Golden Ratio vs Tower - what should I do?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi;
Trying to update a 3 way system I built in 1980 based on reading Weems' book.
I used an 8" woofer (which I upgraded in ~2006 to a Dayton RS225S), a 2" mid and I recently got a pair SEAS 19mm tweeters to replace the original Piezo tweeters. Based on Weems' book I built the cabinet to the golden ratio dimensions and placed the drivers asymmetrically to avoid strengthening any interference effects - but of course that was way pre PC and internet era and by using "The Edge" software now I can see the placement was ok but not optimal. (Box is ported and f3=fs=fb=27Hz)
I have two choices - redo the front baffle or go for a whole new cabinet - perhaps a tower. What are the tradeoffs between a tower and a more conventional "golden ratio" shaped cabinet? From "the edge" software it looks like I can get a flatter response over 1K-5Khz range for the mid in the "golden ratio" enclosure. Won't I also reduce the impact of the baffle step in the base response by having the woofer near the floor in the "golden ratio" enclosure?
On the other hand - towers look cool, and I like the idea that the drivers would be higher up near ear level.
(is it crazy to have the woofer near the floor in a tower and the mid and tweet in the upper half? - does that affect imaging or something like that?)
Usage is in a 20'x40' Family room with cathedral ceilings. Couch is placed about 16' from the short wall where the speakers would be, a 3'x3'x18" high coffee table is between the speakers and the couch. There are 2 arm chairs on one side of the room. Back half of room is relatively empty with a couple of arm chairs in front of a fireplace.
Thanks in advance for any advice!
 
Who cares about golden ratios?! I mean I do, but I don't stop to think about it, unless you are inside the pyramid. Don't let it bring you down in your decisions. Check instead the golden ratios on the speakers of your future choices. If a tower, don't make it to much squarish or with double internal dimensions. From modern to vintage...
Look at a series of good 3-way designs for creativity, stand-mount and floorstanders.
DIY Loudspeakers
Welcome to Avalon Acoustics
Kharma Midi-Grand Ceramique 1.0 loudspeaker | Stereophile.com
kharma
http://www.diyaudio.com/forums/mult...ays-loudspeakers.html?perpage=25&pagenumber=1
http://www.diyaudio.com/forums/multi-way/149246-bass-enclosure-3-way.html
Dionysus
http://www.diyaudio.com/forums/multi-way/147632-classic-monitor-designs-7.html
JBL L100 Century
 
It's really a crapshoot designing a speaker without software. Jeff Bagby has a whole suite of free software available that can transform manufacturers measurement data into specific enclosure optimized data, and then model a crossover from it (or if you can get more accurate data like Zaph has posted that will work better).

jbagby

And yes placing a woofer near the floor can be useful as it eliminates the baffle step loss. It will also limit your crossover based on ctc spacing.
 
It's really a crapshoot designing a speaker without software. Jeff Bagby has a whole suite of free software available that can transform manufacturers measurement data into specific enclosure optimized data, and then model a crossover from it (or if you can get more accurate data like Zaph has posted that will work better).

jbagby

And yes placing a woofer near the floor can be useful as it eliminates the baffle step loss. It will also limit your crossover based on ctc spacing.

Thanks Jay1111 - Jeff's passive crossover designer software looks like exactly what I have been looking for!
Can you elaborate on what you meant by "limit your cross over based on center to center spacing"?
Thanks!
 
Keep in mind that a tower is more typically an ML-TL than a BR so a typical modeler willnot be able to accurately do a sim.

dave
Thanks Dave - trying to learn as much I can here - the more I know, the more I know I don't know crap..
What part of the sim will not be good using a tower for a BR? Is it just the ability to predict the fb? If so - as long as one can tune the port later with measurements is that ok or are there other concerns?
Thanks!
 
Who cares about golden ratios?! I mean I do, but I don't stop to think about it, unless you are inside the pyramid. Don't let it bring you down in your decisions. http://www.troelsgravesen.dk/JBL_L100.htm

Don't worry - not hung up about them. I went with them for aesthetically pleasing design and because Weems recommended them as a good starting point to design the box shape.
Let me rephrase the question - any advantages over more traditional boxy speaker enclosure layouts over long skinny ones or visa versa?
Thanks!
 
Thanks Jay1111 - Jeff's passive crossover designer software looks like exactly what I have been looking for!
Can you elaborate on what you meant by "limit your cross over based on center to center spacing"?
Thanks!

Ideally you want the crossover to be within the wavelength of the drivers ctc spacing. So if for example you have the woofers center 6.5" above the floor, and the mids center 35", you would need to crossover at or under 475hz.

Boxycad is another nice excel based program that shows you all of this info when you design a simple enclosure

http://audio.claub.net/boxycad2.xls
 
Ideally you want the crossover to be within the wavelength of the drivers ctc spacing. So if for example you have the woofers center 6.5" above the floor, and the mids center 35", you would need to crossover at or under 475hz.
http://audio.claub.net/boxycad2.xls
Ah hah - now you and Dave have answered my question - so you can't have your cake and eat it too (woofer low to mitigate baffle step and mid/tweet high to be closer to ear level) - unless the woofer is a subwoofer I guess, or active crossover I suppose. Certainly not a option for an 8" woofer crossed over at 1kHz
THANKS!
 
The BR sim does not take the 1/4 wave resonance into account and cannot predict response. One really needs to use a modeler that can do ML-TLs (ie Martin King)

dave

Thanks - this is really helpful - given this and the other comment on keeping the driver CTC closer than 1/4 wavelength of the x-over frequency, I think I will rebuild it in it's current dimensions but optimize the driver locations for edge diffraction (and go for a more rounded edge on the cabinet).
thx
 
frugal-phile™
Joined 2001
Paid Member
Ah hah - now you and Dave have answered my question - so you can't have your cake and eat it too (woofer low to mitigate baffle step and mid/tweet high to be closer to ear level) - unless the woofer is a subwoofer I guess, or active crossover I suppose. Certainly not a option for an 8" woofer crossed over at 1kHz
THANKS!

The MTMs i am listening to at the moment have a midtweeter that allows for C-C <1/4 wavelength.

dave
 
The 1/4 wavelength expectation is completely un reasonable, and the vast majority of speakers do not fit that requirement.

What's worse then being outside of a 1/4 wavelength is pushing a driver way outside of its ideal range. Start with 1 wavelength as the minimum and work out the most minimal set of compromises from there.
 
frugal-phile™
Joined 2001
Paid Member
The 1/4 wavelength expectation is completely un reasonable, and the vast majority of speakers do not fit that requirement.

What's worse then being outside of a 1/4 wavelength is pushing a driver way outside of its ideal range. Start with 1 wavelength as the minimum and work out the most minimal set of compromises from there.

With most available drivers, yes, but what you say is not absolutely true, as i said i am listening to such a system, and with 1st order XOs to boot.

dave
 
Any advantages over more traditional boxy speaker enclosure layouts over long skinny ones or visa versa?
Thanks!
BS. The baffle step compensation is done at different frequencies. Today speakers are for small houses/apartments (design) besides and most important you can have small drivers 6", 7", 8" going all the frequency range (low freq. extension) to 20Hz -20KHz on the new speakers specifications when the vintage studio/cinema speakers where good producing to 50Hz. The vintage boxy sound is very well appreciated for the good reproduction of bass from the big baffle dimensions. The skinny ones are more appreciated for their low refraction of the mid/high frequencies. Also if you think of it they are helpful because with the same volume or more they bring the voices frequencies up to listen level where the same foot print doesn't change or they are deeper instead of large. :)
Baffle Diffraction Step
Baffle Diffraction Simulator
 
This is not true. You can most certainly put the woofer near the floor and keep the mid and tweeter higher. Doesn;t need to be a sub, doesn't need to be active. The designer just needs to account for the response (i.e. measurements).

https://sites.google.com/site/undefinition/diy-sunflowers

http://speakerdesignworks.com/Cinderella.html

There are more.


Ah hah - now you and Dave have answered my question - so you can't have your cake and eat it too (woofer low to mitigate baffle step and mid/tweet high to be closer to ear level) - unless the woofer is a subwoofer I guess, or active crossover I suppose. Certainly not a option for an 8" woofer crossed over at 1kHz
THANKS!
 
He now knows one of the compromises in his speaker system.

dave
Yes - thanks, it is always a series of tradeoffs isn't it. What exactly is the impact of having the drivers further apart than a fraction of a wavelength (particularly the mid and the woofer)? Will there be much phase error for someone sitting say 16 feet away in this case? (just trying to understand this side of the compromise vs for instance changing the crossover frequency - the consequence of which I can directly see in simulations)
thanks
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.