I see no need for that unless the intent is to use the speaker with a tube amp. But there are much better ways to address that issue too.

Bill Fitzmaurice Loudspeaker Design

The Laws of Physics aren't swayed by opinion.
Bill Fitzmaurice is online now  
post #11 of 26 Old 12-07-2012, 04:07 PM
Senior Member
 
kcnitro07's Avatar
 
Join Date: Jul 2011
Posts: 303
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
Quote:
Originally Posted by tuxedocivic View Post

Bill, you're talking about impedance compensation such as a zobel or an LCR around a tweeter or mid's Fs. The one in the picture is for the whole system.

So are you suggesting you don't need the "LCR" or the zobel? or either? I'm looking into buidling cross overs as well and was reading about the zobel and it seemed beneficial to me.

kcnitro07 is offline  
post #12 of 26 Old 12-07-2012, 04:13 PM
AVS Special Member
 
tuxedocivic's Avatar
 
Join Date: Feb 2010
Location: Ladysmith, BC
Posts: 5,080
Mentioned: 11 Post(s)
Tagged: 1 Thread(s)
Quoted: 554 Post(s)
Liked: 657
Zobels are different, and can be useful. I don't use them often. Sometimes I'll use a "soft" zobel (not sure if there is a technical term for it) where it's like another pole in the cross over but I add some series resistance, like 2ohms. Like a shaping circuit or something.

Anyways, I think Bill and I are saying that a global impedance compensation network is a waste of amp power and money. Individual driver impedance compensation can be very useful. They are different, and you have to know what they do in order to use them properly.
tuxedocivic is online now  
post #13 of 26 Old 12-07-2012, 04:44 PM
Senior Member
 
kcnitro07's Avatar
 
Join Date: Jul 2011
Posts: 303
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
Quote:
Originally Posted by tuxedocivic View Post

Zobels are different, and can be useful. I don't use them often. Sometimes I'll use a "soft" zobel (not sure if there is a technical term for it) where it's like another pole in the cross over but I add some series resistance, like 2ohms. Like a shaping circuit or something.
Anyways, I think Bill and I are saying that a global impedance compensation network is a waste of amp power and money. Individual driver impedance compensation can be very useful. They are different, and you have to know what they do in order to use them properly.

Thats what I thought I was picking up from what you guys were putting down, but figured I'd double or triple verify :P

kcnitro07 is offline  
post #14 of 26 Old 12-07-2012, 08:36 PM
AVS Addicted Member
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 10,129
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 1650
Quote:
Originally Posted by kcnitro07 View Post

was reading about the zobel and it seemed beneficial to me.
Only if you don't know how to design crossovers. If you do they're just wasted parts. But designing crossovers is even more complicated than designing cabs, so if you just want to go with a simple crossover calculator a zobel will allow you to do so.

Bill Fitzmaurice Loudspeaker Design

The Laws of Physics aren't swayed by opinion.
Bill Fitzmaurice is online now  
post #15 of 26 Old 12-08-2012, 07:27 AM
AVS Special Member
 
Looneybomber's Avatar
 
Join Date: Sep 2006
Location: Lawrence, KS
Posts: 4,668
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 23
Quote:
Originally Posted by tuxedocivic View Post

the low pass and high pass combine to form an impedance resonance, which is a spike in the impedance near the cross over. I never bother, but some designers don't want that there, so they use an LCR (not left, right, center - rather inductor, capacitor, resistor) across the circuit to flatten the impedance at the resonance point.
How does this work? I know the RCL is a 2nd order ODE (Ordinary Differential Equation), so how, mathematically, does this circuit interact with the other circuit?

YID DIY
Looneybomber is offline  
post #16 of 26 Old 12-08-2012, 08:32 AM
AVS Addicted Member
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 10,129
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 1650
Quote:
Originally Posted by Looneybomber View Post

How does this work? I know the RCL is a 2nd order ODE (Ordinary Differential Equation), so how, mathematically, does this circuit interact with the other circuit?
That can happen, but it doesn't have to, and in most cases with a well designed system, it doesn't to the extent shown.

Bill Fitzmaurice Loudspeaker Design

The Laws of Physics aren't swayed by opinion.
Bill Fitzmaurice is online now  
post #17 of 26 Old 12-08-2012, 09:24 AM
AVS Special Member
 
Looneybomber's Avatar
 
Join Date: Sep 2006
Location: Lawrence, KS
Posts: 4,668
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 23
I'm a sophomore (technically a senior by credits tongue.gif) working on a BSEE, so I know a little about circuits, and I'm trying to figure out how someone can put an RCL infront (in series or parallel?) of another circuit to reduce impedence spikes.

YID DIY
Looneybomber is offline  
post #18 of 26 Old 12-08-2012, 10:01 AM
AVS Special Member
 
tuxedocivic's Avatar
 
Join Date: Feb 2010
Location: Ladysmith, BC
Posts: 5,080
Mentioned: 11 Post(s)
Tagged: 1 Thread(s)
Quoted: 554 Post(s)
Liked: 657
I don't know much about the electrical side of xo design. Your the one half way to an EE. I only took 1st and 2nd year EE courses. My background is more wave physics related. Anyways. The way I look at it is this. The L and C are gate keepers on either side of the impedance peak. The R controls how much to damp the impedance peak. It's related to the formula you know:

1/R1 + 1/R2 + 1/Rn ... = 1/R

So in the example I posted above, the LCR values were something like 0.5mH, 15uF, and 12ohm. The L sets the compensation region below about 4khz and the C above about 800hz (just guessing here). The R mixes with the system impedance in between those gates according to the R formula above. That is the final impedance that the amp sees. So some of the amp current passes through this LCR and is totally wasted.

I could be totally off with that explanation. biggrin.gif
tuxedocivic is online now  
post #19 of 26 Old 12-08-2012, 10:06 AM
AVS Addicted Member
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 10,129
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 1650
Quote:
Originally Posted by tuxedocivic View Post


I don't know what the upper chart shows, as there are no impedance peaks where the woofer is loaded by the cab and there's no rising impedance with frequency to correspond with the tweeter voice coil inductance.

Bill Fitzmaurice Loudspeaker Design

The Laws of Physics aren't swayed by opinion.
Bill Fitzmaurice is online now  
post #20 of 26 Old 12-08-2012, 11:23 AM
AVS Special Member
 
tuxedocivic's Avatar
 
Join Date: Feb 2010
Location: Ladysmith, BC
Posts: 5,080
Mentioned: 11 Post(s)
Tagged: 1 Thread(s)
Quoted: 554 Post(s)
Liked: 657
Those are not real. Just made up.
tuxedocivic is online now  
post #21 of 26 Old 12-08-2012, 12:48 PM
AVS Addicted Member
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 10,129
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 1650
Quote:
Originally Posted by tuxedocivic View Post

Those are not real. Just made up.
IMO that further confuses the issue.

Bill Fitzmaurice Loudspeaker Design

The Laws of Physics aren't swayed by opinion.
Bill Fitzmaurice is online now  
post #22 of 26 Old 12-08-2012, 03:03 PM
AVS Special Member
 
tuxedocivic's Avatar
 
Join Date: Feb 2010
Location: Ladysmith, BC
Posts: 5,080
Mentioned: 11 Post(s)
Tagged: 1 Thread(s)
Quoted: 554 Post(s)
Liked: 657
Ok post your own.

It show a spike in the impedance around a cross over, and what the compensation network does to it.
tuxedocivic is online now  
post #23 of 26 Old 12-08-2012, 04:06 PM
AVS Addicted Member
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 10,129
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 1650
This is the measured impedance of a 2 way speaker, crossed over at 1.6kHz. Note there is no impedance spike there, in fact there's a dip, as is usually the case, depending on the individual driver impedances and filter Qs.


Bill Fitzmaurice Loudspeaker Design

The Laws of Physics aren't swayed by opinion.
Bill Fitzmaurice is online now  
post #24 of 26 Old 12-08-2012, 04:17 PM
AVS Special Member
 
tuxedocivic's Avatar
 
Join Date: Feb 2010
Location: Ladysmith, BC
Posts: 5,080
Mentioned: 11 Post(s)
Tagged: 1 Thread(s)
Quoted: 554 Post(s)
Liked: 657
Your graph does not show us in any way what the LCR circled in purple does.

Oh and in my experience a dip in impedance at the cross over is not usually the case.
tuxedocivic is online now  
post #25 of 26 Old 12-08-2012, 05:20 PM
AVS Addicted Member
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 10,129
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 1650
Quote:
Originally Posted by tuxedocivic View Post

Your graph does not show us in any way what the LCR circled in purple does.
Oh and in my experience a dip in impedance at the cross over is not usually the case.
I have no idea what it does, as I don't have either the schematic nor the interest in creating one. IME, going back 45 years and five hundred odd designs, is that an impedance dip at the crossover frequency is usually present; one common error newbies make is tuning the crossover by ear without measuring the impedance, ending up with a dip so low that it creates loading issues.

Bill Fitzmaurice Loudspeaker Design

The Laws of Physics aren't swayed by opinion.
Bill Fitzmaurice is online now  
post #26 of 26 Old 12-08-2012, 05:34 PM
AVS Special Member
 
tuxedocivic's Avatar
 
Join Date: Feb 2010
Location: Ladysmith, BC
Posts: 5,080
Mentioned: 11 Post(s)
Tagged: 1 Thread(s)
Quoted: 554 Post(s)
Liked: 657
Bill, the OP asked what the LCR circled in purple does. I posted a before and after effect of the circuit. You then said it confuses the issue and posted a random impedance curve.

As for the dip, it's irrelevant. I was only saying the way I design I usually get a peak because I choose drivers with overlapping passband and horn loading. Correcting for baffle step and the horn loading with the cross over often gives me a peak. I see it in other peoples designs also. Probably in the design posted in th OP also.

As I said. It's irrelevant. I was trying to explain visually what the LCR do. My graphs do it, IMO. I'm not sure why you were critical of a fake sim to illustrate a point. Would you like me to post some real ones? It'll take a few days as I never use this type of circuit so I don't have any. I could post another simulation. But you can do that to. You have years of experience after all, not sure how you couldn't see what my graphs were illustrating. Thanks for the random impedance curve of a speaker that does not require this circuit (at least doesn't require it near the cross over).
tuxedocivic is online now  
Reply DIY Speakers and Subs

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