How to Get Low Response Right on an OB Speakers? - diyAudio
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 May 2014, 01:31 AM   #1
ttan98 is offline ttan98  Australia
diyAudio Member
 
Join Date: Apr 2006
Location: Melb
Default How to Get Low Response Right on an OB Speakers?

I seek some inputs/advice on how to get the Bass Freq. Response right on my OB speakers.

Background
==========
My OB consists of top half which comprise tweeter, mid-range and mid woofer are these
drivers are OB and xover using passive parts. I refer to this combo as top section.
The bottom section consists of twin woofers connected in parallel and mounted on an U-
frame.
I use a 2*4 Minidsp to cross over between the top and bottom section at 200Hz, via LR2
filter from Minidsp.
It took me many months to get the top and bottom sections to get x-over right(Flat) still
the in room bass response is still not ideal.

I have read the Gainphile blog on 5 steps to OB design, and I have gained valuable insights
and info. First I measured the NF response of woofer and I applied Linkwitz Transform and
managed to get a NF response to be flat to 20 Hz. After adjusting the gains from bottom
and top sections I managed to get fairly flat response(ie top and bottom section) measured
at about 1m at tweeter level gated at 100ms in room measurement. I know this is not ideal, I still need to measure them outside.

My take so far
==============
The flat bass response so far is not really breadth taking just adequate and I need your
input to improve the in-room bass response. Should I

a. Adjust a low freq.shelf filter to improve on bass response on a trial and error basis
for better in-room bass response, any other alternatives if so where should I begin.

OR

b. Use REW software for any room bass improvements, I haven't use this software before and suggestions where I should begin and how to get the best of it.

Any inputs/advice is welcome and thanks for reading.
  Reply With Quote
Old 29th May 2014, 02:30 AM   #2
puppet is offline puppet  United States
diyAudio Member
 
Join Date: Dec 2011
Location: The Dells, WI
After reading your post I wonder how much of the "flat" measurement is the woofers rising response and cavity resonance .. 200hz ( u framed) seems quite high of a xo for a four-way design.
So your mid woofers response must be lacking in the low end?

Does miniDSP allow for a simple 12db roll off as a HP .. one that has an adjustable "Q"? .. such as : 12db .7 ... 12db .8 .... 12db 1.0 then a roll off frequency .. xxxhz . You could pick up some low range output from the top section with a little higher "Q" setting.

What I'm getting at here is you'd want the xo in the 100-120hz range. You could try a 3rd order bessel on the u frames @ 100hz with the top section rolling off around 125hz. This might add some punch but for sure you'll hear more of the LF's and probably find yourself increasing their output a little bit to match the top. Doing so will bring the LF level up .. but not in the range of 200hz. This should clean response (coherency) up a bit as well. You might not need as much shelving now in the u frames. Keeps the u frame out of the vocal range ... all good things I think.

Not sure at what frequency you have the u frames roll off but if your shooting for 20hz consider raising that up to 35hz. If you absolutely must have 20hz solid ... build a closed/vented mono sub. I don't push woofers that low in OB anymore. The trade offs in distortion aren't worth it. By raising the cut off your giving back some headroom to the woofers. Another good thing.

Hope you figure it out.
  Reply With Quote
Old 29th May 2014, 03:13 AM   #3
diyAudio Member
 
Join Date: Oct 2005
You said bass response is flat to 20Hz. That is a lot of bass energy for music listening, so your dissatisfaction may not be related to FR, but with poor transient response instead. There are two things you can do to improve transient response.

Improve woofer damping by reducing resistance in the woofer circuit. Use high quality, thick gage, pure copper speaker cable and hookup wire, not cheap junk zip cord. Use quality copper (or brass if on a budget) speaker posts, not cast zinc posts, or better still connect amp direct to drivers without posts. You would not believe how much poor wire quality and poor connections destroy bass joy, by adding source resistance which kills the ability of the amplifier to control the motion of the woofer. Hardware store lamp cord or zip wire is not pure copper even if it is for speaker hookup.

Also, don't use woofers with high Qts, because their suspension self-damping is poor. Woofers marketed "for Open Baffle use" have high Q (.7) because it allows them to play loud and low with passive crossover (no electronic boost.) The price is poor transient performance and loss of detail, no audiophile thrills, just loud and low. The driver suspension is deliberately loose and sloppy to allow the cone to fling around, using its excursion momentum to move out farther than the signal demands, this increases SPL but you lose control of the cone and the detail. You have electronic correction so you can use lower Q speakers which are better at self damping and will allow the amp to more easily control them and that gives you the detail back. You still get the LF SPL because you are putting in more power with your EQ. Pro-sound woofers are often good enough for OB with electronic boost, even though Q is .4, not as low as hifi bass drivers. Qts of .4 makes a very sexy and musical yet detailed bass quality, just add boost to taste. Pro woofers have very high sensitivity and power rating, so you can apply lots of boost with smaller amplifier. There are 12" woofers for hifi use that have very low Q, like Eton, SB Acoustics, etc. But it may sound too dry and it needs much more power and costs a lot more.

Make sure your Uframe wood is not vibrating in sympathy with the woofers. It is easy to grossly color the sound of the drivers when large area of wood is moving only a small amount. You probably don't need U frame if you have big enough area woofers and enough boost, it will only add reflection distortion.

With better amplifier control of the driver from these changes the slam factor will improve and you'll probably find that flat to 20Hz is too punchy and distracting. Good for dancing though!
  Reply With Quote
Old 29th May 2014, 05:13 AM   #4
Juhazi is offline Juhazi  Finland
diyAudio Member
 
Juhazi's Avatar
 
Join Date: Mar 2007
Location: Jyväskylä
My guess is that it is mainly the inroom conditions making things difficult. Puppet's and Ricidoo's comments are valid too I guess. I use a sealed monopole sub and go dipole mid-high, with minidsp, here is something that I have learned.

Rooms have a Schröder frequency, below this soundwaves can't roll, they oscillate and room modes are dominant. Typically this means below 2-300Hz - right where we are crossing from (sub)woofer to others. This phenomenom makes settig the system difficult and measuring it very tricky! With dipoles we can't use nearfield measurements either, then we loose the dipole effect from the measurements.

ttan, perhaps it is possible for you to set up your system outdoors or in a huge hall? This way you will find out what is the capacity of your system and you can find nice acoustic slopes for xo.

However, when this optimized speaker is brought back in a room, you are in trouble again. Dipole has a strong backwards radiation and when placed near a (front)wall, response curve is different from free air. Personally I am in even more trouble because low pass is monopole - high pass a dipole, and therefore at xo backwaves behave differently!

To sum this up - measure and set xo slopes outdoor first. Then when indoors - trust your ears!
__________________
AES Associate Member / My DIY speaker history: -74 Philips 3-way, -82 Hifi 85B, -07 Zaph L18, XLS10+PR/Hypex, -08 CSS125FR, -08 Hifitalo AW-7, -08 TangBand FR, -09 MarkK ER18DXT, -13 PPSL470, -13 AINOgradient

Last edited by Juhazi; 29th May 2014 at 05:31 AM.
  Reply With Quote
Old 29th May 2014, 11:39 AM   #5
diyAudio Member
 
Join Date: Jan 2006
Location: Cz Rep.
Just crank the bass up and enjoy the music. A flat response is only for the academics listening to weird sounds;o) Most people likes a fatter sound….

I listening to 4 naked 12 inches and have so far the best bass I ever had. And yes Juha, I trust my ears…..


Peter
  Reply With Quote
Old 29th May 2014, 03:08 PM   #6
Eldam is offline Eldam  France
diyAudio Member
 
Join Date: Sep 2012
Location: At home, sweet sound home...
Really naked , On an vertical array ? Or W arrangement ?
  Reply With Quote
Old 29th May 2014, 04:12 PM   #7
DavidL is offline DavidL  United States
Banned
 
Join Date: Jul 2010
Honestly, flat response sounds best but only near the loudest level that you will listen to. Anything lesser in volume requires a boosting in frequency at both the low and high end of the spectra.

I always thought that Yamaha had it right when they had a separate "loudness" knob on their amps/receivers. You first turned the loudness control to "Flat" so it provided a flat response, then turned the volume knob to the loudest output that you were comfortable with, then to reduce the volume to normal listening levels, you turned the loudness knob back down, "10" providing the least output but the most bass/treble boost.

Of course I always set the bass/treble controls to a bit more than flat myself. Personal preferences and all that.

Fletcher?Munson curves - Wikipedia, the free encyclopedia

Equal-loudness contour - Wikipedia, the free encyclopedia

Click the image to open in full size.

Last edited by DavidL; 29th May 2014 at 04:14 PM.
  Reply With Quote
Old 29th May 2014, 04:45 PM   #8
diyAudio Member
 
Join Date: Mar 2007
Location: California
The flaw may be in how you measured the lower cabinet response.

You obtained the nearfield response. For an open baffle system, or any other system for that matter, the nearfield response only captures what is happening very near the diaphragm. As you move out into the far field, other sources of sound interact with the sound coming from the driver itself, and this causes some changes in the response. For a typical boxed speaker, the soundwave that radiated along the plane of the baffle reflects/refracts at the cabinet edges (even if rounded) and this causes a "re-radiation" of sound as if the edges were "producing" sound. These secondary sound sources interact with the sound coming from the driver, causing ripples in the response. Also, as the sound wave from the driver propagates past the edge of the cabinet it transitions from a smaller "space" into a "larger space" and this leads to the baffle step. Neither of these can be detected in a nearfield measurement.

For an open baffle you also have the rear wave coming into play. When the microphone is very close to the front of the cone (assuming that is facing "forward") then the amplitude of the front wave is much greater than the back wave, which has to pass around the baffle. So the back wave is significantly "attenuated" relative to the front wave and you don't see it in the nearfield response either. So your nearfield measurement is not capturing the interaction of the front and rear waves that happens farther away from the speaker, e.g. at the listening position.

For example, you think your response looks like this:

Click the image to open in full size.

But it is actually more like the blue line in this plot:

Click the image to open in full size.


The difference between the above plots is the bass you are missing. Your LT has only made the nearfield response flat, but you have not accounted for the open baffle response (from what I can tell).

A nagging problem for the DIY loudspeaker designer is that it is not easy to make accurate measurements at low frequencies. The nearfield technique is a very useful tool, but you must correct for what is not measured in the nearfield. Moving outdoors and measuring at a distance will work, but this introduces lots of other problems and you need to be quite far away from obstacles like buildings, etc. to get accurate measurements to 20Hz.

Luckily, there is a clever way to "construct" a free field response for an open baffle that has been described by John Kreskovsky:
measure a dipole
In essence, you take a nearfield at the front and a nearfield at the rear, and then you combine the two into a model of the loudspeaker by reversing the polarity and adding delay to the rear nearfield measurement that corresponds to the pathlength traveled around the baffle (your Uframe in this case) to the front of the cone. This model is only valid at "low" frequencies where the radiation from the driver is non-directional, but this applies to your system. You can consider this "hybrid" response as if it was coming from the front of the cone alone. Then all you need to do is scale the SPL level down to account for the distance to the position where the other drivers were measured and, viola, you have an accurate, low frequency, free-field, open baffle response!
__________________
"Hardware eventually fails. Software eventually works" - Michael Hartung
  Reply With Quote
Old 29th May 2014, 04:54 PM   #9
diyAudio Member
 
Join Date: Jan 2006
Location: Cz Rep.
@ Eldam: In a vertical position, two/side of course with lots of amp behind and DSP. They are hanging on chains, a while I had a 6 inch wideband driver handling the entire register from approx. 200Hz. Nice but required lot of correction from 500-4500Hz. My Nad amp is broken now so I use a pair of Genelec. However new amp is on its way with 6 channel DSP/crossover/eq and Lord knows what else…

@ DavidL: Yes of course! I mostly listen at low levels but as you say, if volume is turned up, I have to lower the amp driving bass speakers….

Peter

Last edited by peterbrorsson; 29th May 2014 at 04:57 PM.
  Reply With Quote
Old 29th May 2014, 10:29 PM   #10
ttan98 is offline ttan98  Australia
diyAudio Member
 
Join Date: Apr 2006
Location: Melb
Hi,

Many thanks to everyone for their inputs especially taking the time and effort to response to my query. I find them invaluable in one way or another. I will go through them in more detail and if there are any further questions I will put them here.

Again thanks everyone. This has been an invaluable forum.
  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
REALLY extended low frequency response? Adam Reed Chip Amps 6 3rd February 2014 04:43 PM
Ribbon Low-End Response rockguitardude Planars & Exotics 14 11th July 2011 06:38 PM
My S13 OB. Uniform polar response to tweeters at last! gainphile Multi-Way 12 24th February 2010 08:03 AM
Low-distortion OB gainphile Multi-Way 88 7th May 2009 11:43 AM
OB in room response valleyman Full Range 11 6th August 2008 07:16 PM


New To Site? Need Help?

All times are GMT. The time now is 12:48 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