Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Multi-Way Conventional loudspeakers with crossovers

DIY 2-way design for a bookshelf
DIY 2-way design for a bookshelf
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
Thread Tools Search this Thread
Old 22nd January 2019, 09:15 PM   #21
zermak is offline zermak  Italy
diyAudio Member
Join Date: Mar 2017
Sadly I don't have the tools to simulate the crossover using the baffle and bass reflex/in cabinet responses (I am using VituixCAD right now); I am stuck with the infinite SPL provided by SB Acoustics so I can't actually give you any feedback about your changes. And if I amy ask what software are you using?

About the tweeter SPL, I am pretty sure my dad prefers a more present high end so I guess a 0,75 or 1 Ohm resistor will be enough but then this is going to be a subjective matter and like you said I can tune it with a combination of resistors later when the cabinet is done.
  Reply With Quote
Old 23rd January 2019, 01:46 AM   #22
zermak is offline zermak  Italy
diyAudio Member
Join Date: Mar 2017
I was able to run a simulation using the Boxsim software (which I am still learning to use properly and I can't properly design your front baffle design, I am stuck at 45 degree angles for the edges and I used a 2x2 cm value for all the sides).
After setting a few things up I was able to run the optimization script on the XO network and this is what it outputted using the available items from the store I will buy from:
Click the image to open in full size.

How does it look on your simulator?
Considering that the cost of the crossover is around 95 euros for each speakr I better pick up the best solution here and I am also curious to compare the results.
  Reply With Quote
Old 23rd January 2019, 07:38 PM   #23
LineSource is offline LineSource  United States
diyAudio Member
LineSource's Avatar
Join Date: Sep 2003
Location: SiliconValley
A crossover optimizer which only uses the driver impedance cannot adjust for irregular driver SPL effects like cone break-up, rising SPL at high frequency from beaming, baffle edge effects, and baffle-step effects from the low frequency transition from 2Pi -> 4Pi room radiation.

1) Start construction of the cabinets now
---do not route baffle speaker holes until you have actual drivers
2) Continue education in simulation
--- Use FPGraphTracer to import into Vituix the on-baffle SPL measurement data from SBAcoustics 61-NAC as an FRD file
3) Continue to request support on diyAudio website until you feel confident in $$ purchase
--- will not affect schedule until cabinets are completed

P.S. Your optimizer crossover circuit shows:
1) +SPL bump from 200-2000Hz created by the baffle effects
2) no low frequency baffle step compensation
3) does not adquately attenuate the metal cone break-up effects
4) the midbass SPL does not closely follow the ideal LR4 curve because it does not factor in the imperfect rising high frequency SPL from the driver

P.S. You still need to decide if you want a crossover optimized for against the wall placement, or a more common circuit which includes modest bass baffle step compensation.
  Reply With Quote
Old 23rd January 2019, 09:00 PM   #24
zermak is offline zermak  Italy
diyAudio Member
Join Date: Mar 2017
Thanks for the feedback and as you can see I am still learning about a lot of stuff.
There is a but tho. I am already using the frequencies responses of the drivers; I've traced them with VituixCAD (and I guess SPLs are on an "infinite baffle"), as you suggested, and I've been using them to run the simulations since your first crossover option.

I am even using LspCAD 6 now with the advanced options on (meaning 19L bass reflex cabinet with your port's sizes and diffractions of the speakers and bass reflex port on and placing them based on your design) and the SPL response curve is kinda weird; it goes linearly up ending with 20kHz over 10db the average (and in this case even if weird in its way the crossover from BoxSim has the least "wobbling" response making the "/"-shaped curve the most linear).

On the could list:
1) That is the easy part and I will as soon as the weather gets a little warmer (I can't use the garage right now, it's packed ahah) EDIT: what about the inside bracing? Do you have any design to show? Can there be none and use only damping material? Or it is important for cabinet solidity?;
2) I'll keep experimenting indeed;
3) here I am and so far I am only sure on the drivers to buy; these SB Acoustics.

Abotu the PS list.
1) which is the opposite that I get running your crossovers on my simulations but again in many I can't add bass reflex and diffaction responses.
2) is that important? I mean is the size of the baffle this wide to have bumps in the low frequencies? I've read the paper and it seems the higher frequencies are more subject to this, because of their nature (higher frequencies, closer the audio beams).
3) I guess this comes from experience.
4) I don't know about it because even if it doesn't follow the ideal LR4 curve it doesn't mean it's bad for this driver but again I am just talking looking at simulated datas and I am pretty sure you have more experience than me.

And least but not least.
As I said the speakers will be close to the wall; let's say 25 cm away from the back wall and the left one will be close to the side wall (25 cm, maybe less honestly) plus there is a furniture at 94 cm from the top of TV furniture on where the speakers will be placed.
So a crossover optimized for against the wall placement is the only option and it makes my simulations even more harder.

PS: I own a Umik-1 and I will gladly use it to measure the speakers and make XO adjustments but considering the costs of the crossover itself the less I will spend after for the tuning, the better.

Last edited by zermak; 23rd January 2019 at 09:07 PM.
  Reply With Quote
Old 23rd January 2019, 09:51 PM   #25
zermak is offline zermak  Italy
diyAudio Member
Join Date: Mar 2017
Another post, sorry.
And so what if I make the cabinets, buy the drivers and test them into the cabinet and then build the crossover using the measurentes I will take with the Umik-1 with the speakers in their listening position? Seems the best and smartest move here, right?
  Reply With Quote
Old 24th January 2019, 05:20 PM   #26
Qts is offline Qts
diyAudio Member
Join Date: May 2016
With Umik-1 you can't use second channel for timing reference. Without accurate timing, acoustic centers of drivers and crossover simulation are inaccurate.

I bought Umik first and then Focusrite Scarlett 2i2 2nd Gen and Sonarworks xref 20.

Last edited by Qts; 24th January 2019 at 05:26 PM.
  Reply With Quote
Old 24th January 2019, 05:38 PM   #27
zermak is offline zermak  Italy
diyAudio Member
Join Date: Mar 2017
Well even if it is cool to play with such stuff and monkeys are everywhere (indeed I have bought the Umik-1 to measured my box design for the Markaudio CHR70 and tune it with Equalizer APO since I use them on my desktop PC) I prefer not to spend any more money on gear to test loudspeakers.

My idea is to singularly measure the drivers response (with the mic set on tweeter center height at maybe 1 meter, still wondering on this) but with them mounted in the cabinet and the cabinet placed in the listening position.
Can't I use the plots from REW or any other analyzing audio software to find out how to time align them properly?

PS: I just ordered the drivers and the materials for the cabinet; I'll buy XO materials after I finish the build.

Last edited by zermak; 24th January 2019 at 05:41 PM.
  Reply With Quote
Old 25th January 2019, 12:58 AM   #28
pliedtka is online now pliedtka  Canada
diyAudio Member
pliedtka's Avatar
Join Date: Jul 2008
Location: Ottawa
You can make the cabinet so the front baffle is removable to try different driver placement - often moving tweeter off center helps smoothing out peaks and dips and you don't have to rely only on crossover to smooth out the freq response. For most 4ohm 6" SB, the first coil in x-o is around 1mH, I would get a bigger value and unwind just in case, or make one with two values, say 1 and 1.2mH to have some flexibility in tonal adjustment. I use around 1.6mH for free standing monito with 17NRX-4, 50cm from back and 1.5m from side walls. Of course you will have to attenuate tweeter accordingly, with SBs less than 0.5dB here and there makes quite a difference as they are very transparent speakers. Also don't take as final 100% what the simulator tells you, experiment a bit around the values - use your ears as final judge, I was very surprised that being a bit off produced more satisfying sound.
  Reply With Quote
Old 25th January 2019, 06:10 AM   #29
Qts is offline Qts
diyAudio Member
Join Date: May 2016
Wavequide, for example Monacor WG-300, is one way to move tweeter acoustic center back (about same amount than Gravesen stepped baffle) and reduce diffraction significantly.
  Reply With Quote
Old 25th January 2019, 05:40 PM   #30
zermak is offline zermak  Italy
diyAudio Member
Join Date: Mar 2017
I think I am gonna try the Gravesen front design from his SBA61-NAC build first to address the polar center of the speaker (the loudspeakers will be placed on the TV furniture making the tweeter face a very low point, and maybe I'll just angle the front baffle of the speaker properly to address the issues even better, I have stil some time to virtually experiment) but using a bigger volume cabinet if it's worth it (SB Acoustics suggests a 17L vented box with port tuned at 35Hz amd using the Illuminator dimensions and with everything ion place I guess I am there with the net liters).

About the crossover, thank you for the tip and so far every simulations run never suggested a value over 1,5mH if I recall so we agree on that. By the way this midwoofer is a bad boy and have a very difficult SPL to address but maybe I found the trick, looking at Gravesen XO network: use a resistor (between plus and minus, sorry I don't know the proper term, maybe short?) after the first inductance to attenuate the high frequencies even further.

Thank you for the suggestion but I guess I am going to stick with the front baffle design/experimentation to address these issues. I already have a decent starting point which is, as said before, Gravesen's design.
  Reply With Quote


DIY 2-way design for a bookshelfHide 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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
looking for Morel bookshelf design msh1999 Multi-Way 11 8th March 2017 06:06 PM
Bookshelf Design for FE126 pcgab Full Range 11 1st November 2015 11:46 PM
need a crossover design for bookshelf arunpanchami Multi-Way 4 6th February 2014 04:50 AM
Bookshelf TL design kctess5 Full Range 7 6th February 2012 07:27 PM

New To Site? Need Help?

All times are GMT. The time now is 05:18 AM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio