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

Passive woofer crossover peaking issues
Passive woofer crossover peaking issues
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 28th June 2020, 06:30 AM   #21
wintermute is offline wintermute  Australia
just another
diyAudio Moderator
 
wintermute's Avatar
 
Join Date: Aug 2003
Location: Sydney
Passive woofer crossover peaking issues
Doing a little more experimenting I came up with a slightly more practical resonant peak filter. This is probably doable. (all inductors are erse laminated core 16awg available at PE) The tradeoff for smaller inductor is bigger cap. Won't be cheap though.

I'm curious about why no active, is it philosophical, or maybe a design challenge?

Tony.
Attached Images
File Type: png take_2.png (55.4 KB, 78 views)
__________________
Any intelligence I may appear to have is purely artificial!
Photography (another hobby)
  Reply With Quote
Old 28th June 2020, 05:59 PM   #22
Rallyfinnen is offline Rallyfinnen  Sweden
diyAudio Member
 
Join Date: Jan 2017
I have been working on a similar problem with a ongoing project (second order filter though), and went with impedance flattening LCR of the woofer resonance, and series resistance on the cap to gnd just like in the previous example, that made it a lot better, but it's still peaking a little bit compared to the raw driver response, because I had to 'twist' the phase a little to match with the mid.

For the impedance correction air wound coils with thin wire are not that expensive, in my case the resistance of the coils (needed two) matched the needed resistance. I got Jantzen coils from here: Kamm Lautsprecher Internet-Shop
  Reply With Quote
Old 30th June 2020, 06:41 PM   #23
DBMandrake is offline DBMandrake  Scotland
diyAudio Member
 
DBMandrake's Avatar
 
Join Date: Nov 2010
Location: Glasgow, UK
Quote:
Originally Posted by wintermute View Post
Doing a little more experimenting I came up with a slightly more practical resonant peak filter. This is probably doable. (all inductors are erse laminated core 16awg available at PE) The tradeoff for smaller inductor is bigger cap. Won't be cheap though.
After playing around in Vituixcad with a notch here's what I came up with - similar to yours, which I hadn't seen until later - although mine has a zobel which changes the stop band response significantly and therefore all the other values as well:

Passive woofer crossover peaking issues-visaton-low-pass-notch-png

The notch has a Q of exactly 1.0, centre frequency of 49Hz, and flattens the impedance curve perfectly. The blue line is the filter response, the red line is the response with the peak with the notch disconnected, and the purple line is the target response which you almost can't see because it overlays so closely even below -50dB.

Passive woofer crossover peaking issues-visaton-low-pass-notch-six-pack-png

Minimum impedance is still 5.4 ohms at 22Hz, but rises significantly above that and is above 6.8 ohms above 37Hz)

It's interesting to consider that not only is the filter applying a 4th order low pass response, baffle step correction (rollover point 230Hz) is also built in with no additional components just by retuning the low pass filter from the theoretical.

When I tried to do the same thing with a 2nd order filter I was only able to get a rough approximation not the very close match I get with 4th order. It just goes to show that the more degrees of freedom you get with a higher order filter can be put to good advantage in shaping the response. (I made use of that shaping ability extensively in my last design)

So it's certainly possible to achieve the target response passively but 363uF and 28.9mH are quite daunting and would require a (by-passed) bipolar electrolytic and laminated iron core inductor to achieve in any reasonable space. While I don't mind a laminated core inductor (the series coils for the woofer will be laminated iron anyway) the bipolar cap is more problematic.

I have no idea what kind of current rating and ESR would be typical of a bipolar electrolytic in the 300uF range, and the prospect of increasing ESR and decreasing capacitance with age isn't that appealing.

Interestingly though because the notch is very low Q the capacitor value isn't particularly critical - it can vary all the way from about 300uF to 500uF before the response changes by 1dB. So a small amount of error in the capacitance value over time wouldn't be that critical. And up to an addtional 2 ohms of resistance in the string of cap/coil/resistor can be tolerated before the response goes out of shape by more than 1dB.

With a maximum input of 28.3v (100w referenced to 8 ohms) the notch string would peak at 1.8 amps at resonance at 49Hz. Is that something a ~330uF bipolar cap could cope with or is that pushing it too hard ?

Power dissipation of the resistance in the notch filter string (mainly coil and resistor) with 28.3v input would be a maximum of 35 watts at 49Hz. Feasbile, especially when you consider that 100 watts at that frequency would be a very, very loud noise, and well above any normal sort of playback level.
Quote:
I'm curious about why no active, is it philosophical, or maybe a design challenge?
Active is just not the design I set out to build, it was always intended to be a passive design, right from when the basic design was conceived more than 10 years ago.

It would be like starting to build a set of drawers, running into some small snags with the drawer runners and saying, you know what, I think I'll just turn it into a wardrobe instead...

An active design would certainly be a hell of a lot easier in terms of applying exactly the transfer functions I want to each driver, however the hassle of three amplifiers per speaker, DSP circuitry etc just isn't worth it for me right now, especially on such a large speaker.

Doing the design passive is more of a challenge, perhaps that's part of the appeal.
Attached Images
File Type: png Visaton low pass with notch.png (10.5 KB, 166 views)
File Type: png Visaton low pass with notch Six-pack.png (72.5 KB, 166 views)
__________________
- Simon

Last edited by DBMandrake; 30th June 2020 at 06:46 PM.
  Reply With Quote
Old 30th June 2020, 06:54 PM   #24
DBMandrake is offline DBMandrake  Scotland
diyAudio Member
 
DBMandrake's Avatar
 
Join Date: Nov 2010
Location: Glasgow, UK
Quote:
Originally Posted by Zvu View Post
I'm notching the hell out of those bumps in the bass region.
Ah good, so I'm not the only one who has considered doing this.
Quote:
I'm buying wire and wind my own coils. Copper wire costs 10euros/kg so that's cheap. For testing i've reused burned (that's why the plastic former is a bit deformed) regular EI transformer core. I wound 54mH with 0.6mm wire by hand on transformer core by mistake. I had to unwind three layers to get to 16mH that i needed. Dip the plastic former in laquer and voila. PS i don't have large hands.

Attachment 856073 Attachment 856074
It's looking like I would only need about 28mH for the proposed notch, and a peak current at resonance of about 1.8 amps as in my previous post. Would an EI laminated core of that size be able to carry that sort of current without saturation/distortion ?

Did you perform any high power level distortion measurements of your notch to make sure it wasn't saturating ?

Some cores saturate a lot easier than you'd think - in my previous 2 way (which is only rated at about 15-30 watts) the baffle step coil - about 4mH, which carries all the bass current was originally a reasonably large permite core which was claimed to be suitable for the application, however I was easily able to hear saturation (midrange break through) on certain kinds of bass transients. I swapped it for a laminated I core inductor of a very similar size and the saturation is gone.
Quote:
For caps use two 100V polarized elcos connected back to back like this --|(--)|-- it's a notch filter at 80Hz anyways. Notch such as this is working for few years now in Kef R300's i reworked and with a guy who is not being gentle to them (he listens his music with lots of bass very loud). Works like a charm.
Would a bipolar electrolytic not have been better in this application ? Did you measure the ESR of your configuration ?
__________________
- Simon
  Reply With Quote
Old 30th June 2020, 07:02 PM   #25
DBMandrake is offline DBMandrake  Scotland
diyAudio Member
 
DBMandrake's Avatar
 
Join Date: Nov 2010
Location: Glasgow, UK
Quote:
Originally Posted by AllenB View Post
I seem to have come up with three observations. Firstly, the inductor resistance seems to be a useful tool both in terms of the resonance and the sensitivity.
Are you referring to the series inductors or the one in the notch ?
Quote:
Secondly, in this example I'm using a 10 ohm parallel resistor and the impedance still shows as being above 5 ohms. In fact reducing this resistance will in some circumstances at some frequencies actually increase the impedance as a result of reducing the phase angle of the filter load, if you look close enough.
Yes, this impedance transformer effect operates over quite a wide range of frequencies, from approx 40Hz to over 150Hz in this filter. This is because the 4th order filter is effectively two series tuned resonators with resistive damping (the speaker) connected at the centre point of the second one.

If the speaker impedance is too high (at the driver resonance) this reduces the damping at that point and therefore the current drawn from the source increases. (And the voltage output increases above unity)

A 10 ohm paralel resistor not part of a notch wouldn't be very feasible though as the power dissipation is high right through the entire bass region.
Quote:
The other thing, if I were to do this again is to reduce the target sensitivity to the level at 110Hz, and let the 70Hz peak ride because not only is it a place where such a peak can be tolerated, it is also workable with the room.
Unfortunately 71Hz is already a room mode in my room that needs notching out.

The main room modes in my room are 41Hz, 71Hz and 106Hz. All are between 4-6dB peaks, so I don't want to add another 4 dB at 70-80Hz.
__________________
- Simon
  Reply With Quote
Old 30th June 2020, 07:14 PM   #26
DBMandrake is offline DBMandrake  Scotland
diyAudio Member
 
DBMandrake's Avatar
 
Join Date: Nov 2010
Location: Glasgow, UK
Quote:
Originally Posted by Rallyfinnen View Post
I have been working on a similar problem with a ongoing project (second order filter though), and went with impedance flattening LCR of the woofer resonance, and series resistance on the cap to gnd just like in the previous example, that made it a lot better, but it's still peaking a little bit compared to the raw driver response, because I had to 'twist' the phase a little to match with the mid.
When I tried to do mine as a 2nd order as a test I found the same problem - due to having to "twist" the response to include the baffle step correction I wasn't able to get a close match to the target curve, giving either a peak, or the roll off not being correct. With the 4th order it is possible to get an accurate match to the target curve as you have two more degrees of freedom to adjust.
Quote:
For the impedance correction air wound coils with thin wire are not that expensive, in my case the resistance of the coils (needed two) matched the needed resistance. I got Jantzen coils from here: Kamm Lautsprecher Internet-Shop
You make a good point about the coil resistance in a notch filter where you need added resistance anyway to meat the target curve.

My previous design had two notch filters in the upper midrange and I was able to make the coils a lot smaller than they otherwise would have been by exploiting the fact that I needed several ohms resistance anyway. So the coils provided a couple of ohms of the total required R and were much smaller than a similar coil with say 0.3 ohms resistance.

This can be done to some degree for this bass notch, however I'd have to be careful that the coil's R is not too high, otherwise its power dissipation could become quite high given the low Q (therefore wide bandwidth) and high power levels at bass frequencies compared to my previous upper midrange notches where the power levels are low.

So it would be a matter of balancing the total power disipation between the resistor and coil in a sensible way given their relative thermal ratings.
__________________
- Simon

Last edited by DBMandrake; 30th June 2020 at 07:19 PM.
  Reply With Quote
Old 1st July 2020, 04:21 AM   #27
AllenB is offline AllenB  Australia
diyAudio Moderator
 
AllenB's Avatar
 
Join Date: Oct 2008
Passive woofer crossover peaking issues
Quote:
Originally Posted by DBMandrake View Post
Are you referring to the series inductors or the one in the notch ?
Series.
Quote:
Unfortunately 71Hz is already a room mode in my room that needs notching out.
I don't usually concern about such coincidences where you plan to deal with each issue anyway.

So about that, each issue is likely to respond to minimum phase filtration. Maybe even at the same time should the circumstances arise. Is global EQ out of the question? Of course speaker level will give you the lowest inductor values, low frequency line level inductors could be found in power transformers etc, if you can measure their self resonance. Or use a non inductor style of EQ.

There are alternatives, box alignment is one.

EDIT: I see that you have found a way.
  Reply With Quote
Old 1st July 2020, 12:11 PM   #28
Zvu is offline Zvu  Serbia
diyAudio Member
 
Zvu's Avatar
 
Join Date: Jan 2009
Location: Belgrade
Quote:
Originally Posted by DBMandrake View Post

Would an EI laminated core of that size be able to carry that sort of current without saturation/distortion ?

Did you perform any high power level distortion measurements of your notch to make sure it wasn't saturating ?
I've measured distortion with few types of core. Here are the results:

Influence of inductor types on distortion performance

These EI cores tested aren't the ones i winded but i've bought them locally. They failed the distortion test miserably so i won't be buying them anymore. Bad alloy i guess. It would be nice to know what type of alloy is used by Mundorf in their EI coils.

Bottom line, getting the right core could be a problem. I would advise to use ERSE i-core inductors (they're cheap) and use your own wire. Kef R300 use ferrite coils in their notch (22mH) and in series with woofer (2.2mH). Their cores behave well in distortion tests.

Quote:
Originally Posted by DBMandrake View Post

Some cores saturate a lot easier than you'd think - in my previous 2 way (which is only rated at about 15-30 watts) the baffle step coil - about 4mH, which carries all the bass current was originally a reasonably large permite core which was claimed to be suitable for the application, however I was easily able to hear saturation (midrange break through) on certain kinds of bass transients. I swapped it for a laminated I core inductor of a very similar size and the saturation is gone.
I'd use Dayton or ERSE I core for that. I don't know what exactly causes the distortion in core material but here are few tests:

YouTube

YouTube


Quote:
Originally Posted by DBMandrake View Post
....Would a bipolar electrolytic not have been better in this application ? Did you measure the ESR of your configuration ?
ESR and ESL of capacitor in series RLC is practically irrelevant. You are adding 28mH and few ohms in series with that capacitor.
__________________
Tesla; George Carlin

Last edited by Zvu; 1st July 2020 at 12:16 PM.
  Reply With Quote
Old 1st July 2020, 12:22 PM   #29
chris661 is online now chris661  United Kingdom
diyAudio Member
 
chris661's Avatar
 
Join Date: Feb 2008
Location: Sheffield
Quote:
Originally Posted by DBMandrake View Post
Unfortunately 71Hz is already a room mode in my room that needs notching out.

The main room modes in my room are 41Hz, 71Hz and 106Hz. All are between 4-6dB peaks, so I don't want to add another 4 dB at 70-80Hz.
How're you notching those? Could the impedance-related peak be treated similarly?

IIRC, in a couple of his bigger 3-way designs Troels implemented a plate amp for the LF driver, and set it up to accept speaker-level inputs. ie, still usable with a standard 2-ch HiFi amp.

Apart from doing that, it looks to me like a big LCR network is going to be the only way to fix this one. You could use an air-cored inductor and build most of the series R into it. More wire, though.

Chris
__________________
Bits of information and other things I've worked on: https://www.grimshawaudio.com/lab
  Reply With Quote
Old 1st July 2020, 02:04 PM   #30
wintermute is offline wintermute  Australia
just another
diyAudio Moderator
 
wintermute's Avatar
 
Join Date: Aug 2003
Location: Sydney
Passive woofer crossover peaking issues
OK understand on the why, but it if gets too much, you could go hybrid (it's what I did) I used (analogue) active for the woofer to mid crossover, and passive on mid to tweeter. Then you only need four channels of amplification rather than six...

The way my active crossover works, you design as if you were doing passive. I really should do a more accessible version of it using opamps for the buffers rather than B1's. I think one channel uses 8 jfets...

Tony.
__________________
Any intelligence I may appear to have is purely artificial!
Photography (another hobby)
 

Reply


Passive woofer crossover peaking issuesHide 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
Passive woofer crossover redux... aceinc Multi-Way 6 19th July 2018 03:16 AM
Khozmo Passive Pre Remote Issues tlarwa Analog Line Level 6 2nd December 2015 01:48 AM
About the woofer:Passive Radiator or Double Woofer for Vetend kevinlin1013 Multi-Way 8 12th January 2011 06:13 PM
corrections for LF and HF peaking pdcaceo Tubes / Valves 4 14th November 2010 12:58 PM
passive woofer crossover. gvr4ever Full Range 3 15th May 2009 01:02 AM


New To Site? Need Help?

All times are GMT. The time now is 07:10 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2020 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2020 DragonByte Technologies Ltd.
Copyright ©1999-2020 diyAudio
Wiki