Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Multi-Way Conventional loudspeakers with crossovers

The rationale behind choosing particular drivers?
The rationale behind choosing particular drivers?
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 9th January 2018, 04:56 PM   #11
andy2 is offline andy2  United States
diyAudio Member
 
Join Date: Nov 2004
Location: StupidVille
I would advise to buy the best drivers you can afford. There are "real" reasons why they are expensive. You can't go wrong with either Seas or ScanSpeak. SBAcoustic are OK if that is all you could afford. Accuton is best if you are good with xover. The rests are hits or misses.
  Reply With Quote
Old 9th January 2018, 06:18 PM   #12
Bill poster is offline Bill poster  Thailand
diyAudio Member
 
Bill poster's Avatar
 
Join Date: Jun 2008
Location: Bangkok/London
SB are more than ok. But not always easy to use

Check out Tangband's W5 poly and W3/W4 bamboo drivers.
Also the Peerless HDS ppb and nomex range of midwoofers

Easy to use, great sounding and not expensive

Last edited by Bill poster; 9th January 2018 at 06:25 PM.
  Reply With Quote
Old 9th January 2018, 10:33 PM   #13
ScottG is offline ScottG  United States
diyAudio Member
 
Join Date: Feb 2003
Location: US
Quote:
Originally Posted by CharlieLaub View Post

..One way to get a feel for these thigns, and how important (or not) they may be, is to start with a not-too-expensive but robust woofer and tweeter (say a 6" woofer and 1" dome tweeter), put these in a cabinet, and then use a DSP to implement an active crossover for the loudspeaker pair. DSP makes it so easy to play with all the variables in the crossover, even in real time. Then you can make a change and listen. This will help you gain invaluable experience with crossover design choices without having to build a bunch of loudspeakers to get it.

You could buy a miniDSP 2x4HD and a 4-channel power amp and have all that you need to implement this scheme. It would also be good to have a calibrated measurement mic (starts around $50) and a computer. Taking certain measurements of the loudspeaker system will cut down on the guesswork significantly and reduce frustration.

This makes me think of workflow. Of late I've seen all sorts of variations, and it's not that they are necessarily bad or end in a bad result, but that things really could be done a bit better. With that in mind:


1st: Break-in drivers (..see other sources on that).

2nd: Run driver through T/S testing with Impedance measured and a "rub-buzz" test. Look at parameters to see if they aren't horrible. Look at impedance to see if there is a serious problem. Listen during rub buzz test for, you guessed it: rubbing and buzzing. Sort out faulty drivers for return.

3rd: test on large open baffle with chamfered rear (of baffle) for driver if required. Make sure front of baffle is "clean" usually with re-usable putty near driver and baffle intersection. Listen to the signature as it goes through its "sweep".

Zaph's (always a good source) is a good example of a good test baffle:

Zaph|Audio

Zaph|Audio


4th: Import that data into a modeling program and start modeling. Make sure level differences are accounted for (..not just with respect to the overall design, but particularly level differences between the same drivers for the left and right loudspeaker). Generate a variety of designs based on physical constraints for your intended loudspeaker. Look at axial response variations (both vertically and horizontally) at different distances.


5th: ONLY once you've got a few models you'd like to try: then physically create them (the boxes) and usually with re-usable parts. I tend to use cardboard + round-over stock of varying edge profiles (a grouping of stock for various lengths) + tape (particularly for corners and around drivers) + damping for the cardboard (mostly interior of "box") + wood frame (again from grouping of stock of various lengths). You can build a nice "erector"-set of materials for less than $100. As you become more proficient - you can add clay sculpting to create more complex baffles (..I'd personally really like a nice 3D printer for this portion).


6th: Then try them with various crossovers DIGITALLY (..I like Soundeasy for this). Play with loudspeaker tilt a bit. Constantly measure the result. Alter crossovers. Constantly listening for differences.


7th: Only once you have a design you feel comfortable with should you start cutting into wood.

8th: listen again, if excellent - "push" to passive component crossover and listen again (make adjustments as required) for final result.


IF you don't do it like this you are asking for:

A. a mediocre result (..though who knows, you could get lucky). It's AMAZING what modest alterations in design can do for a loudspeaker with respect to their "sound". Near identical designs can sound quite different with just modest changes in crossover and BAFFLE (and driver placement). Box volume on the other hand is pretty much as modeled, but external shape (beyond the baffle) and internal shape (relative to how the drivers react) can also impact the result significantly (though usually not as much as the crossover/baffle/driver placement).

B. a LOT of scrap wood from just a few changes.



..of late, I've been seeing people do very basic box modeling for volume, cutting wood and building the box, putting drivers in, measuring, and then modeling a crossover from that in something like xsim. You can get a decent result from this, but you'll rarely get that "dialed-in" extraordinary result. You can even add-in the digital crossover to this work-flow to get a better result, but that does substantially limit the baffle shape, driver placement, etc. that combine with the crossover for that excellent result. If you want to make a box change - then you are back to cutting wood.
__________________
perspective is everything

Last edited by ScottG; 9th January 2018 at 10:37 PM.
  Reply With Quote
Old 9th January 2018, 11:05 PM   #14
ScottG is offline ScottG  United States
diyAudio Member
 
Join Date: Feb 2003
Location: US
As for the thread topic..

1. "Efficiency" (average) + bandwidth + linearity for the operating pass-band and matching to other drivers, respective of a range of suitable baffle shapes.

2. Directivity pattern linearity (off-axis performance) in conjunction with various crossovers (below, at, and above crossover).

3. Average impedance, Mms, Fs, Vas, linear excursion (and a few other parameters) - relative to intended us.


Note: 2 and 3 are often "flipped" depending on driver type and basic design. IF I'm just looking for a woofer for a 3-way I look at #3 and don't care much about #2. But if I'm looking for a 2-way midbass then #2 gets priority.


4. Materials choice, particularly surround material and shape for a given driver. Then motor and basket (..and air flow).

5. Resulting decay profile.

6. Resulting non-linear distortion profile.
__________________
perspective is everything

Last edited by ScottG; 9th January 2018 at 11:11 PM.
  Reply With Quote
Old 10th January 2018, 12:50 AM   #15
Flaxxer is offline Flaxxer  United States
diyAudio Member
 
Join Date: Dec 2017
Location: Albion, CA
Wow ... Thanks to you CharlieLaub. I love this !!!

Damn Scott ... talk about owing a big "thank you". Your posts above just opened up everything to me now. Very basic information I had not considered.
I am very good with wood. I'm going to build three total enclosures. Large, med, and small. Each COMPLETELY modular via rabbetting. I will be able to alter volume, sealed or ported, and baffle infinitely via slide or bolt in walls, ports, slots, and baffles. This will also allow me to experiment with damping options as well. One for bottom cabinets. One for large to smaller bookshelf speakers, and one for smallish speakers. Thanks for quoting CharlieLaub, and expounding.

I already have a 4 channel amp, and a MiniDSP 2x8, with a measurement mic on the way as of yesterday. I have a good laptop and REQWizard installed.

And Bill Poster ... THIS is the kind of info on drivers I need! Saves me funds on notoriously difficult drivers to design with.

I feel like I just won the lottery !
  Reply With Quote
Old 10th January 2018, 12:16 PM   #16
TMM is offline TMM  Australia
diyAudio Member
 
Join Date: Sep 2007
Location: Australia
Send a message via AIM to TMM
I choose drivers by looking at their objective performance / distortion. Descriptions like "muddy", "laid back", "snappy", "warm" etc are subjective descriptions of a complex cocktail of distortion. It is extremely difficult to judge a single driver's performance when it is part of a complete multi-way speaker. Ever listened to speaker after disconnecting one or more of the drivers? It sounds like garbage, even if the remaining drivers are performing well. I could go on forever about the limitations of subjective analysis but there is a good run down here: Zaph|Audio


The idea of a crossover is to 'hide' as many driver defects as possible by choosing a crossover frequency such that none of the drivers have to play in frequency ranges which they have defects. If you have a woofer with a defect at 3kHz and a tweeter which distorts below 1kHz, by choosing a crossover frequency of 2kHz you hide the problems of both drivers.

So what defects are there? Well personally I consider (in order of importance):

-Non linear distortion i.e. harmonic distortion, intermodulation distortion. This is rarely published by the manufacturers. Zaph Audio is one of the few places you will find comparable non-linear distortion data for many drivers. In most cases you will have to measure non-linear distortion yourself. I prefer doing harmonic distortion sweeps since they are easier to interpret than IMD measurements and they show the non-linear behaviour of a driver vs. frequency. If you use a driver over a frequency range which it delivers low harmonic distortion in a harmonic distortion sweep, you will generally avoid most non-linear distortion problems. Whatever non-linear distortion is produced by the driver, you are stuck with it. The only thing you can do is avoid using a driver through frequency ranges where it is unacceptably bad. Crossover to another driver which doesn't have a non-linear distortion problem through that frequency range.

-Linear distortion i.e. frequency response. Generally the usable frequency range of a driver is determined by it's mechanical resonance (fs) at the low end and cone breakup at the upper end. A driver with a flat frequency response will be easier to integrate into a system than a driver which has a response that looks like a rollercoaster. That said, linear distortion can be 'fixed'. With enough crossover/filter work you can hammer any response into a flat line, just that you should prefer to not have to resort to it (or put up listening to a non-flat response!).

-Impedance and Sensitivity. Generally you want the tweeter to have the highest sensitivity and the woofer to have the lowest sensitivity. It is much easier to attenuate a tweeter in a passive crossover than it is to attenuate a woofer because to create a constant attenuation vs frequency usually requires building a conjugate network to flatten the drivers impedance. Flattening the impedance of a driver with a low resonant frequency requires large valued, big and expensive crossover components. The other thing to consider - especially for tweeters - is that there aren't multiple resonant impedance peaks or other impedance abnormalities. If there are, this may require additional work flattening the impedance to get your crossover filters to behave. If you plan to use multiple woofers in parallel then you may need to consider that the overall impedance doesn't become too low for your amplifier to drive.

-T/S parameters. Really only applicable for bass drivers, determines how big the enclosures will be and how low they will go.

-Physical Appearance. Pretty obvious this one - who doesn't like their speakers looking pretty? That said, just because some speakers drivers belong to the same 'series' does not mean that they are a good match for each other. Fair warning that just because say a 7" woofer is a good performer doesn't mean that it's 4" brother in the same 'series' is also a good performer or that they will integrate well together.
There's no reason you can't match a paper woofer with an aluminium midrange and a soft dome tweeter. That said, it is likely that drivers with contrasting construction materials will look strange together so often people avoid using sets of drivers that visually contrast each other. I'll admit that I have been guilty of avoiding using certain drivers together for that reason too.

-Price. I have to disagree that 'you get what you pay for'. There are some very expensive drivers that are simply garbage, and there are some very affordable drivers which could be used in a world-class system provided that you implement the crossover perfectly. Many manufacturers produce both excellent drivers and mediocre ones at similar price points. Some manufacturers have consistently mediocre or poor drivers at ridiculous price points, and some manufacturers have consistently good to excellent drivers but at consistently high price points.
What does seem to be the case is that the more expensive drivers from any given manufacturers line up generally have more benign distortion problems, so you are more likely to produce acceptable sounding speakers without being an expert crossover designer.
As an example, there is a $200 7" Danish-built woofer that does not have any severe distortion problems out to 10kHz so you could literally cross it at any frequency without major disaster. There is also a $50 7" Asian-built woofer which performs on par with the Danish woofer to about 1kHz, after which it suffers some rather severe linear and non-linear distortion problems due to cone breakup. As long as you cross it below 1kHz, it performs as well as the Danish woofer. Yet many people do not like this Asian woofer because they tried to cross it above 1kHz and did not get good results. If you are building a 3-way system which requires a crossover point of 800Hz, which one do you buy? I'd buy the Asian one of course! Then I have $150 to spend on a better mid/tweeter. That is why you should never make a subjective assessment of a driver when it is in a finished speaker - how would you know about the problems of a driver if the crossover has been skillfully designed to hide them?
...And finally there are also manufacturers that produce 7" woofers which are more expensive than said Danish 7", yet are out performed in every possible way by both the Danish 7" and the Asian 7"

Last edited by TMM; 10th January 2018 at 12:34 PM.
  Reply With Quote
Old 11th January 2018, 03:53 AM   #17
Flaxxer is offline Flaxxer  United States
diyAudio Member
 
Join Date: Dec 2017
Location: Albion, CA
Quick questions ... I am not trying to do this. But knowing the answer will help me understand something I am reading abt.

If you design a well integrated 2 way xover. Then later want to change to a three way, does the low pass you add effect what you had in the two way already ? Do you have to re-tweak, if the two way is staying the same ?

How does adding a low pass, now making a bandpass, effect the slope or inductance of the existing two way ?

This is hard putting in words

And the second questions is, I understand choosing low QTS drivers, and why. But only from curiosity, what things can be done to make high QTS drivers work in enclosures? Does AP membranes work ? Or are they only suitable for OB ?
  Reply With Quote
Old 11th January 2018, 12:45 PM   #18
Zvu is offline Zvu  Serbia
diyAudio Member
 
Zvu's Avatar
 
Join Date: Jan 2009
Location: Belgrade
It has huge effect on performance. Please read this:

why
__________________
Not for fainthearted
  Reply With Quote
Old 11th January 2018, 02:45 PM   #19
Flaxxer is offline Flaxxer  United States
diyAudio Member
 
Join Date: Dec 2017
Location: Albion, CA
Quote:
Originally Posted by Zvu View Post
It has huge effect on performance. Please read this:

why
Perfect! That is just the info I was looking for! I'm currently working on learning how everything effects the freq response. As well as how to choose drivers, based on different parameters. So this helps a lot. I can't wait until I receive the microphone, where I can start seeing it first hand.

PS: Mic not here yet, but SE Scans are arriving sometime today.
  Reply With Quote
Old 11th January 2018, 04:35 PM   #20
Zvu is offline Zvu  Serbia
diyAudio Member
 
Zvu's Avatar
 
Join Date: Jan 2009
Location: Belgrade
Nah. I don't believe anything without pics
__________________
Not for fainthearted
  Reply With Quote

Reply


The rationale behind choosing particular drivers?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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Help choosing drivers please CUTTING EDGE CNC Full Range 9 9th October 2016 06:41 PM
Rational Audiophile Loudspeaker System danvprod Multi-Way 56 23rd March 2015 10:59 AM
rational approach traw Multi-Way 2 15th August 2006 07:39 AM
Choosing drivers pia2o Multi-Way 2 28th February 2004 08:30 PM
Need help choosing 3-way drivers AudioIsFun Multi-Way 7 3rd January 2004 04:03 PM


New To Site? Need Help?

All times are GMT. The time now is 10:33 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.79%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio
Wiki