S15: Econowave DSP - a Constant Directivity vs Dipole study

[gainphile]

I had been intrigued by constant directivity loudspeaker topology since reading rave reviews of Geddes speakers (gedlee.com). The cost of building one, however prevented me to have a go. Soon I found out that Zilch and his mates at audiokarma had been measuring impressive commercial waveguides available to the public. This had opened new possibilities.

I have since incorporated the waveguides to my dipoles with somewhat success. Then I also read a rather disturbing find when a group of listeners preferred a Behringer over the Orions which I look up to. The Behringers are in the same school of thought with CD design. Lacking tools, I pulled out my fingers and built these.

Unpainted and unglued at the moment. But sealed well enough.

Design
The design is loosely based on Standard Econowave.

External dimensions
36.8cm (W) x 60.6cm(H) x 33cm (D). Sealed box.
Basically made from 33cm x 1.8cm cutouts.
Stuffed with pillow fill liberally.

Drivers
High: Selenium D220Ti-8, a screw-on compression driver
Low: Woofer: Eminence Beta 12a - 8 Ohm.

Waveguide:
DJCity YD-L033 - HORN ABS RESIN 162X303X107MM. $20.
This is similar to the Econowave Standard waveguide: JBL 1" thread-on Progressive Transition, Part# 338800-001,
or its clone: Pyle Pro PH612 (partsexpress.com) or Dayton clone

XO: Active, using MiniDSP

High:
High pass filter: 1khz LR4
Horn Notch: 1900hz, -4db, Q=1.4
Gain: -11db
Delay: 0.1ms
Tweeter protect: 68uF non-polar capacitor

Low:
Low pass filter: 1khz LR4
Driver notch 1 : 704hz, -3db, Q=3
Driver notch 2 : 1037hz, -3db, Q=6
Linkwitz transform (note: will depend on box volume & stuffing)
- measured f and Q: f=84.59, Q=0.82
- Target f and Q: f=20hz, Q=0.5. Calculated values:
a0= 1
a1= 1.994770836782980
a2= -0.994777672819971
b0= 1.004151794671880
b1= -1.994713110995730
b2= 0.990683603935339
LT graph


Global (Baffle step compensation):
Shelving low-shelf: 186hz, 6db, Q=1

MEASUREMENT

I encountered dramas with the measurement. My turntable is too small for the speakers (designed for flat dipole baffles). I could not rotate the speakers on its acoustic axis so take this with a grain of salt. I also had to normalise to 0deg. btw. it was unstable enough to fall and smash my window.

IMPRESSIONS

I'm in an early stages of understanding these types of speakers. But they are very positive. Let me address my intial skepticm with box speakers first:
- No, these speakers don't sound "boxy".
- No, the bass does not drone as I observed with many commercial designs.
- Yes, the big 12" drivers do reproduce high frequencies fine. Very fine in fact.


Comparing to the Dipoles, these are my first impressions which may or may not change over time:

BASS
The bass is not bad. In fact I thought they are tighter and went deeper than the dipoles although both are shaped to reproduce 20hz Q=0.5. Initially I thought the midbass was missing, but then baffle step compensation fixed it. The dipoles sounds more "even". This is perhaps why the CD speakers need multisub (?).
Score: DRAW

MID-HIGH
The CD speakers reproduce mid-high flawlessly and this manifest in very realistic female vocals. The dipoles have long had issues with high frequency dispersion which is not ideal and is work in progress.
Score: CD Speakers

IMAGING
The presentation of these speakers is definitely different than dipoles. They are more "forceful" to the ears/brain and very focussed. I am not saying the imaging collapses like generic hifi speakers. They are open, but not like dipoles. This was very impressive at first but I grew tired of it over time like in a few days.

I can relate people who owns dipole listening to these types of speakers and gobsmacked by the imaging. But I really don't think it holds well over time. We'll see.
Score: Dipoles

SPL/VOLUME
Although I knew their sensitivity is ridiculously high (97+db), it is only when turning the knob that I knew the effect. They are willing to play at any level long....long .... before my ears gave up.

Comparing to dipoles, however, it is different. When the dipoles volume is turned up they don't "feel" loud but "more envelopment" rather. The CD speakers on the other hand sounds *very loud*. Again, feels like a forceful speakers. I don't know if it is due to pressure source vs. velocity source nature of both speakers.With the same level of comfort, I actually listen to the dipoles louder.
Score: Unsure (CD speakers if mere SPL is the concern)

Not a straight "These speakers are better than those" it seems. The CD speakers are great speakers in a lot of ways and are cheaper to build than the dipoles. But they may be tiring to listen to over time. If I have Home Theatre (which I don't) then these are the speakers which I would choose. These would be the speakers to impress friends.

The dipoles, on the other hand, are very neutral speakers to listen to music to. The mid-high transition needs to be fixed somehow to be on par with the CD. The dipoles would be the speakers to choose for long term listening, which I do (generally 3-4 hours of listening every day).

I will report in a few weeks time should my opinion change. If you have experience with both types of speakers please do post your opinion

 

[MisterTwister]

I like these macho speakers )
you may try to listen to woofer and compression driver separately.
that "foceful" sound may come from 12" woofer being crossed too high.
also I'd try lower crossover.

 

[gainphile]

"Forceful" is due to lack of better word. I am trying to describe the way these speakers presents the wall of sound.

Perhaps someone else can propose a better word.

SL did a similar study, and his choice of word was "rectangular window"
Spatial distortion

I cannot relate a "rectangular window" in my observations with these speakers. It's hard to describe subjective observation.

I can, however relate to SL's finding that the speakers can be tiring to hear over time.

 

[ra7]

Excellent work and superb summary. You have a way of expressing thoughts in a simple manner that is easy to understand.

I've been enjoying my dipoles too and I'm reading about the Econowaves..... I may build something similar... so keep the thoughts coming.

 

[gainphile]

Dynamics

Thanks. I am struggling with English actually as my native language does not have any Gender, Time (past, present, etc), nor Singular-plural concept of expression

If you're planning to build the ewaves, I can only recommend them. They are superb speakers and the design approach I think is valid. Then hopefully we can compare notes !

I missed this out:

DYNAMICS
One ascpect which I anticipated with monopoles is the "slam" factor. I expected that the CD will perform better in terms of dynamics, the snaps of snare drums, for example. But as I found out, this is not the case. The dipoles are actually more dynamic and snare drums and percussions are reproduced with better transients and gusto. The CD excel in low frequency slam like bass drums. No clear definite winner.
Score: DRAW

 

[Godzilla]

Thanks for your comments. Some may look at what you built and suggest it's nothing more than a dj speaker. I am familiar with dj speakers and the latest ones i 'fixed' had blown tweeters. They were being blasted for hours a day at a karate school. The crossover was 6khz... mated to 15" woofers. Obviously this was done to allow the speaker to handle more power rather than sound its best. Your speaker is built for good sound in a home environment. I like pro sound drivers and bet you have something interesting and worthwhile there.

Zilla
PS, currently i am building something with the Beta 12lta and CD tweeter... but crossed around 8 - 10khz... i am hoping for something good.

 

[ra7]

Thanks Zilla! I can provide some perspective on the other end... I've heard a big tractrix horn with JBL CDs mated to two sealed kappa pro 12" woofers. The crossover was 400 Hz. Ohhh... what realistic sound. Simply effortless and real. Trumpets sound like trumpets... one of the hardest instruments to reproduce correctly.

I bet the ewaves don't do this and probably because the horn is crossed too high. Anyway, crossing it lower using a horn is a whole different ball game... it certainly won't be 'econo!'

 

[DQ828]

Hi Gainphile

Looking good, great post as usual, will follow with interest.

I assume the "Tweeter protect: 68uF non-polar capacitor" is just incase the MiniDSP goes haywire & sends the LF to the tweeter?

How did you implement the "Tweeter protect: 68uF non-polar capacitor" a photo would be good

How did you learn how to do the the "Transform Thingy" Would it be useful in my new OB's I'm going to start playing with the MiniDSP in a couple of days.

David

 

[gainphile]

I assume the "Tweeter protect: 68uF non-polar capacitor" is just incase the MiniDSP goes haywire & sends the LF to the tweeter?

How did you implement the "Tweeter protect: 68uF non-polar capacitor" a photo would be good

How did you learn how to do the the "Transform Thingy" Would it be useful in my new OB's I'm going to start playing with the MiniDSP in a couple of days.

David

Actually in case someone turned the DSP off before the amps. It produces a whacking thump

The cap is basically a 300hz highpass 1st order passive xo:

amp ==> 68uf caps ==> tweeters

The implication is there will be phase shift so this needs to be accounted in the delay.

About Linkwitz transorm, yes it's one of the most powerful filter in existence. Even more important with dipoles. Have a read here:
Active Filters

And for miniDSP implementation, download the spreadsheet:
MiniDSP - Linkwitz Transform

 

[DQ828]

About Linkwitz transorm, yes it's one of the most powerful filter in existence. Even more important with dipoles. Have a read here:
Active Filters

And for miniDSP implementation, download the spreadsheet:
MiniDSP - Linkwitz Transform

Are yes, that's the stuff that makes my brain hurt

Thanks

David

 

[DQ828]

Actually in case someone turned the DSP off before the amps. It produces a whacking thump

The cap is basically a 300hz highpass 1st order passive xo:

amp ==> 68uf caps ==> tweeters

The implication is there will be phase shift so this needs to be accounted in the delay.

About Linkwitz transorm, yes it's one of the most powerful filter in existence. Even more important with dipoles. Have a read here:
Active Filters

And for miniDSP implementation, download the spreadsheet:
MiniDSP - Linkwitz Transform

Gainphile

The info says "First you have to determine the f(0) and Q(0) for your box." I don't have a box, I have a "H" frame, what should I do

David

 

[gainphile]

Email sent as it's a bit OT

 

[john k...]

Gainphile

The info says "First you have to determine the f(0) and Q(0) for your box." I don't have a box, I have a "H" frame, what should I do

David

You may be interested in this. The LT is not really the way to go with dipole woofers.

Also, you can download my ABC Diople for free here. It has a variety of filter calculators in the Excel spread sheet.

 

[gainphile]

The LT should be used to eq the driver. On top of that the 6db/oct dipole rolloff eq is still needed.

 

[john k...]

The LT should be used to eq the driver. On top of that the 6db/oct dipole rolloff eq is still needed.

The problem with the LT in general is that it can not be used for all values of Q. The second thing is that if it is applicable and is used to shift the poles, the lower pole of 6dB eq must be well below the cut off frequency. It is basically what I suggest for drivers with Q > 0.5 and acceptable Fs and Qts for a 2nd order roll off but with the LT shifting Fs and Qts. The response rolls off 3rd order below the lower pole of the 6dB eq.

However, in general a dipole woofer should roll off at least 3rd order below the specified Fs to prevent excessive excursion below Fs.

 

[bwaslo]

High:
High pass filter: 1khz LR4
Horn Notch: 1900hz, -4db, Q=1.4
Gain: -11db
Delay: 0.1ms
Tweeter protect: 68uF non-polar capacitor

Gainphile - how are you doing CD compensation (response lift starting at ~6kHz)? I don't see that in your settings.

Without that, the high end would seem pretty lifeless I'd think. Compression drivers have a natural mass rolloff at high frequencies, which normal horns tend to flatten due to their beaming. but constant directivity waveguides don't on their own). See the ewave passive crossovers, where they use a series resistor shunted by a cap to do this.

 

[cuibono]

Nice work gainphile! I've been interested in this design for a while, and will someday build a set, maybe for a HT setup.

Could you post a regular polar graph? And have you considered adding an edge radius to the box and foam plug to the horn? I've read that these make a big difference, and people find the speakers sound less fatiguing.

BTW, what is your native language? That is interesting it doesn't have past-present-future distinctions.

 

[CLS]

Well done!

...

DYNAMICS
One ascpect which I anticipated with monopoles is the "slam" factor. I expected that the CD will perform better in terms of dynamics, the snaps of snare drums, for example. But as I found out, this is not the case. The dipoles are actually more dynamic and snare drums and percussions are reproduced with better transients and gusto. The CD excel in low frequency slam like bass drums. No clear definite winner.
Score: DRAW

Maybe the dipole has smaller 'tail' after the impulse?

A well-designed sealed box usually has a tight and clean bass sound, might be somewhat muted when overdamped. It's a (pleasingly) surprise that a severely EQ'ed box has a good lower bass.

And, well, it can never be apple to apple in this comparison, but once you have bigger cone area for the dipole, it can also excel in low bass, at least for hearing. (good low 30's give a sense of 'very deep', under that, it becomes vibrations but not sound.... )

Also, on dipole, I found multiple subs benefit, too. I'm using 3 channels now. It's obvious better than 2.

Talking about the senses of dynamic and impact (or slam... etc) of bass sound, the most powerful one I've heard is horn bass - front loaded with sealed back chamber and full size mouth (not those commercial ones with severe truncation). Just overwhelming.

Interestingly, dipole bass (in its best condition) can deliver that kind of 'feel' - huge contrast, very fast rise and fall... Not the same level of sheer power, but the 'feel' can be very close

 

[dantheman]

Great write up Gain! I had a similar experience with dipoles(coax pro 15 and compression driver) and direct radiators(Pro 10" and dome tweeter)--the dipole sounding more dynamic. However, when I went with a Dayton WG, D220, and Eminence Delta Pro 12A, in a similar fashion to the E-wave, dynamics seemed restored. I changed that woofer to a Peerless 10" and it seemed I again lost the dynamics but got a better sound balance, polar response, and deeper bass in return.

I wouldn't give up on either of your systems yet.

I'm curious as to what you make of the pseudo Pluto vs. pseudo EW vs. pseudo Note CD dipole. The Pluto essentially fits Dr. Toole's criteria fairly well. You're sort of in a unique position to have a toned down, but still awesome, version of Dr. Toole's findings/S. Linkwitz design, John K OB implementation, and Dr. Geddes perspective on loudspeakers at your fingertips.

Have you set those EWs up like Dr. Geddes recommends? The 45 degree cross. The most spacious presentation I've heard from a pair of loudspeaker requires that set up.

Great thread,

Dan

 

[gainphile]

Gainphile - how are you doing CD compensation (response lift starting at ~6kHz)? I don't see that in your settings.

Without that, the high end would seem pretty lifeless I'd think. Compression drivers have a natural mass rolloff at high frequencies, which normal horns tend to flatten due to their beaming. but constant directivity waveguides don't on their own). See the ewave passive crossovers, where they use a series resistor shunted by a cap to do this.

My implementation is active, so I simply notch the lower hump with appropriate cut and Q. It is very effective and easy to do.