MiniDSP H-Frame. - Page 4 - 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 14th February 2013, 10:51 AM   #31
diyAudio Member
 
Join Date: Jun 2011
Location: Kent
I had used an LR4 at the FC as that was recommended in the ABC dipole design, I shall give your suggestion a go in using a B2 at 15Hz.

I understand that Minidsp has it's limitations regarding max gain, that's why I used a combination of both a high shelf and low shelf filter using the same center frequency 142Hz for each with 12dB and -12dB gain. What I don't understand is how you came up with two different center frequencies?

I hope I am on the right track, and thank you for taking the time to explain it to me.

Is it SQRT of FH x center shelf frequency and FC x center shelf frequency, so... (SQRT 575 x 142=285) and (SQRT 35 x 142=70)

Last edited by rich31td; 14th February 2013 at 11:11 AM.
  Reply With Quote
Old 14th February 2013, 12:41 PM   #32
diyAudio Member
 
john k...'s Avatar
 
Join Date: Aug 2004
Location: US
The LR4 recommended by ABC is for the low pass filter between the woofer and mid. ABC does not address the high pass I have recommended for limiting excursion. I have placed it at 15 Hz so that the initial woofer roll off will still follow a 2nd order, Q = 0.5 high pass response around the 35 Hz cut off. If you want to have greater protection you could use a higher order filter.

The thing about the shelves is that they have one pole and one zero at frequencies Fp and F0. The center frequency is the sqrt(Fp x F0) and gain = 20 log((F0/Fp)). So the dipole EQ has pole at 35 and zero at 575 HZ. This can be divided into two shelves by placing a pole and zero anywhere between 35 and 575 Hz. If placed at 142 Hz then both shelves with have the same gain, 12dB. One shelf has Fp = 35, F0 = 142. The other has Fp = 142, F0 = 575. When combined the pole and zero at 142 Hz cancel and you have the original response. You already understand the trick of replacing one low shelf with a high shelf to limit the total gain while retaining the correct shape of the EQ.

Hope this helps. Have fun with your woofer experiment.
__________________
John k.... Music and Design NaO Dipole Loudspeakers.
  Reply With Quote
Old 14th February 2013, 12:56 PM   #33
diyAudio Member
 
Join Date: Jun 2011
Location: Kent
Ahhhh, I see, my mistake. I had worked out how you came upon those center frequencies but not why. All has been explained now. Thanks again.

Next step, the mid range!
  Reply With Quote
Old 14th February 2013, 04:20 PM   #34
Davey is offline Davey  United States
diyAudio Member
 
Join Date: Dec 2001
Location: Bremerton, WA.
John/Rich,

To create a standard, first-order shelving filter (with gain greater than 16db) you would indeed use two filters, but the settings should be at the same frequency. If you separate the filter center frequencies (and leave the Q setting at 1) you no longer have a textbook 6db/octave shelving filter.
With some of the digital boxes (like the DCX2496) you are required to separate the programmed frequency of each filter to create the proper cascaded filter function, but not so the miniDSP since it uses "center" frequencies for the shelving settings.

Anyways, your example of 35Hz-757Hz yields a gain of 24db at center of 142hz. So, to program into a miniDSP you have three options. (All three have exactly the same curve, just offset in gain for utilizing different gain structure.)

F = 142Hz, Gain = 12, Q = 1, Low-shelf,
F = 142Hz, Gain = 12, Q = 1, Low-shelf,

Or

F = 142Hz, Gain = 12, Q = 1, Low-shelf,
F = 142Hz, Gain = -12, Q = 1, High-shelf,

Or

F = 142Hz, Gain = -12, Q = 1, High-shelf,
F = 142Hz, Gain = -12, Q = 1, High-shelf,

Attached is a photo of the middle example:

Cheers,

Dave.
Attached Images
File Type: png Untitled.png (110.3 KB, 146 views)

Last edited by Davey; 14th February 2013 at 04:38 PM.
  Reply With Quote
Old 14th February 2013, 05:51 PM   #35
diyAudio Member
 
john k...'s Avatar
 
Join Date: Aug 2004
Location: US
Quote:
Originally Posted by Davey View Post
John/Rich,

To create a standard, first-order shelving filter (with gain greater than 16db) you would indeed use two filters, but the settings should be at the same frequency. If you separate the filter center frequencies (and leave the Q setting at 1) you no longer have a textbook 6db/octave shelving filter.
With some of the digital boxes (like the DCX2496) you are required to separate the programmed frequency of each filter to create the proper cascaded filter function, but not so the miniDSP since it uses "center" frequencies for the shelving settings.

Anyways, your example of 35Hz-757Hz yields a gain of 24db at center of 142hz. So, to program into a miniDSP you have three options. (All three have exactly the same curve, just offset in gain for utilizing different gain structure.)

F = 142Hz, Gain = 12, Q = 1, Low-shelf,
F = 142Hz, Gain = 12, Q = 1, Low-shelf,

Or

F = 142Hz, Gain = 12, Q = 1, Low-shelf,
F = 142Hz, Gain = -12, Q = 1, High-shelf,

Or

F = 142Hz, Gain = -12, Q = 1, High-shelf,
F = 142Hz, Gain = -12, Q = 1, High-shelf,

Attached is a photo of the middle example:

Cheers,

Dave.
That is incorrect, Dave. This figure compares a 1st order shelf with Fc = 142, +24dB with two shelves cascaded, each with Fc = 142, +12dB.

Click the image to open in full size.

The difference is clear. A 1st order shelf has one pole and one zero in the transfer function. Cascading two identical shelved with 1/2 the gain and the same Fc yields a transfer function with two poles and two zeros. What I posted previously is correct.

When the center frequencies are staggered as I specified, at 70 and 285Hz the result is as below, perfect match.

Click the image to open in full size.

The gain of a 1st order shelf is F0/Fp= 575/35 = 16..42 > 24dB.

What you have is a 2nd order shelf with gain = (Fo/Fp)^2 = (285/70)^2 = 16.57 > 24dB. (ignoring round off errors)
__________________
John k.... Music and Design NaO Dipole Loudspeakers.

Last edited by john k...; 14th February 2013 at 06:04 PM.
  Reply With Quote
Old 14th February 2013, 07:10 PM   #36
Davey is offline Davey  United States
diyAudio Member
 
Join Date: Dec 2001
Location: Bremerton, WA.
Yep, sorry I forgot about the known "Q-issue" with the miniDSP.

You do need to alter the Q setting (on some) of the miniDSP plug-ins to achieve the textbook first-order shelf if using filters with the same frequencies.

So, for this 35-575 example set both filters to 142Hz, and 12db, and change Q from 1 to 0.8. It looks good then.

MiniDSP has adjusted/corrected the Q "problem" with some of the other/newer plug-ins, so a person needs to be aware of which plugin (and kit) is being used and check accordingly. Sorry for the possible confusion.

Cheers,

Dave.

Last edited by Davey; 14th February 2013 at 07:15 PM.
  Reply With Quote
Old 14th February 2013, 07:14 PM   #37
6.283 is offline 6.283  Germany
diyAudio Member
 
Join Date: Aug 2008
Location: Black Forest
Quote:
Originally Posted by Davey View Post
MiniDSP has adjusted/corrected the Q "problem" with some of the other/newer plug-ins, so a person needs to be aware of which plugin is being used and check accordingly. Sorry for the possible confusion.
Or use the available spread sheet and enter the coefficients in advanced mode ?
__________________
2Pi-online.de
  Reply With Quote
Old 14th February 2013, 07:42 PM   #38
Davey is offline Davey  United States
diyAudio Member
 
Join Date: Dec 2001
Location: Bremerton, WA.
Yep, that will work too. IMHO, it's preferable to keep the miniDSP programming in the "basic" mode if possible.....especially for those that are new to the platform.

It doesn't help that the default Q setting for miniDSP shelving filters is "0.5".....which doesn't correspond to a first-order shelving filter if you were to program it without changing that parameter.

There are many idiosyncrasies of the miniDSP operation a user needs to be aware of. John K. himself noticed a Q problem with the notch filtering settings a year or two ago.

Cheers,

Dave.
  Reply With Quote
Old 14th February 2013, 09:53 PM   #39
diyAudio Member
 
john k...'s Avatar
 
Join Date: Aug 2004
Location: US
Quote:
Originally Posted by Davey View Post
Yep, sorry I forgot about the known "Q-issue" with the miniDSP.

You do need to alter the Q setting (on some) of the miniDSP plug-ins to achieve the textbook first-order shelf if using filters with the same frequencies.

So, for this 35-575 example set both filters to 142Hz, and 12db, and change Q from 1 to 0.8. It looks good then.

MiniDSP has adjusted/corrected the Q "problem" with some of the other/newer plug-ins, so a person needs to be aware of which plugin (and kit) is being used and check accordingly. Sorry for the possible confusion.

Cheers,

Dave.
It is not a Q issue. It is not possible to cascade two 1st order shelves with the same gain and center frequency and have the same response as a single 1st order shelf. There cascaded pair will have two zeros and two poles. The net transfer function will be 2nd order have s^2 in both the numerator and denominator where as the 1st order shelf only has s^1.
__________________
John k.... Music and Design NaO Dipole Loudspeakers.
  Reply With Quote
Old 14th February 2013, 10:26 PM   #40
diyAudio Member
 
john k...'s Avatar
 
Join Date: Aug 2004
Location: US
Quote:
Originally Posted by Davey View Post
It doesn't help that the default Q setting for miniDSP shelving filters is "0.5".....which doesn't correspond to a first-order shelving filter if you were to program it without changing that parameter.

Cheers,

Dave.
This also depends on what version you have. I have a version of the 2way advance and the 4way advanced that require Q = 1 for a "correct" 1st order shelf. I have a newer version of the 2way advanced that requires Q = 0.5. And my 2x8 plug-in requires Q = 0.5.

For this reason, and the Q peak/notch this, I always input biquad coefficients for any EQ and then check by measuring the results and comparing them to known behavior.

I guess if I had one complaint about miniDSP it would be their deviation form the standard analog definitions of Q. Point in fact, a true, analog, 1st order shelf doesn't have a Q.
__________________
John k.... Music and Design NaO Dipole Loudspeakers.
  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
Basic miniDSP volume control question for a miniDSP n00b architect7 miniDSP 8 26th January 2012 02:40 PM
H frame vs W frame dipole woofers rick57 Multi-Way 18 8th October 2011 05:58 PM
miniDSP 2x4, a new home for miniDSP minidsp miniDSP 8 27th August 2010 02:11 AM
W-Frame or H-Frame dipoles? performace issues? JinMTVT Subwoofers 34 5th October 2006 05:44 PM
H frame/ Dipole/U frame ???????? j.burtt Multi-Way 24 1st May 2006 04:50 AM


New To Site? Need Help?

All times are GMT. The time now is 01:05 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