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

S. Harsch XO
S. Harsch XO
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 17th November 2019, 09:33 PM   #661
soundchaser is offline soundchaser  Canada
diyAudio Member
 
soundchaser's Avatar
 
Join Date: Oct 2007
Location: Kanata
Default I must be daft or somthing

Hi All;

Great work here with some really great insights and results. I am only doing simulations right now but am planning a build for the near future (Dec/Jan). It's all DSP so I can do whatever I want with biquads and canned crossovers and filters in Audio Weaver. Don't have enough horsepower for FIR arbitrary phase/amplitude filters unless I play with sample rates which I may do.

I will play with filler driver designs (which I have also learned about in this thread;THANKS to all) as well as conventional LR4 and BW4 etc designs. It's just a file and an upload once the drivers are measured.

I am simulating biquads in SPICE and I can make the Butterworth/Bessel slopes but I cannot get the phase to make sense. The coefficients are from the Mini DSP spreadsheet. I am sure that they are correct. I have simulated a 2 way with a crossover at 375Hz (a FAST like xrk's) and I have the 1.3333ms delay in the tweeter path (1/375/2). Note that there are no drivers. The free space summation is an ideal summing node so driver response/displacement is not a concern (but I still can't get it to look like a Harsch response )

Intuitively what the simulation says makes sense. A constant delay (1.333ms) will result in huge phase wraps at higher frequencies. I simply cannot get the little phase ripple that you guys can make. 1.333ms at 1kHz is 480, at 5kHz it's 2400.

Am I doing something goofy or does what I am simulating make sense and I am just interpreting it incorrectly.

Thanks for any insight.

Matt
Attached Images
File Type: jpg 2way_375_harsch.JPG (199.2 KB, 110 views)
File Type: jpg 2way_harsch_amplitude.JPG (146.0 KB, 111 views)
File Type: jpg summed_response.JPG (188.8 KB, 106 views)

Last edited by soundchaser; 17th November 2019 at 09:35 PM.
  Reply With Quote
Old 17th November 2019, 11:10 PM   #662
BYRTT is offline BYRTT  Denmark
diyAudio Member
 
BYRTT's Avatar
 
Join Date: Dec 2009
S. Harsch XO
Think looks you probably close enough, filter is a compromise to transient perfect behavour summing two non symetrical out of phase minimum phase filter slopes plus add a physical delay to HF part, on axis sum is not as perfect as for example LR or 1st order slopes, for HARSCH filter sum have a cost in some more or less amplitude ripple.

Error looks you view that summing ripple on a 1,5dB vertical scale and there it looks bit scary, i setup your project specs into REW and if you look below we can see vertical scales mean a lot : )
Attached Images
File Type: png soundchaser_1.png (18.5 KB, 90 views)
File Type: png soundchaser_2.png (24.3 KB, 90 views)
File Type: png soundchaser_3.png (18.8 KB, 24 views)
  Reply With Quote
Old 18th November 2019, 07:09 AM   #663
phase_accurate is offline phase_accurate
diyAudio Member
 
Join Date: May 2002
Location: Switzerland
Hi Matt

It looks as if you were able to program your DSP quite freely. Why don't you try subtractive delay in this case ?

Regards

Charles
  Reply With Quote
Old 18th November 2019, 07:30 PM   #664
soundchaser is offline soundchaser  Canada
diyAudio Member
 
soundchaser's Avatar
 
Join Date: Oct 2007
Location: Kanata
Thanks Guys;

Brytt; I am not concerned about the amplitude ripple. I suspect that this gets lost in the noise of actual driver FR variations. Who knows; it might even help. It's the huge phase slope that I can't reconcile with the results presented here (small phase bump up to 60-90). Again these are SPICE simulations. There are no drivers, offsets or lobing issues.

Charles; Thanks; will check what is available but this is not DSP that I have presented; it is a time domain based biquad simulated in SPICE (Micro Cap). Like the MiniDSP SW though I have many building blocks available including biquads that behave like the MiniDSP types. I do not have a subtractive delay however. There is a block for a two-way subtractive type crossover that i can play with.

Thanks
Matt
  Reply With Quote
Old 18th November 2019, 09:01 PM   #665
BYRTT is offline BYRTT  Denmark
diyAudio Member
 
BYRTT's Avatar
 
Join Date: Dec 2009
S. Harsch XO
Okay phase was culprit subject : )

In REW i get the same phase wrap when i sum the two IR (impulse responses) with HF bandpass delayed 1,333mS because IR for the new summed bandpass continue sit 1,333mS delayed on timeline in IR window, but then one can re allign IR to time zero and phase then change to the characteristic HARSCH filter hump. Other method in REW that output the correct phase without one re alligning IR is instead of a delay to IR for HF band then set IR for LF band 1,333mS advanced in time, so wonder if you can do the same in remove the delay on HF band and try with for example a negative sign -1,333mS set for woofer, else try put a normal delay on both bands in front of their filter sections and see what happens, for example a delay of 1,333mS for LF band and then 2,666mS for HF band.

Last edited by BYRTT; 18th November 2019 at 09:04 PM.
  Reply With Quote
Old 18th November 2019, 10:05 PM   #666
soundchaser is offline soundchaser  Canada
diyAudio Member
 
soundchaser's Avatar
 
Join Date: Oct 2007
Location: Kanata
Thanks Byrtt; I will give those ideas a try tonight after my chores .

I apologize; I have been mis-spelling your name.

Best
Matt
  Reply With Quote

Reply


S. Harsch XOHide 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
convert passive XO speaker to active XO halibutman Multi-Way 10 16th September 2019 06:26 AM
FS: Tent Labs XO 3.2, XO Supply, XO Module [New In Box] $135/all lupin..the..3rd Swap Meet 2 17th April 2015 07:10 PM
50Mhz XO vs. 49.152Mhz XO? Mull3t Digital Line Level 3 31st May 2013 02:07 AM
Digital XO vs Op-amp XO, which would be best to use in active speakers? MikeHunt79 Analog Line Level 11 16th November 2009 06:06 PM
1st Order XO with -6dB XO frequency. Has anyone tried this? primalsea Multi-Way 20 28th May 2004 07:10 AM


New To Site? Need Help?

All times are GMT. The time now is 09:55 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio
Wiki