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 29th April 2013, 10:13 AM   #131
AllenB is offline AllenB  Australia
diyAudio Member
 
Join Date: Oct 2008
Quote:
Originally Posted by cT equals piD View Post
then I think the designer would also want to base the values of inductance and capacitance on a load impedance that is at least a good deal closer to the actual impedance than nominal.
I think it's more to do with how different the response can be by using any single value for the impedance rather than the full plot, and the fact that tweaking is to be expected when measurements are not taken. I set up an example of this with a simple crossover changing only the impedance.

The tweeter (green) has the crossover applied (blue) for the target (black).

The pink trace shows the effect of using the full impedance plot after the phase has been set to zero. Zero degrees has no significant meaning in this case, it could be set to anything but it gives a clue toward the benefit of including it, as usually when it is left out it is left at zero.

The yellow and cyan traces show the result of setting the impedance to equal an eight ohm resistance, and a twelve ohm resistance. The peaking is because this crossover was designed for the true impedance, rather than a compensated one but this is of course irrelevant. It is the difference from the target that I'm trying to show.

The more thoroughly you can compensate the impedance, the less significant all this becomes but at some point this too is going to need measurement, and once these initial measurements are done there is no longer a need to choose a value for the impedance.
Attached Images
File Type: gif xo.gif (17.6 KB, 255 views)

Last edited by AllenB; 29th April 2013 at 10:15 AM.
  Reply With Quote
Old 30th April 2013, 12:41 AM   #132
diyAudio Member
 
Join Date: Sep 2012
Location: Maine, Bangor-area
AllenB, thanks for the feedback.

This concerns designing crossovers without measurement. So if the cross-over frequency is in the rated impedance region of a driver, it is just plain sensible to calculate the values of capacitance and inductance based on rated rather than nominal impedance. In practice, it might not make a big difference in terms of the acoustic response. But using rated instead of nominal would be logical.

A difficulty I suppose is that most often the crossover frequency cannot practically be in the rated impedance region of both of the involved drivers.

Regards,
Pete
  Reply With Quote
Old 30th April 2013, 04:47 AM   #133
Jay is offline Jay  Indonesia
diyAudio Member
 
Join Date: Feb 2003
Location: Jakarta
Quote:
Originally Posted by AllenB View Post
The more thoroughly you can compensate the impedance, the less significant all this becomes but at some point this too is going to need measurement, and once these initial measurements are done there is no longer a need to choose a value for the impedance.
Quote:
Originally Posted by cT equals piD View Post
A difficulty I suppose is that most often the crossover frequency cannot practically be in the rated impedance region of both of the involved drivers.
I'm a little bit confused regarding how you put the impedance plot (e.g. ZMA file) into the perspective of your discussion.

I haven't read/followed this thread. I was familiar with the FRD Consortium tools and method so I didn't find a need to read the thread when it was posted in 2011. But I think there may be interesting discussion so I will read the thread from the beginning now.
  Reply With Quote
Old 30th April 2013, 04:59 AM   #134
AllenB is offline AllenB  Australia
diyAudio Member
 
Join Date: Oct 2008
I went and had another think about how to show the differences in a way that would help others decide how significant this can be. The behaviour of a second order electrical filter tends toward being correct at both frequency extremes regardless of the impedance, while being more sensitive to it in the middle. A lower Q filter (such as the Linkwitz-Riley (Q=0.5) that we often start with) will tend to be more sensitive over a wider region.

I modelled an impedance that is resistive at 12 ohms below 1kHz, and 8 ohms above that. I set a target L-R rolloff at 1kHz, and also in half octave steps for two octaves either side of 1kHz. I then used calculated values of filter for each of these using 8 ohms (blue curves) and 12 ohms (red).

This way, one could look at the driver impedance curve and roughly estimate how much the response could be affected by how far away from the crossover point the impedance peak is.
Attached Images
File Type: gif 8.gif (36.0 KB, 241 views)
File Type: gif 12.gif (36.2 KB, 235 views)

Last edited by AllenB; 30th April 2013 at 05:01 AM.
  Reply With Quote
Old 8th May 2013, 05:10 PM   #135
diyAudio Member
 
Join Date: Jul 2010
Location: Western China
Ok - a little forbearance if this is posted in the wrong place.

First Issue:
I have a pair of $1 drivers from my first project that are rated on the back at 4Ω apiece.
They measure at:
Tweeter measures at 6.1Ω
Mid/bass measures at 5.4Ω
Their crossover was a 4.7uF capacitor wired to the mid/bass in series with the tweeter.

A First Order Butterworth, but without an inductor?
So what would this be called and how calculated?

Using several sources for the calc of a first order I used these values:
High Pass Impedance: 6.1Ω
Low Pass Impedance: 5.4Ω
xo Frequency Hz 5546 hz
To reverse engineer a capacitor value of 4.699uF

1) Have I performed these calculations correctly?
2) Have I assumed the correct layout of the xo as a first order has coil value and this xo had just a capacitor?

5546 seems very high even for the low price boxes this first project came from.

=============================

Second Issue:
I have purchased an AS-23 xo by Kasun, 3200hz pt, 150W rated, 3rd order with a Zobel.
I want to reverse good specs for two drivers to place into this xo in a bookshelf box (the same from the drivers above) and not mess with it anymore and move on to the next project. A 1" or 1.5" tweeter and a 3" mid/bass. The box is confining driver size choice and the xo will confine the specs.

What specs am I looking for and what calculators/software should I be using?

I realize this is backwards, but I am not up to designing an xo at this stage, nor building a box, and simply wish to pair this xo with some compatible drivers based up T-S and other specs, and finish this project and move on to the next one.

I don't want to just toss in two good drivers without being thoughtful and having intent with this first project.
Currently the drivers in it are $3 replacements I bought just to get basic project skills up to speed. Now that proof of concept is at hand, I want to put some better drivers ($30 apiece) in and finish this.

Many thanks in advance!

Last edited by miragem3i; 8th May 2013 at 05:33 PM.
  Reply With Quote
Old 9th May 2013, 01:10 AM   #136
diyAudio Member
 
Join Date: Sep 2012
Location: Maine, Bangor-area
Quote:
Originally Posted by miragem3i View Post
Ok - a little forbearance if this is posted in the wrong place.

First Issue:
I have a pair of $1 drivers from my first project that are rated on the back at 4Ω apiece.
They measure at:
Tweeter measures at 6.1Ω
Mid/bass measures at 5.4Ω
Their crossover was a 4.7uF capacitor wired to the mid/bass in series with the tweeter.

A First Order Butterworth, but without an inductor?
So what would this be called and how calculated?

Using several sources for the calc of a first order I used these values:
High Pass Impedance: 6.1Ω
Low Pass Impedance: 5.4Ω
xo Frequency Hz 5546 hz
To reverse engineer a capacitor value of 4.699uF

1) Have I performed these calculations correctly?
2) Have I assumed the correct layout of the xo as a first order has coil value and this xo had just a capacitor?

5546 seems very high even for the low price boxes this first project came from.
1) Yes, but some might object to you calculating the cross-over frequency based on the DC resistance of the tweeter. The impedance of the tweeter is always at least 1.2 times the DC resistance. I'm assuming that you got the values of impedance (actually direct current resistance) by measuring with an ohmmeter, but I could be wrong.

2) The cross-over is a high-pass filter for the tweeter only. There is no filter on the mid-range. The design relies on the natural roll-off of the mid-range.

A cut-off frequency of the tweeter at 5 kHz is not unusually high. If the size of the mid-range is fairly small, a cross-over at 5 kHz is fine.

Regards,
Pete

Last edited by cT equals piD; 9th May 2013 at 01:13 AM.
  Reply With Quote
Old 9th May 2013, 03:47 AM   #137
mart.s is offline mart.s  United Kingdom
diyAudio Member
 
Join Date: Dec 2008
Quote:
Originally Posted by miragem3i View Post
Ok - a little forbearance if this is posted in the wrong place.

First Issue:
I have a pair of $1 drivers from my first project that are rated on the back at 4Ω apiece.
They measure at:
Tweeter measures at 6.1Ω
Mid/bass measures at 5.4Ω
Their crossover was a 4.7uF capacitor wired to the mid/bass in series with the tweeter.

A First Order Butterworth, but without an inductor?
So what would this be called and how calculated?

Using several sources for the calc of a first order I used these values:
High Pass Impedance: 6.1Ω
Low Pass Impedance: 5.4Ω
xo Frequency Hz 5546 hz
To reverse engineer a capacitor value of 4.699uF

1) Have I performed these calculations correctly?
2) Have I assumed the correct layout of the xo as a first order has coil value and this xo had just a capacitor?

5546 seems very high even for the low price boxes this first project came from.

=============================

Second Issue:
I have purchased an AS-23 xo by Kasun, 3200hz pt, 150W rated, 3rd order with a Zobel.
I want to reverse good specs for two drivers to place into this xo in a bookshelf box (the same from the drivers above) and not mess with it anymore and move on to the next project. A 1" or 1.5" tweeter and a 3" mid/bass. The box is confining driver size choice and the xo will confine the specs.

What specs am I looking for and what calculators/software should I be using?

I realize this is backwards, but I am not up to designing an xo at this stage, nor building a box, and simply wish to pair this xo with some compatible drivers based up T-S and other specs, and finish this project and move on to the next one.

I don't want to just toss in two good drivers without being thoughtful and having intent with this first project.
Currently the drivers in it are $3 replacements I bought just to get basic project skills up to speed. Now that proof of concept is at hand, I want to put some better drivers ($30 apiece) in and finish this.

Many thanks in advance!
Quote:
Originally Posted by miragem3i View Post
Ok - a little forbearance if this is posted in the wrong place.

First Issue:
I have a pair of $1 drivers from my first project that are rated on the back at 4Ω apiece.
They measure at:
Tweeter measures at 6.1Ω
Mid/bass measures at 5.4Ω
Their crossover was a 4.7uF capacitor wired to the mid/bass in series with the tweeter.

A First Order Butterworth, but without an inductor?
So what would this be called and how calculated?

Using several sources for the calc of a first order I used these values:
High Pass Impedance: 6.1Ω
Low Pass Impedance: 5.4Ω
xo Frequency Hz 5546 hz
To reverse engineer a capacitor value of 4.699uF

1) Have I performed these calculations correctly?
2) Have I assumed the correct layout of the xo as a first order has coil value and this xo had just a capacitor?

5546 seems very high even for the low price boxes this first project came from.

=============================

Second Issue:
I have purchased an AS-23 xo by Kasun, 3200hz pt, 150W rated, 3rd order with a Zobel.
I want to reverse good specs for two drivers to place into this xo in a bookshelf box (the same from the drivers above) and not mess with it anymore and move on to the next project. A 1" or 1.5" tweeter and a 3" mid/bass. The box is confining driver size choice and the xo will confine the specs.

What specs am I looking for and what calculators/software should I be using?

I realize this is backwards, but I am not up to designing an xo at this stage, nor building a box, and simply wish to pair this xo with some compatible drivers based up T-S and other specs, and finish this project and move on to the next one.

I don't want to just toss in two good drivers without being thoughtful and having intent with this first project.
Currently the drivers in it are $3 replacements I bought just to get basic project skills up to speed. Now that proof of concept is at hand, I want to put some better drivers ($30 apiece) in and finish this.

Many thanks in advance!
Your calculations seem correct, and although quite crude, designs like this are still used today (I believe Epos still do this) When the drivers are well matched you can get away with very minimal X-Over like a first order, even without a Woofer inductor. 5000Hz and above X-Over is Ok with smaller woofers under 5 inches as most 4-5" woofers roll off at about 3500 - 5000Hz.
Personally Unless the drivers are especially well matched I have not had much success with first order X-Overs, Tweeters will usually need a low Fs (resonant frequency) of about 550Hz for low order X0's Tweeters over 1000Fs usually need at least a second order and preferably a third order which means a lot more expensive components. Quality tweeters from the likes of SEAS and Scanspeak with low Fs are expensive but you can then use a first or second order X-over with less parts, unfortunately you cant normally have it both ways. The problems I had with first order XO's was getting the teeters to blend in properly and not sound aggressive, I've played around for months trying to get it to sound RIGHT by ear as I don't have all the measuring equipment like most beginners. I have always reverted to 2nd order LWR or third order butterworth.
The other problem is driver impedance's are always changing when being driven by a signal (and many other reasons) so a lot of the standard formulas go out of the window and it becomes a bit of trial and error using the standard calculators just as a starting point.
With the drivers you already have I would just stick to the original design and buy some small (And preferably) pre-cut (non ported) cabinets for them and get a decent 4.7uF mkp cap and a bit of cheap foam damping to line them with and away you go.
As far as your second issue goes that would be totally the wrong way to approach it unless you are willing to basically rip the AS-23 to pieces time and time again and change values etc to get it to work with drivers you dont even have yet. You must start with the drivers and then design the X-Overs to match the specs of the drivers. Of course there is an easier way to get what you want and that is to scour the net to find all the projects using those AS-23's and see what drivers they are all using and the different mods to the X-Overs.
Unfortunately speaker design is a complicated and time consuming affair, especially if you want perfection from a limited budget.
If you are prepared to spend a bit of cash then I would look at drivers from the SEAS Prestige range or Vifa. Madisound.com and parts-express.com are good places to start.
Have a look at the Peerless 830855 SDS 4" Woofer and Vifa D27TG35-06 tweeter which are in your suggested price range. -Have fun
  Reply With Quote
Old 9th May 2013, 04:40 AM   #138
mart.s is offline mart.s  United Kingdom
diyAudio Member
 
Join Date: Dec 2008
Sorry chaps I hadn't realized you had replied to my post regarding 2nd order crossover.
The original crossover values were not known, But originally I believe it used a Seas 27TBFC/GH1212 tweeter. The x-over consisted of an ferrite core inductor for the bass driver and a 5uF cap with a 5 ohm resistor for the tweeter. The nearest x-over I could come up with that matched these parts was a first order Solen split crossed somewhere around 3000Hz with the woofer inductor about 0.6mH.
I reduced the capacitor value in the new 2nd order to 5uF and it sounds about right now with the Audax tweeter.
The new 2nd order was based on the following circuit from madisound (based on the closest match to my drivers) with a few slightly different values, And yes I have done it all by ear with no measuring tools or software, just a friend's Proac Studio 100's as reference.
I am STILL getting a little sibilance on the tweeter which I guess is due to a rise at the tweeter crossover point and I just cant be bothered to mess around with notch filters or re designing the x-over to a 3rd order which I know would help. I am lucky that the Audax tweeter I have is buttery smooth which helps somewhat. I also tried a SEAS 25tdfc a while back which I didn't like the sound of so sold on for peanuts and went back to the Audax.

Madisound x-over below:
Attached Images
File Type: png madisound crossover.png (250.7 KB, 187 views)
  Reply With Quote
Old 9th May 2013, 10:08 AM   #139
AllenB is offline AllenB  Australia
diyAudio Member
 
Join Date: Oct 2008
Quote:
Originally Posted by miragem3i View Post
A First Order Butterworth, but without an inductor?
No, but probably close.

Firstly, you will only create a first order butterworth filter with a capacitor or an inductor if the impedance is flat (and impedance phase is zero), meaning it is behaving like a resistor. Otherwise such a Butterworth filter would require more components (which in this thread has been addressed in the form of impedance compensation).

Secondly, the choice of filter depends on the response of the driver. In other words getting the driver response to resemble a Butterworth slope might require a completely different electrical filter. For example: using only a capacitor on a tweeter above the resonance causes the response to not roll off so readily. It will come back up near the resonance due to the impedance peak. Combined with the driver's own rolloff, the response may resemble the slope of a higher order filter at a lower frequency.

Quote:
Using several sources for the calc of a first order I used these values:
High Pass Impedance: 6.1Ω
Low Pass Impedance: 5.4Ω
Take a look at the impedance plot for a driver similar to yours. It will vary, sometimes higher than the nominal value (ie: 4 or 8 ohms) and in some places maybe lower. A starting point is to use a value of impedance that you feel best represents the region near the crossover point on the plot. If you don't have a plot then perhaps you can just use a slightly smaller capacitor than you calculate.

Quote:
5546 seems very high even for the low price boxes this first project came from.
This is a common approach when cost is a major factor. A woofer will be chosen that can play without a crossover and still sound half OK, then a small tweeter is chosen to fill in the top end. This method has several things wrong with it as you can see.

Quote:
I want to reverse good specs for two drivers to place into this xo
Specs are probably not what you are looking for, but rather two drivers with a similar response. This may possibly mean that the new drivers will sound similar to the originals.

Now, I suppose that if your tweeter has a lumpy response then finding a new one without the peaks could make an improvement because a simple crossover does nothing to the middle and upper treble (but a good crossover would). If you found such a tweeter it would need a similar response and impedance if you do not want to change the crossover (ie it might have a similar resonance and rated impedance). You'll also need to match the sensitivity unless you are going to change the woofer as well.

If it works now, it may or may not work with new drivers unless you are prepared to make adjustments. Perhaps you can practice with the drivers you have now? They may sound good with a better crossover.


Quote:
Originally Posted by mart.s View Post
When the drivers are well matched you can get away with very minimal X-Over like a first order, even without a Woofer inductor.
I think I can see what you're saying here, but I'd just like to clarify that matching drivers in general may not always be about things that are obvious at first glance. While some would, my own speakers wouldn't sound right with such a simple crossover and I wouldn't say my drivers were not well matched.

Quote:
driver impedance's are always changing when being driven by a signal
Such a change would be incidental. For all intents and purposes you can consider that the impedance remains what it is when measured if the operating conditions aren't changed.

Quote:
Originally Posted by mart.s View Post
The new 2nd order was based on the following circuit from madisound (based on the closest match to my drivers) with a few slightly different values, And yes I have done it all by ear with no measuring tools or software, just a friend's Proac Studio 100's as reference.
I am STILL getting a little sibilance on the tweeter which I guess is due to a rise at the tweeter crossover point and I just cant be bothered to mess around with notch filters or re designing the x-over to a 3rd order which I know would help.
Congratulations

Some may not be convinced that tweaking by ear can produce a good result. Don't get me wrong, there is a point where a plethora of measurement and calculation along with various sensible acoustic choices becomes the only way through, but even the knowledge of this can go a long way. I still tweak by ear in between other things (even when it's only to compensate for a bad recording).
  Reply With Quote
Old 9th May 2013, 05:14 PM   #140
diyAudio Member
 
Join Date: Jul 2010
Location: Western China
Thanks for the replies!

I tried the old drivers with the new xo first thing and they sounded terrible. Both tweeters had clearly seen over-loaded output from a low wattage SS amp and they were cooked above 9k.

The two replacement drivers I put into the boxes with the Kasun sounded ok until I pop tested for polarity and now I want to replace them. The originals and replacements are of such low grade that specs beyond ohms are not to be had and measurements are beyond my scope at this point.

I am going to get some drivers that would seem to work well with the natural roll offs, impedance peaks and a 3200 xo point and consider this first project as complete. It has fulfilled the goals of trouble shooting, systems processes, solder skills and basic novice procedure t/s. It does not sound bad at all, no sound-stage though. I will find some wooden boxes for the outboard xo's and move on.
  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
WinSpeakerz vs others for measurement & crossovers rick57 Multi-Way 8 12th December 2010 10:20 PM
Designing crossovers using Vidsonix Virtual Crossover triode4 Multi-Way 0 28th October 2007 01:22 AM
A brief introduction WayneM Introductions 7 30th April 2005 01:23 AM
Designing passive tri-mode crossovers? Datoyminaytah Multi-Way 9 4th September 2004 12:33 PM
Designing for guys who like designing Sch3mat1c Tubes / Valves 14 31st August 2003 11:46 AM


New To Site? Need Help?

All times are GMT. The time now is 08:21 AM.


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