Welcome, Guest
Username: Password: Remember me
This is the optional Forum header for the Suggestion Box.

TOPIC:

Re:Linkwitz transform 11 years 9 months ago #919

  • devteam
  • devteam's Avatar
  • Away
  • Administrator
  • Administrator
  • Posts: 9049
  • Thank you received: 1376
Geez, you guys are starting to seriously compete with the LT spreadsheet running around the DIY audio forums...Neat!

Keep up the good works and let us know how this implementation works on a real speaker prototype. Anybody out there already at that stage, maybe with measurements?
miniDSP, building a DSP community one board at a time.

Please Log in or Create an account to join the conversation.

Re:Linkwitz transform 11 years 9 months ago #944

  • knuisje
  • knuisje's Avatar
  • Offline
  • Senior Member
  • Senior Member
  • Posts: 48
  • Thank you received: 4

Attachment LT_digital_coefs_v1-20100812.zip not found

In this latest version I added to the graph the equalizer as calculated from the digital coefficients. This is usually the same as the targeted equalizer as calculated from the original and target response. But when you fiddle with Fc and Fs you can see it can be different.
I also added a digital filter stabilty calculation. When the poles are within the unit circle the digital filter is stable. Enjoy!
Attachments:

Please Log in or Create an account to join the conversation.

Last edit: by knuisje.

Re:Linkwitz transform 11 years 9 months ago #946

  • devteam
  • devteam's Avatar
  • Away
  • Administrator
  • Administrator
  • Posts: 9049
  • Thank you received: 1376
Very nice... I'm glad to see how you guys are handling this task and doing a very good job at keeping everybody in the loop. Any news on a real implementation? I'm guessing that you're tweaking your system as you're loading new biquad settings into the board?

If you have any results/graphs to showcase on how your LT improved your system, sharing it with us on the Showcase section would definitely be appreciated!

Keep up the good work!
miniDSP, building a DSP community one board at a time.

Please Log in or Create an account to join the conversation.

Re:Linkwitz transform 11 years 9 months ago #952

  • knuisje
  • knuisje's Avatar
  • Offline
  • Senior Member
  • Senior Member
  • Posts: 48
  • Thank you received: 4
Ok, I did some measurements to test the correct implementation. This is a close range measurement of a closed box. I used these parameters in my spreadsheet to calculate the digital coefficients for the biquad LT equalizer:

f(0) = 87
Q(0) = 1.1
f(p) = 50
Q(p) = 0.707

In the graphs you can see the results.

Attachments:

Please Log in or Create an account to join the conversation.

Last edit: by knuisje.

Re:Linkwitz transform 11 years 9 months ago #953

  • knuisje
  • knuisje's Avatar
  • Offline
  • Senior Member
  • Senior Member
  • Posts: 48
  • Thank you received: 4
And the measurements.

Attachments:

Please Log in or Create an account to join the conversation.

Last edit: by knuisje.

Re:Linkwitz transform 11 years 9 months ago #954

  • knuisje
  • knuisje's Avatar
  • Offline
  • Senior Member
  • Senior Member
  • Posts: 48
  • Thank you received: 4
The biquad parameters.
Attachments:

Please Log in or Create an account to join the conversation.

Re:Linkwitz transform 11 years 9 months ago #964

  • devteam
  • devteam's Avatar
  • Away
  • Administrator
  • Administrator
  • Posts: 9049
  • Thank you received: 1376
Great stuff! Thanks for posting your measurements Peter.. Very useful info!

By the way, you guys are featured in the latest newsletter for your great job at documenting custom filters.
Hoping to see others follow your lead with some similar ideas, and keep up the good work!

DevTeam
miniDSP, building a DSP community one board at a time.

Please Log in or Create an account to join the conversation.

Re:Linkwitz transform 11 years 9 months ago #984

  • knuisje
  • knuisje's Avatar
  • Offline
  • Senior Member
  • Senior Member
  • Posts: 48
  • Thank you received: 4
Yes, another update of the LT spreadsheet. In this version I added a pole-zero plot. For more information see Wikipedia .

Attachment LT_digital_coefs_v1-20100821.zip not found

Attachments:

Please Log in or Create an account to join the conversation.

Last edit: by knuisje.

Re:Linkwitz transform 11 years 8 months ago #1062

  • KAMIKAZE
  • KAMIKAZE's Avatar
  • Visitor
  • Visitor
Hello to all, I just purchased minidsp kits and 4-way adv. plugin. Waiting for parcel :)
I already build box for my Peerless XXLS 830846 - 58 liters. Here is measured parameters of it
Fs = 25.43 Hz
Qms = 10.67
Vas = 66.53 liters
Cms = 0.376 mm/N
Mms = 103 g
Rms = 1.56 kg/s
Xmax = 12.5 mm
Xmech = 18.75 mm
P-Dia = 212 mm
Sd = 353 sq.cm
P-Vd = 0.441 liters
Qes = 0.4
Re = 2.6 ohms
Z = 3.12 ohms
BL = 10.23 Tm
Pe = 175 watts
Qts = 0.39
no = 0.257 %
1-W SPL = 86.25 dB
2.83-V SPL = 91.13 dB

Could someone please help me with calculating Linkwitz transform for my box? And give some example. I just not understand all process... :-[

Please Log in or Create an account to join the conversation.

Re:Linkwitz transform 11 years 8 months ago #1063

  • knuisje
  • knuisje's Avatar
  • Offline
  • Senior Member
  • Senior Member
  • Posts: 48
  • Thank you received: 4
Hi Kamikaze,

First you have to determine the f(0) and Q(0) for your box. The spreadsheet in this link can help you, assuming you have a closed box. In the tab "box" you fill in your parameters Fs, Qts, Vas and Vb. The spreadsheet calculates the Qtc [= Q(0)] and Fsc [= f(0)]. With your parameters these are 0.57 and 36.6Hz.
Now determine the desired Q(p) and f(p) and use the spreadsheet in this thread to calculate the biquad coefficients.

Please Log in or Create an account to join the conversation.

Re:Linkwitz transform 11 years 8 months ago #1064

  • KAMIKAZE
  • KAMIKAZE's Avatar
  • Visitor
  • Visitor
Hi knuisje,

Thank you for answer. As I understand, final Q(p) must be 0.5-0.7? and F(p) as lower as better? Can you please give some example with it(with calculated biquad's), maybe I do not really understand this. Thank you in advance.

Please Log in or Create an account to join the conversation.

Last edit: by KAMIKAZE.

Re:Linkwitz transform 11 years 8 months ago #1065

  • knuisje
  • knuisje's Avatar
  • Offline
  • Senior Member
  • Senior Member
  • Posts: 48
  • Thank you received: 4
Kamikaze,

please read the article linked in my previous post on how the Linkwitz transform works.

Please Log in or Create an account to join the conversation.

Re:Linkwitz transform 11 years 8 months ago #1069

  • KAMIKAZE
  • KAMIKAZE's Avatar
  • Visitor
  • Visitor
Ok, I think all done.
Attachments:

Please Log in or Create an account to join the conversation.

Last edit: by KAMIKAZE.

Re:Linkwitz transform 11 years 8 months ago #1070

  • KAMIKAZE
  • KAMIKAZE's Avatar
  • Visitor
  • Visitor
and xls calculations

p.s. can't attach image. here is the link rghost.ru/2624222

Please Log in or Create an account to join the conversation.

Last edit: by KAMIKAZE.

Re:Linkwitz transform 11 years 7 months ago #1118

  • knuisje
  • knuisje's Avatar
  • Offline
  • Senior Member
  • Senior Member
  • Posts: 48
  • Thank you received: 4
I added some tabs for LPF, HPF, BPF and APF calculations based on the cookbook formulas. Also included are the coefficients for a RIAA correction.

Edit: Added are the remaining cookbook filters (BPF1&2, notch, PeakingEQ, lowshelf, highshelf). Also added approximation plots of the phase and group delay for the APF.
Attachments:
The following user(s) said Thank You: CharlieLaub

Please Log in or Create an account to join the conversation.

Last edit: by knuisje.
Moderators: devteam