Featherweight Title Fight - Page 2 - diyAudio
Go Back   Home > Forums > Loudspeakers > Subwoofers

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 23rd December 2012, 08:00 PM   #11
diyAudio Member
 
Join Date: Jan 2008
Quote:
Originally Posted by Patrick Bateman View Post
I took the images available online and 'reverse engineered' the box. This is what I came up with.

The image is a bit small, but the important dimensions are in the port. Here are the details:[*]The final segment of the vent is 10.2cm long. The vent height at the final segment is 12.7cm. Note that the vent turns 90 degrees.[/list]
The 90 degree "turn" makes the vent air velocity more symmetrical in both directions, the inside is more similar to the outside than if it ended before the turn.

You will be surprised to find in the real world (not in your sim) that the 90 degree wide turn adds little to the port length.

Have fun with your sims!

Art
  Reply With Quote
Old 23rd December 2012, 09:23 PM   #12
diyAudio Member
 
Patrick Bateman's Avatar
 
Join Date: Oct 2006
Location: San Diego
Quote:
Originally Posted by weltersys View Post
I'll be looking forward to your sims:^).

When box size is reduced, and Fb is kept the same, the vented peak reduces.

The Smaart comparison I posted was a single Lab 12 in this dual design:
FREE SUB PLAN: Dual Lab12 (Front Loaded) by Welter Systems
The "too big" box results in nearly the same LF port output, with a falling upper response, very similar to the DSL CS-30.

The single Lab 12 is 5 dB more "sensitive" than the dual in the 40-50 Hz range.

Art
Click the image to open in full size.
Here's a hornresp sim of my 'guesstimate' of what the CS-30 is. The black line is with one woofer. The grey line is with two woofers, minus 3dB.*

In the sim, I'm seeing that output at resonance is basically determined by the dimensions of the vent. This makes sense, as the woofer cones are barely moving, so output is largely determine by the size of the vent, it's efficiency, and how much power is going into the woofer voice coils.

Once we're about half an octave above the tuning frequency, we see that the output of the vent is adding less and less, so the response shape is largely dictated by the woofer. (IE, the second woofer adds about 3dB more output.)

Click the image to open in full size.
Here's the impedance curve. It looks like we *do* have the classic twin impedance peaks of a vented box. The reason we don't see them in the DSL graph is that the graph is cut off at 20hz.

Click the image to open in full size.

Here's the phase response of the two vented boxes. Looks fairly close to the DSL graph.

Click the image to open in full size.
The group delay curve of the box with one woofer is quite a bit worse than with two.

I'm a bit of a noob when it comes to group delay. If I understand it correctly (and I probably don't), the reason that the single woofer box has higher group delay is that the peak in the single woofer box is narrower. (IE, the closer you can get to a 'flat line' in the SPL graph, with flat phase, the closer you'll get to flat group delay curve.)

Click the image to open in full size.
Click the image to open in full size.

It's interesting that Ivan Beaver recommended the use of the CS-30 *and* the TH-Mini. I'm guessing that it's a bit of a yin-yang thing, basically the CS-30 has a big bump in output right where the TH-Mini needs it, around 35hz. So the CS-30 almost acts like a bandpass subwoofer, augmenting the low-end of the CS-30, while also buying a bit more efficiency where the CS-30 is acting like a direct radiator (above 50hz or so.)





* if anyone is curious how I modeled these, I'll explain how in the next post

Last edited by Patrick Bateman; 23rd December 2012 at 09:26 PM.
  Reply With Quote
Old 23rd December 2012, 09:40 PM   #13
diyAudio Member
 
Patrick Bateman's Avatar
 
Join Date: Oct 2006
Location: San Diego
In my last post, I mentioned that I'd explain how I simmed these with hornresp. Here's how:

1) AFAIK, hornresp can't model vented boxes where the port isn't a constant cross section.
2) In order to model these, I basically turned the model inside out.
Click the image to open in full size.
It's the same method that I use to model back loaded horns in hornresp. You set the volume of the rear chamber "VRC" to zero. Under the 'tools' menu, you set the chamber to 'rear lined'. (Which doesn't do anything, since it's zero, but it eliminates the throat adaptor.) The volume of the throat chamber becomes the volume of the vented box. In the pic above, you can see how it's similar to the CS-30; there's a coupling chamber on one side of the woofer.
3) When doing your sims, be sure to select 'combined response' from the menu 'tools'. If you don't do that, all you'll see is the output of the port alone.

In order to compare the output of one woofer versus two, in the same box, here's what I did:

1) Model the first box
2) Copy that box to a second entry in Hornresp
3) Double the volume of everything in the second box. (Double the area of the horn segments, double the volume of the throat chamber, etc.)
4) Under the 'tools' menu, model the second sub box with two drivers in parallel, and two in series
5) Set the voltage on the second box to 2.0 volts, and the voltage on the first box to 2.83

I know that some might argue that you could get the same results by simply creating a second box, and dropping the voltage from 2.0 to 2.83V. But that only holds true for the SPL graph; the impedance graphs will be skewed by the parallel wiring. Which is why I do a series-parallel wiring scheme, with twice the box volume. Makes for an 'apples to apples' comparison of the most important graphs. (phase, spl, impedance, group delay.)
  Reply With Quote
Old 23rd December 2012, 10:06 PM   #14
diyAudio Member
 
Patrick Bateman's Avatar
 
Join Date: Oct 2006
Location: San Diego
Click the image to open in full size.
Click the image to open in full size.

Definite a yin-yang thing with these two subs. In the sims above, I've simulated a pair of CS-30s* and a single TH-Mini. It's really interesting that the Mini is struggling right where the CS-30 has displacement to burn, and vice versa. (See how the xmax peaks and dips complement each other?)

I think this is really clever. There's no reason that our subs need to match, in fact there are a lot of good reasons to have asymmetrical subs:

1) the last thing you want with your subs is to have them all run out of output at the same time
2) Vented boxes and tapped horns 'ring' due to delay; so asymmetrical impedance curves might prevent a situation where all the subs are hitting a note that's particularly obnoxious. You might notice this with a single vented sub, where certain notes can sound really obnoxious
3) Combining vented subs with tapped horns gives you the small footprint of vented boxes for ULF along with the high efficiency of tapped horns
4) Multiple subs smooth out room peaks, a la Geddes

Click the image to open in full size.
Here's the impedance curve. We see the same thing in the impedance curve that we see with the displacement curve; basically the dips in one sub complement the peaks in the other. I wonder if this might make it possible to run multiple subs in parallel? Someone who understands impedance phase would be better suited to answer this than me

Click the image to open in full size.
And here's the SPL curve, with 400 watts going into each of three subs. What a gorgeous curve! The two complement each other so well. The CS-30 extends the low frequency output of the TH-Mini, while the mini belts out the SPL above 50hz.

Note that the three subs combined may generate as much as 130dB in the three octaves from 32.5hz to 250hz. Total overkill for most homes, but would be a heck of a setup for a bar or club. And all in three boxes that weigh less than 100lbs each! In a pinch you could almost cram this into a Volkswagen Golf. (My TH-Mini clone fits just fine on the back seat of my Accord.)

WOW.
  Reply With Quote
Old 24th December 2012, 01:49 PM   #15
diyAudio Member
 
Join Date: Jan 2011
Great looking graph for the given cabinet space
But now another channel needs to have its own processing, own amp etc.

If you have the DSP power to spare and an amp to spare, this looks to be a very good combo.
  Reply With Quote
Old 24th December 2012, 06:25 PM   #16
diyAudio Member
 
Join Date: Jan 2008
Quote:
Originally Posted by Patrick Bateman View Post
Here's a hornresp sim of my 'guesstimate' of what the CS-30 is. The black line is with one woofer. The grey line is with two woofers, minus 3dB.*

In the sim, I'm seeing that output at resonance is basically determined by the dimensions of the vent. This makes sense, as the woofer cones are barely moving, so output is largely determine by the size of the vent, it's efficiency, and how much power is going into the woofer voice coils.

* if anyone is curious how I modeled these, I'll explain how in the next post
Fb is determined by the length and area of a vent.
Low level output is largely independent of vent size, though at high drive levels, a small vent is not adequate. Hornresp does show port velocity, one can estimate whether the vent is large enough by velocity.

The CS 30 looks like a Lab 12, has the power rating of a Lab 12, but could be a variant with different TS parameters.
Your simulation shows less of a low peak than the DSL charts, only 5 instead of 10 dB, while Hornresp usually exaggerates them.
I'm curious as to why you would use 2 volts and 2.83 volts on a speaker with a claimed minimum impedance of 6.5 ohms (though the CS 30 chart looks like the minimum is around 2 ohms at 20 Hz), and what you used for speaker parameters.
The Lab 12 has an impedance minimum of about 4.29 ohms, the Lab 12C (which looks the same) 3.11.

Art
  Reply With Quote
Old 24th December 2012, 07:50 PM   #17
diyAudio Member
 
Join Date: May 2009
Location: New Zealand
Patrick, take a look at the relative phase of the outputs of the CS-30 and TH-mini when used together. Trap for young players... hint: think about what happens to the phase of the output of a vented speaker below resonance.

Regarding your comments about the interleaved impedance peaks, look at Horst Moller's work with his "doppelhorns".
  Reply With Quote
Old 24th December 2012, 09:47 PM   #18
tb46 is offline tb46  United States
diyAudio Member
 
Join Date: Jan 2006
Location: Texas
Hi Patrick Bateman,

This thread may be of interest to you: Project Eburon - Step 1 - Advanced AkAbak modeling of Bassreflex and QB5 Alignments

Regards,
__________________
Oliver
  Reply With Quote
Old 25th December 2012, 02:27 PM   #19
diyAudio Member
 
Join Date: Feb 2004
Very interesting stuff. My 15" dayton ho box has very similar dimensions and response curve to the cs30, with slightly low extension, possibly more output and power handing. I had not thought about using a complex port shape.

Ultra compact ported 15" Dayton Ref HO
  Reply With Quote
Old 25th December 2012, 09:55 PM   #20
diyAudio Member
 
Patrick Bateman's Avatar
 
Join Date: Oct 2006
Location: San Diego
Quote:
Originally Posted by weltersys View Post
Fb is determined by the length and area of a vent.
Low level output is largely independent of vent size, though at high drive levels, a small vent is not adequate. Hornresp does show port velocity, one can estimate whether the vent is large enough by velocity.

The CS 30 looks like a Lab 12, has the power rating of a Lab 12, but could be a variant with different TS parameters.
Your simulation shows less of a low peak than the DSL charts, only 5 instead of 10 dB, while Hornresp usually exaggerates them.
I'm curious as to why you would use 2 volts and 2.83 volts on a speaker with a claimed minimum impedance of 6.5 ohms (though the CS 30 chart looks like the minimum is around 2 ohms at 20 Hz), and what you used for speaker parameters.
The Lab 12 has an impedance minimum of about 4.29 ohms, the Lab 12C (which looks the same) 3.11.

Art
It's funny, for the longest time I've operated on the assumption that big vents were more efficient than small vents. For instance, I'd always believed that a 20cm vent would produce more SPL than a 10cm vent, due to the increase in vent area.

But I ran some sims, and you have proved me wrong. Based on the sims, a small vent is just as efficient as a big vent.

Now I know there are some factors that hornresp can't compensate for:

1) turbulence in the port
2) compression in the port

Admittedly, these are actually less of a factor in a design like the CS-30, where there's a high tuning. (Since port compression becomes a bigger problem at lower frequencies.)

I would post some pics of the simulations of an 8" port and a 4" port, but they're practically the same picture. Much to my surprise, a larger vent doesn't increase SPL. (At least as far as Hornresp is concerned, see the caveats above.)
  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
Asynchronous I2S FIFO project, an ultimate weapon to fight the jitter iancanada Digital Line Level 3435 Yesterday 05:58 PM
Editing thread first post title doesn't change actual thread title muckrake Forum Problems 2 15th August 2012 10:13 PM
Cabinet resonance - when and why NOT to fight it? hebrer Full Range 15 10th March 2011 05:41 AM
what is the title or name of this track ravslanka Music 2 6th October 2009 07:45 PM


New To Site? Need Help?

All times are GMT. The time now is 11:56 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