difraction compensation - 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 2nd January 2002, 11:42 AM   #1
navin is online now navin  India
diyAudio Member
 
navin's Avatar
 
Join Date: Jan 2002
Location: Mumbai (Bombay), India
Send a message via MSN to navin Send a message via Yahoo to navin
Default difraction compensation

I am planning to build 2 speakers that are thin tall deep towers looking physically like NHt's 3.3 or 2.5.

The Side facing woofer is a 12" Audio Concepts DV12. The Front drivers are a 6" 18W8546 ScanSpeak and a 1" 9900 ScanSpeak tweeter. Before I start I must warn that my post is long and since I do not have regular email access I would appreciate it if I am emailed at navin@vsnl.com.

I have the following requirements:
1. As high efficiency as possible
2. Use for Home Theater as well as Stereo
3. Resonable Power handling (100-150W rms)

I hope to use the 12" to compensate for the difraction loss of the 18W8546. Difraction loss is a loss (claimed to be 6db) of low frequencies based on the width of the cabinet (in this case 8" or 20cm). Hence to compensate for Difraction loss most 6" 2 way spekaers employ either a 2nd 6" covering only the lower frequencies or reduce the
upper frequencies to make the speaker sound better balanced.

Not having a 2nd 6" I hope to use the output of the 12" but dont know if this make sense as the 12" will be quite some distance from the 1st 6" and the frequencies that are compensated for are from 100-1800Hz. The 12" might not be able to reproduce 1800 Hz accurately and in any case 2 drivers reprducing the midrange but being far apart might create problems (lobing, delay, etc.). Any advise would be welcome.

Also to increase power handling it makes sense to roll of the low frequencies to the 6". Power handling for this driver is not limited by it's thermal limits but by it's physical excursion limits. Hence the same power being supplied
to the 12" and 6" operating in parallel at low frequencies (below 100Hz) can stress the 6" while not even exciting the 12". However to employ a low frequency high pass crossover using passive crossovers is a challenge. Again I welcome any advise on this front.

Lastly given the physical constructon of a 8" wide, 44" high, 30" deep tower will be difraction loss be 4db since the only edges are 2 sides (4" from center of woofer) and the top (10" from center of woofer) or more? I assume this
because it commonly held that difraction compensation for a 6" 2 way is about 6db-8db. However 6" 2way monitors have 3 edges close the the woofer (4" from center of woofer) and a 4th edge about 10" from center of woofer.

So there you have it, can one use a side facing 12" for difraction compensation?, Can one increase the power handling of a 6" using passive crossovers?, and what it the difraction compensation for a 6" on a thin, deep, tower cabinet?.

Regards
Navin
  Reply With Quote
Old 2nd January 2002, 07:54 PM   #2
Wizard of Kelts
diyAudio Moderator
 
Join Date: Sep 2001
Location: Connecticut, The Nutmeg State
Even side-facing, the 12" will have output well above 100 Hz coming to the front of a speaker.

If you are worried about diffraction loss, try crossing the 12" half an octave above 100 Hz and see if that fills it in.

There are a lot of 8" wide speaker cabinets on the market and most of them go below 100 Hz. Are you certain this phenomenon is going to be a problem?

The best thing to do is take some frequency measurements in the room. The web has free programs for that, if you are interested. An inexpensive Radio Shack sound level meter and a free program is probably all you need. With some programs, yo might be able to substitute an inexpensive microphone for the Sound Level Meter.
  Reply With Quote
Old 3rd January 2002, 07:47 AM   #3
navin is online now navin  India
diyAudio Member
 
navin's Avatar
 
Join Date: Jan 2002
Location: Mumbai (Bombay), India
Send a message via MSN to navin Send a message via Yahoo to navin
Thanks for your reply.

What I am worried about are:

1. How much difraction do I need for a tower cabinet assume I round edges. I think I will still need 6db as the driver will still be radiating into half space. Am I right?

2. Assuming I need 6db difraction compensation from about 100Hz to about 1500Hz the 12" woofer will have to produce frequencies upto 1500hz which might pose 2 problems: a) the 12" woofer might not produce these frequencies (above 300Hz) accurately and b) with 2 drivers producing teh same midrange frequencies quite fasr apaart I might run in to problems of imaging, lobing etc... Right?

3. The 6" woofer (like any small woofer) is lim9ited as to how much bass it can produce. Hence it makes sense to roll it off at the low frequencie (below 100 Hz). However designing a accurate PASSIVE HIGH PASS crossover at 100Hz is difficult and expensive as big capacitors are needed.

If you have any suggestions let me know.

The speaker i am hoping ot build will be about 8" wide, 44" tall and 28" deep.

Regards
navin
  Reply With Quote
Old 3rd January 2002, 10:03 AM   #4
diyAudio Member
 
Join Date: Dec 2001
Location: Milano, Italy
Hi Navin,

I had analyzed a bit the NHT 3.3 project, and maybe my considerations can help you in the design of your 'clone'

The 3.3 are carefully optimized for a strongly defined project philosophy: obtain a full range system with optimal image and soundstage recreation. The system deals with room interactions by defining a 'controlled' placement of the speakers in the room and of the uniots in the cabinet.

Bass section
NHT reccommends to place the system close to the room's rear wall, and at some distance from the room's side walls.

The whole design of the speaker is optimized for this placement:
1 - the side placed 12" works in about 70 lt sealed box, and is placed as close as possible to the floor and to the rear of the speaker.
This means that the 12" alignement is designed to work using the reinforcement of the 2 walls near him
I don't have the unit TS parameters, but probably the way chosen by Ken Kantor (who is surely not a newcomer in design of sealed box systems...) is to obtain a very low Qtc (0.6 ?) for fast transient response and control, and to 'compensate' the roll-off at low frequencies with the contribution of floor and wall.

2 - This unit is more a built-in sub than a true woofer. The x-over freq is around 100 Hz. This means that the unit will not substantially participate to the recreation of the ambience and soundstage.
To filter with a passive x-over, in my experience this is the best way: a) use a woofer with an early 'natural' roll-off. Units designed for sub use (high mass, low Q) often shows this behavior. b) compensate the impedance peak with a RLC net; a sealed box with low Q is not difficult to compensate, also if you use a big qty of filler the imp. peak will be already dumped. c) stay, if you can, on a 1st order x-over cell.

Mid-High section

The 3.3 has a sort of 3-way box, with a 6.5" driver in about 10 lt, a 4" (probably a custom version of the Seas 11 unit) in about 3 lt and a dome tweeter (still Seas, I guess).
The cabinet design brings the mid-high section 'far enough' from the room rear wall, resolving the possible interaction problems with the wall itself. This, added to the minimized front section, makes me guess that KK tried to obtain a kind of 'free standing' optimization for the mid high section (remember the good results achieved in ambience, soundstage and 'air' by projects like the ProAC tablette?)

The problem for a replica is that this section is optimized for a well defined dispersion behavior, helped also by the inclination of the front panel and by the block of felt placed in the outer side of the panel itself.
If you want to replicate the project (and not only the look of it) it takes to know something more about the original design, choose the appropriate drivers and design the x-over accordingly.
You also need a good simulation software, unfortunately it is almost impossible to design such a complex x-over only applying the book formulas: if you succeed to obtain good results you are either very lucky or a real wizard!

Not knowing the 3.3 project details, I would stay with a 2 way system with a 6" and a 1" dome driver, trying to design it as if it is a stand alone mini-speaker. I personally like very much the Seas P17 RCY for this job (or his son, if the driver is not anymore in production). If you're afraid by the 6" dispersion in the x-over with the tweeter try a 5", trading some bass for upper mid accuracy.

I hope I have not confused you, anyway let us know if you need some help in calculations or other aspects: there is plenty of very experienced people in this list, and NHT 3.3 are well worth the work for a replica!

bye
sandro
  Reply With Quote
Old 3rd January 2002, 11:09 AM   #5
navin is online now navin  India
diyAudio Member
 
navin's Avatar
 
Join Date: Jan 2002
Location: Mumbai (Bombay), India
Send a message via MSN to navin Send a message via Yahoo to navin
Thanks. I am sorry I misled you by using NHT 3.3 as an example. My speaker will inly look physically like the 3.3.

It will effectively be a 2 way (6" woofer and 1" tweeter) with a side firing 12" subwoofer.

What I hope to do is the use the higher midrange output of the subwoofer (above 100Hz till 1000Hz or so) to compensate for the difraction loss of the 6" on the 8" baffle. This will allow for a more efficient speaker.

Also I hope to roll of the 6" below 80 or 60Hz so that the speaker's capacity is not limited by the excursion limitation of the 6". One way to do this is to use a 500uf cap in series with the 6". This can be done by using 2 1000uf caps in series bypassed with a 10uf polyester cap or 2uf Wonder cap/Multi Cap/ or other exotic cap.

Regards
Navin
  Reply With Quote
Old 3rd January 2002, 11:54 AM   #6
Alex M is offline Alex M  United Kingdom
diyAudio Member
 
Join Date: Apr 2001
Location: Hampshire, UK
Hi Navin,

I don't understand your estimate of the cutoff
frequency for your diffraction compensation. For
a 20cm baffle, John Murphy's formula (see his
article at http://www.trueaudio.com/st_diff1.htm )
gives a centre frequency for the step at 575Hz,
which is a lot less than your figure of 1800Hz,
and which is certainly possible with a 12" driver.
At 1800Hz a driver this size is anyway going to be
beaming quite severely and a side-mounted woofer
is pointed in the wrong direction.

My feeling is that you would be better off using
the mid-bass driver to do the baffle step
compensation, and to leave the subwoofer to do
what it does best - in other words the low bass.

Alex
  Reply With Quote
Old 3rd January 2002, 12:20 PM   #7
Warp Engineer
On Holiday
 
AudioFreak's Avatar
 
Join Date: Oct 2001
Location: Queensland, Australia
ok here goes ... 6dB is the theoretical value.... it will be less than this in real life.... how much less? it depends on your drivers, any stuffing in the enclosure, ported or sealed, what material the enclosure is made of and how thick it is.... best is just to listen and see where it sounds best. The centre frequency for the baffle step will be around 575Hz but it starts rolling off @ around 9KHz and finishes @ around 80Hz this is usually corrected for in the crossover.
  Reply With Quote
Old 3rd January 2002, 12:43 PM   #8
diyAudio Member
 
Join Date: Dec 2001
Location: Milano, Italy
Hi Navin,

my fault I had not read carefully your mail, I try to be more on topic this time.
I have run a quick simulation with the Scan 8546, and at first sight it seems that you should not be so concerned about it.

In a 10 lt box (filled) and an additional resistence of 1 ohm (the filter's coil that you must use anyway) you have a Qtc of 0.5 and a F-3 of about 90 hz, both very good for bass-midrange use.
Also with 200W he reaches 6.5 mm exc. at 120 Hz and 10mm at 80: given the normal energy distribution of music you are on the safe side.

I need to go back to work now, but later I will try a simulation of your drivers on a panel with dimension and shape that you want to use, my feeling is that you don't need to play tricks with the woofer to compensate the Scan. btw, if you can post the tS parameters and the freq. and impedance curve of the woofer I can try a simulation of the complete system.

Sorry, I really have to run
bye
sandro
  Reply With Quote
Old 4th January 2002, 08:06 AM   #9
navin is online now navin  India
diyAudio Member
 
navin's Avatar
 
Join Date: Jan 2002
Location: Mumbai (Bombay), India
Send a message via MSN to navin Send a message via Yahoo to navin
Thanks or all these replies.

As was said eariler the difraction loss starts at quite high a frequency 9k by one estimate) and continues to 80Hz.

I am not to worried about the 2 extremes as they are easily compensated with output from the tweeter (2k-9k) and sub (80-200Hz). It is the range between 200Hz and 2kHz that needs to be addressed.

This loss is about 4db from what I understand. To put some more clairity to this my speakers will be 44" high, 8" wide and 28" deep and will sit as close to wall behind them as physically possible.

The speakers will be 3 feet from the side walls and about 7 feet between them. The room is 14.5 feet wide.

So now from what I understand is the sub can operate till about 200Hz. The 6" will operate from 100Hz to about 2500Hz and the tweeter from 2500Hz on up.

The sub (90db 4 ohms nominal 3 ohms DC) can easliy be rolled of using an inductor. The 6" (88db 8ohms nominal 5.5 ohms DC) will need a large cap for high pass and a crossover using some difraction compensation. The 1" (91db 8 ohms nominal 7.1 ohms DC) will have a crossover as well as some damping circuit to normalise the 500Hz impedance peak.

Thanks again. Any more advise is well come.

Regards
navin
  Reply With Quote
Old 4th January 2002, 10:49 AM   #10
navin is online now navin  India
diyAudio Member
 
navin's Avatar
 
Join Date: Jan 2002
Location: Mumbai (Bombay), India
Send a message via MSN to navin Send a message via Yahoo to navin
Default difraction

Thanks for teh help.

What I think I will need to do is to build the box and then install the drivers.

The 12" can initally be full range with any crossover just to see how much response I get and how it's response affects the
midrange (imaging, etc...). It can be rolled of later.

I downloaded LSPCAD. Has anyone used this? Is it accurate for basic simulation?

Using LSPCAD I built a very simple crossover which by no mena is a final crossover. It is just so that one can understand what
the pros and cons of crossing over at different points is and how the upper response of the 12" affects the senstivity of the
system.

I ran the 12" with a series inductor L1.
The 6" with a 500uf cap to stop low frequencies and a LC circuit for the low pass crossover the LC circuit is called L2 C2.
The 1" was first damped with a series resistor to reduce it's level and then had a CL circuit using a 6.8uf cap and 1 mh inductor.
Finally I added a paralell 10 ohm ressitor with the tweeter to damp the 500Hz impedance peak.

Hence having only 4 variables L1, L2, C2, and R1 one can understand the behaviour of the drivers better.

I estimated various responses of teh 12" by making L1 2.5mh, 3.5mH, 4.5mh and 7mh. The increase L1 simulated a quicker
rollof of the upper frequencies of the 12".

These are my results.

L1=2.5mH, L2= 1.1mH, C2=4.7uf R1=2ohms db=87.5db
L1=3.5mh, L2=1.2mH, C2=6.8uf R1=2.5ohms db=87db
L1=4.5mh, L2=1.4mh, C2=6.8uf R1=3ohms db=86db
L1=7mH, L2=1.7mH, C2=6.8uf R1=4ohms db=85db

I hope I am working the right direction.

Regards
  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
Baffle step difraction JC951t Multi-Way 7 18th March 2009 12:09 PM
Gunderson compensation.. mikeks Solid State 155 25th September 2006 02:14 AM
Baffle Difraction & ralph-bway Multi-Way 36 9th January 2005 01:26 PM
Reduce difraction and distortion from the cabinet on DIY designs. Bedroc Multi-Way 1 24th February 2002 03:09 AM


New To Site? Need Help?

All times are GMT. The time now is 08:51 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