A 3 way design study

This is the kind of speaker I was talking about a few posts earlier in this thread 🙂

An all purifi design which I would like to listen to, someday......

https://forum.marchaudio.com/index....ZYDFVgCvvkPwncKjJg_aem_yWHl8AIZ81RAx1Q1c5sAew

1000207726.jpg
 

Attachments

  • 1000207726.jpg
    1000207726.jpg
    9.7 KB · Views: 62
@tktran303: Because of these reasons and more 🙂

Another speaker that I would like to listen to is this:

https://audiolabinsight.com/t/ascilab/143/608

Someday I want to compare something like the March audio speaker pulled out into rooms vs the ascilab cardioid speaker (with all Purifi drivers)
 
  • Like
Reactions: otto88
Finally, I can hear the "PureBliss" speakers in stereo 🙂
1000208733.jpg


The first test print of the shallow T34A waveguide is ready as well..
1000208734.jpg


The waveguide print is not exactly 100 percent fault free but i can live with this for test fitting purposes.. I also forgot to add the screw mounting holes in the waveguide and the spacer between the tweeter body and the waveguide as well, in this iteration.. 🙂
Next time...

The more I have been listening to this Purebliss speaker system, the more I like them..
I can confidently say that this is the overall best speaker I have listened to in the last 4 years of my speaker making journey in terms of balanced sound from bottom to top (comparing to other mono speakers though)..
Rather than me trying to cook up descriptions of sound, let me copy few words from hificompass website about his Pharaoh speaker design (https://hificompass.com/en/projects/2-way-systems/pharaoh) that uses this tweeter (because I felt exactly the same way)

"Just note that the sound of the speaker is different from all those two-way systems with which we had to deal with. First, you do not immediately understand what is happening, since all the test records you know sound with newly opening details, which you simply did not notice before. The accents are placed in the spatial picture by the different way. The sound is very smooth, tonally balanced and, for all its analyticity, preserves musicality.
After 15-20 minutes all the unusual sensations disappear and you just stick to the music, trying to make up for the lost feelings for many years, have forgotten that you were going to sit down for only 5 minutes to test the sound 🙂 "

I also feel that these impressions are true and are due to his axial/listening window flat 'ish' crossover design for his speaker (ignoring any tiny power response bumps) and sitting less than 2.5m away from the speakers.. (off course the directivity and smoothness of all frequency response curves will be slightly better than mine in hificompass speaker due to his baffle shape).

I hope the waveguide makes things everything even smoother with in my speaker (in terms of the graphs, atleast 😉 ) But let's see...
 
Last edited:
I think I have completed a full circle now.. 🙂 Starting this thread with a direct radiating/tweeter on a small-waveguide 3way speaker, going through a lot of deviations and experimentation and at this point, back to the same kind of speaker
1743435823776.png


All that has changed is the mid and the tweeter. And yet there is a feeling that so much has changed in terms of sound.. 🙂

This is the current crossover in use
1743437198303.png


Measurements of both speakers from MLP
1743437365325.png


1743437400244.png


Floor + ceiling (some amount of left side wall) majorly contributes to the nulls around 200Hz, 500Hz & 900 Hz.
That 65Hz null may be a room mode or a reflection I guess

Group delay plots
1743437515730.png

1743437580748.png


To compare, here is how the big, dual 15inch driver-based cardioidish + waveguide speaker FDWed response looked like vs current speaker in almost exactly the same position

Big cardioid 3way
1743437888809.png


Current Purifi + Bliesma + Satori 3way
1743438029073.png

From the above measurements, it looks like my cardioid speaker was relatively worse w.r.t impact of reflections at the above position.
Lesson learnt: So placement does seem to be important even with a cardioid speaker. A nice cardioidish pattern is not the solution to all problems with boundary interference.
Subjective impressions of stereo Purifi+T34A+Satori WO24P system

1) There is an ease/effortlessness/balance in the sound through the whole spectrum of sounds I care about that I haven't found in any of the speakers I have tried so far.

2) That slightly rib cage massaging bass is back with this woofer system and I am happy, again.. 🙂 Maybe even the room mode has something to do with it.😀

3) The imaging is ever so slightly hazy/not as well defined as the horn system, which is probably to be expected. But I don't miss the etched out sort of imaging. It feels like all my brain was doing with that big cardioid + horn system was focussing on the midrange and not paying enough attention to the finer details in treble. This is system is much more balanced in that sense. I can hear a lot of things in music, which I have "never heard" before 🙂 (all those little nuances are so enjoyable to hear that and wonder how I missed all of those earlier).

4) As I have remarked in the previous post, the spatial presentation is completely different to any other speaker I have ever heard. The depth and width of the soundstage have increased significantly. Also, the soundstage is now occupied by all of those little nuances in treble and mids. I wonder what measures that.. Distortion maybe??

5) Someone who has spend years hearing a seriously sound-focussing horn/waveguide system is going to be disappointed for sure.. 🙂 Missing all that emphasis on the mids.. Maybe someone who has been hearing a very well balanced horn system may find things have little dulled out.. 😉 Eaither way, I don't care.. I like this.. I will have a horn/waveguided system as a second system for comparisons and to use it when I get that urge to hear that focussed sound..

To summarize, some of my impressions are very similar to hificompass speaker I referenced above, again. So I quote (https://hificompass.com/en/projects/2-way-systems/pharaoh):

"The sound stage is very impressive - wide and deep with the preservation of the separation of virtual images. Magnificent transparency, localization is excellent too. Most likely, that all are attributes of wide dispersion and hard membranes of the tweeter and midwoofer. The lower register is full and quite impressive for the speakers of this size - this is the second moment that surprises the first time you listen. The transients and dynamics are just as a hurricane, you need to hear it at least once"

But now it is really, finally time to focus on that room treatmen, whatever little more I can do about that... 😉
And to try out that shallow waveguide for the tweeter...................
 

Attachments

  • 1743437964738.png
    1743437964738.png
    21.9 KB · Views: 23
  • 1743437320449.png
    1743437320449.png
    41 KB · Views: 34
Last edited:
My primary interest in this kind of a hybrid speaker was because of something I saw in my (relatively crude) simulations for an open baffle bass+cardioid mid + waveguided tweeter setup in the monkeybox thread
This was my simulation
1744862364988.png


The radiation pattern looks quite similar to what @Juhazi posted above for the silent pound speaker.. 🙂
When I put it in a room in a situation like below and included all the major reflections,
1744862422780.png


I got something like this
1744862461473.png


(neglecting the above 1kHz PIR), I find the impact of the major dips often seen in the 200Hz to 500Hz region much less compared to even full-range cardioid speakers, which looked like below
1744862632912.png

1744862570906.png


Then the silent pound speaker is relatively slim, relatively good looking etc.. 🙂
Interesting concept.. If I didn't have a house full of speakers already, I might have pursued that concept and would have tried to optimize it even more.. 😀
 
In other news, I have been enjoying the current 3way set up thoroughly.. 🙂
I was curious about what kind of response I was getting at MLP and I tried out a simple exercise of try to make some sense of the graphs comparing it with MLP response.
The original speaker response in crossover design looked like this
1744862829366.png


The MLP response looks like this
1744862843895.png


It looks like I might be perceive more of a mix of the listening window response below 3-4kHz and more of the power response curve above 3-4kHz (in the above ML response filter target, the slope down above 4kHz is -0.8 to 0.9dB/oct as is indicated in the target slope in the CTA 2034 curve above).

If that is the case, getting rid of those power response lumpiness above 3-4 kHz might improve the sound and make it even better somehow.. 😀

The radiation pattern might change from this:
1744863077813.png


to this (from @fluid 's post about T34A waveguide):
1744863119994.png


and the power response can change from this:

1744863215947.png


to something like this
1744863242342.png


Got to fit that waveguide on the tweeter...... 😀
 

Attachments

  • 1744863207316.png
    1744863207316.png
    46 KB · Views: 8
(neglecting the above 1kHz PIR), I find the impact of the major dips often seen in the 200Hz to 500Hz region much less compared to even full-range cardioid speakers, which looked like below
That's interesting.
There is clear difference in vertical directivity.

Is it the same drivers?
Could this be due to the drivers vertical arrangement or c2c ratio ?

Or maybe the dipole radiation offers a better load of the room near and below Schroder transition...
 
Last edited:
  • Like
Reactions: vineethkumar01
the most interesting aspect of the silent pound speaker to me is the vents on the front instead of the sides. That is an idea I might want to try, except for the fact that I'm getting very good results from side vents. Perhaps the front vents are needed because of the relatively wide baffle relative to the driver size. We saw that on the Sigberg Manta.

I'm not impressed by dual 12" OB woofers; I've done enough sims to know they would need help from a sub. A single OB 12 above a sealed 12 would do better. I've got that simulation somewhere on my computer. I got the best SBIR results in my OB sims aiming the side nulls at the first reflection points.
I use two metrics in comparing these virtual designs in search of one that merits being built. One is the Vituix predicted preference rating, an albeit poor substitute for listening but the best I can do in that regard and the other is how its SBIR looks. For comparison with your sims, here is what I have for an almost full range cardioid (100 Hz up) floorstander, with an 8" coax at its top. It is remarkably flat up to the floor reflection null.

Floorstand_QuadSides CTA-2034 with Reflections.png

I should caution that these SBIR results are highly dependent on ceiling and driver height. The result above is the best I've been able to do in my cardioid coax paradigm with 8' ceiling. Perhaps with separate waveguide and tweeter, one can find that magic set of driver heights that eliminates that remaining null. I would love to hear opinions on its audibility. Certainly, its something that one has to live with with most designs.
 
So the proof of concept simulations I showed earlier were done using real measurements from my SICA 5.5-inch coaxial driver in passive cardioid cabinet + 2 x 8-inch sealed woofers, one placed in the front of the cabinet and one placed at the back.

With appropriate DSP control, one can change the pattern to dipole bass, omni bass, cardioid bass or something in between, depending upon one's requirement/application.. 🙂

For example, proof of concept dipole bass + cardioid mid configuration
1744890061709.png

1744890180510.png


Now making it full-range cardioid
1744890230908.png


1744890263243.png



As per the application, one can scale up the woofer sizes as per requirements, change coax size, change the coax to a conventional mid+tweeter set up etc and also change the DSP settings to see if nulls can be widened/redirected to other directions as well so as to combat SBIR as per one's p[placement scenario in room.

As @nc535 said, the response is dependent on the placement situation,
Adding side woofers makes it even more configurable and fun.. 🙂
 
All written below is my opinion, not a fact.

As I remember my OB in WMTMW configuration needed much less room correction than WWTM which I have now.

I am not sure, as I have no measurements without correction any more, but OB in WMTMW configuration make sort of line array and significantly reduce floor/ceiling reflection. IMO, having bass drivers distributed on different height (speakers were 180 cm tall) make room modes differently excited (sort of multisub approach), as well.

It is not that I am not satisfied with wwtm but I feel that previous was better.

So probably will go back to WMTMW configuration, only do not know how to involve recently made ESS WG which is to big for it.