A first - sealed book shelf speaker - own use

[Hari Iyer]

Over the past 15 years i have build various types of box loading from - bass reflex, transmission line and open baffle. But i have never done a sealed box before. Most DIY builders start the journey exactly the reverse order - sealed, BR, TL and OB. Its time i complete the cycle with my first ever sealed box. This sealed box will be used in place of my OB midrange. I will still be retaining the H-frame sub for freq below 100Hz.

I will be using my Russian dirver which i had purchased 6 years ago. Initially i had used them in an OB loading, but this driver is meant more for a sealed box because of its extreme high compiance (1.4mm/N), low resonance and Qts 0f 0.72. The measured response by other DIY builders using this driver is from 63Hz till 20KHz. I had done simulation on my 30 liter enclosure using a metronome shaped baffle. The F10 is around 47Hz. I don't think i would be requring the sub in most situations but for now, i shall be retaining them. If not required after listening, the H-frames will be put for sale.

Below is the Basta simulation on my sloped baffle:




The wood work is planned to begin next month and shall post updates as appicable.
Thanks for looking.

 

[Hari Iyer]

IME, most modern architecture homes in Mumbai/ Thane region are not speaker placement friendly. Average living room size are around 10x16 feet which prevents placement of speakers in room without obstruction somewhere. Most of my contacts feel that it's next to impossible to find the right place to put a pair of speakers - FS & BS included. You need to compromise on certain aspects of sound staging for smaller rooms.

Even my observation is similar. I have been doing countless simulations past 2 weeks, but am not able to arrive at the right dimension considering my main door location, TV position and moving space around.

I have been discussing with my elder brother about this past one week and we are exchanging ideas about how this should be finally panning out.

My family members have been considerate enough to allow 2 H-frame woofers and an OB midrange which not only have poor optics but is also obstructive. My new design will be a waste if I don't take that into consideration. After all they have been tolerating my madness all these years. I am even ok to compromise on some aspect of SQ to achieve this.

Let me see what I come up with.

Thanks for looking.

 

[hifimaddy]

It's nice to hear the stories or observations you are considering for designing speaker cabinets and speakers. Very much appreciated for your effort.
You are very right. Every room is not ideal for home theatre use or stereo listing. but due to our budget and space constraints now we all adopted to the space we have, and all the new technologies are capable of optimising the audio according to the room.
I am a professional architect. So we studied and are still considering "golden ratio" dimensions for the spaces that are the optimum dimensions for a stereo listing and also home theatre use. The golden ratio is used in all fields for designing everything we see around us and also pantings in olden days.
You can also consider the "golden ratio" for designing your speaker cabinet, but I don't know how it will affect your situation. Any way, keep going and happy to see your updates.

 

[Hari Iyer]

It's nice to hear the stories or observations you are considering for designing speaker cabinets and speakers. Very much appreciated for your effort.
You are very right. Every room is not ideal for home theatre use or stereo listing. but due to our budget and space constraints now we all adopted to the space we have, and all the new technologies are capable of optimising the audio according to the room.
I am a professional architect. So we studied and are still considering "golden ratio" dimensions for the spaces that are the optimum dimensions for a stereo listing and also home theatre use. The golden ratio is used in all fields for designing everything we see around us and also pantings in olden days.
You can also consider the "golden ratio" for designing your speaker cabinet, but I don't know how it will affect your situation. Any way, keep going and happy to see your updates.

Thanks for your suggestion. Luckily my living room is already a GR - 10x 16. Also i have designed the box to be a GR for all the dimensions. Need to do proper scaling to prevent obstruction of TV viewing and movement space. I have come up with an hybrid design (not BS and not FS - but somewhere in-between) for achieving. My equipment rack is at a 2 feet high. On top of the rack i am planning to put my hybrid speakers, The idea was to have a pyramid shape box so that optics are reasonable and at the same time due to the sloping baffle will not obstruct TV viewing. I may not be pulling the speakers from the wall as that would mean keeping them on the floor and resticting moving space in the room. The base of the pyramid is 11" and base depth is 6.875" which is a GR. similarly the Top width is 6.875" and top depth is 4.2968" which again is a GR. The height is 36.8" from base to top. The driver will be mounted at 39" from floor considering my rack height of 24 inch which is 15" from base of the pyramid for ear height when seated. I have done Basta simulation based on the wall distabce and baffle distance and found some dip around 200Hz, But again t am ok with this dip provided all other parameters are ok.

Below is the baffle shape used for simulation of hybrid box in my room,

 

[Hari Iyer]

My observation of using drivers and box loading with respect to measurements require some analysis to understand why they measure a particular way. TBH, i am neither an expert in measurement and neither on analysis of the measurement. Nor do i have any dedicated anechoic room for measurement to avoid room reflections. Having said that, i have done countless number of measurements of my speakers and have analyzed them in my own way looking at what the experts have to say. Having done some reading about various drivers and box loading and my own built experience, here are my observations that i wish to put into my practice for this built. It will also validate if i am on the right path,

- Low frequency woofers the cone motion is a bit slower than mid-frequency and high. So even if the voice coil of the woofer, mid-range and tweeter are perfectly aligned in time, the cone motion will be different for all the 3 drivers and hence the step response will not be accurate
- If all the 3 drivers woofer (low compliance), mid-range(low compliance) and tweeter are mounted in an OB loading, then the mid-range cone for its frequency will move a bit faster than the woofer cone causing time smeaer even though the drivers voice coil are aligned in time
- If all the 3 drivers above are mounted in a BR enclosure, the woofer will be delayed severely in time compared to the mid-range and even though there is a delay using DSP, there will be time smear and the sound won't be coherent.
- if all the 3 drivers above are mounted in a sealed enclosure, and due to the low Q of the box and if the enclosure is critically damped with the correct amount of stuffing, and the drivers voice coil are aligned in time, the chances of better integration of the 3 drivers and time coherence is high. But to get a reasonable low end such boxes are quite huge and will be difficult to integrate in a standard living room.
- one option is to put the woofer (low compliance) in an h-frame OB so that Q is low and the box size is reasonable and instead of putting the mid-range too in an OB, put them in a sealed enclosure. When the midrange is also place in an OB, there is a velocity mismatch between woofer and midrange cone again it will cause a time smear. But when the midrange is placed in a sealed box with a low Q and proper damping, the cone velocity of the midrange will be controlled by the Q of the box and will be damped a bit more than being in an OB loading. So the chances of the woofer and midrange cone velocity being almost equal is higher with this arrangement. This is what i want to try in this project based on my experience building speakers.

Also as per musicandesign.net website, the free space directivity factor for a monopole = 1 and for dipole = 3. Also the total radiated power for a dipole will be 4pi/3 while for a monopole having a DF of 1 will radiate in 2pi space as 2pi/1. Also a monopole midrange radiating into
2Pi space radiates only 1.76 dB more total power than a dipole midrange radiating into 4Pi space. So the acoustic power matching between the woofer and midrange becomes easy with a dipole woofer and a monopole midrange atleast in theory. Will need to check this aspect too in this built.

Below is the measured step resoponse with a dipole woofer and dipole midrange,


The tweeter/midrange rising and falling response is too fast for the woofer to keep pace within fraction of a msec even though all 3 drivers are aligned in time due to differing cone velocities of the drivers.

By putting the midrange in a sealed enclosure and by damping the cone vibrations by stuffing the enclosure, i am expecting the cone movement of the midrange to be reduced a bit. The driver that is used in this sealed box is a high compliance type so that the stiffness of the cone suspension will get a correct spring action from the cabinet walls.

Thanks for looking.

 

[Hari Iyer]

After a wait of one month the sealed box got completed yesterday. I have yet to polish them and will do that after 2 weeks. I yesterday connected them directly to my tube monoblocks without any regards or attention to room placement. I immediately noticed a very polite mids and highs compared to my OB mids. The OB mids are quite sweet sounding with the same drivers. This could be because my front wall is very reflective because of a 55" TV and other hardware which impacts the reverse phase causing cancellations. Ideally for an OB midrange the front wall should be diffusive. Also noticed roll-off much earlier and there was no deep punch for the bass though it was there. May be i over-damped the enclosure. I did not do any measurements and removed some 20% of the damping material. But this did not help me to increase the bottom end.

Today i measured the box and noticed that the LF roll-off to be around 100Hz and the F10 was at 76Hz as below,



This explains the lack of bass for the speakers because of an earlier roll-off. I went ahead and added my OB-sub and adjusted the response to get the below,



with the H-frame OB the -3dB is now at 30Hz which is quite good for my small room. To my suprise there was no artificial boom or honkiness in the entire response and the sound stage and dynamics were excellent. I am liking this sound more than an entire OB setup for lows, mids and highs. Below are the final setup images,






Thanks for looking.

 

[Hari Iyer]

I was intrigued by the excellent dynamics, sound stage and integration of the H-frame (Dipole) and the sealed box (Monopole) and researched why they sound much better than an full OB setup. The reason lie in how the directivity and power response of both the dipole and monopole. The dipole woofer radiates in a 2Pi/DF area when the woofer is near the ground plane. Typically if you have a dipole midrange they will radiate in 4pi/DF area as the mid-range driver will be around 3 feet above the ground plane. Because of this the power response of the dipole woofer will be -4.33dB below the midrange and it becomes challenging to integrate the two even by intensity matching. Because of this the midrange will be a little bright for an entire OB setup.

For a monopole, the radiation above the baffle step is 2Pi/DF and below the baffle step is 4Pi/DF. If the monopole mid-range is allowed to roll-off around 150Hz then it will radiate in 2pi/DF above 150Hz which will be same as the dipole woofer radiating in 2pi/DF as the woofer are very close to the ground plane. This allows easy power matching and intensity matching. I do not have the means to measure the DF of my speaker. Any FM who have measured their speaker DF in room - please add your inputs / suggestion.

FM interested in know more of this can check this link https://musicanddesign.speakerdesign.net/PowerMatching.html

Thanks for looking.

 

[Hari Iyer]

I discovered the below link today and thought of sharing

 

[Hari Iyer]

After some listening i converted the OB subwoofer into just a woofer and they are now connected in series with my sealed enclosure for a net nominal impedance of 6 ohms which also matches with my softone OPT secondary winding of 6 ohms. With the subwoofer wiring removed, i lost some bass and now LF roll-off has shifted an octave above. Also the dipole roll-off is calculated at 200Hz for my H-frame which also agrees with my measurement. Also the dipole peak is at 200Hz. I have only added a passive low-pass shelving at 200Hz where the -3db is around 340Hz and it flattens at around 480Hz onwards at -6dB. But i have yet to add the notch filter at 200Hz to reduce the dipole peak by -6db. Here are the measurement taken today morning with the mic at the listening position centered between the left and right speakers at a distance of 9 feet from the speakers with the grill cloth installed,



The dipole peak at 200Hz needs to be removed with a notch filter which i have designed with LT spice simulator. I need to check component availability and add them and measure again for checking if the dipole peak at 200Hz is reduced by -6db.

 

[Hari Iyer]

If you look at post #9 then you can see the hump at 200Hz and the following roll-off @ -18dB/octave caused by the dipole peak and the roll-off of the woofer beyond the baffle step below the driver resonance. i tried passive compensation between the preamplifier and my monoblocks but only the LF shelving seem to work effectively. Creating a notch @ 200Hz for -12db was more of a challenge and hence abandon this approach. Without this EQ the system sounds a bit boomy at the mid-bass masking all other frequencies and hence need to be EQed.

I then thought of bi-amping again using my 31+31 band dbx GE as both for boosting the freq from 20Hz till 60Hz and also creating this -12dB cut from 125Hz onwards. But what i needed was a overall roll-off of -18dB after 80Hz and hence introduced a 3.7mH inductor in series with the H-frame to give that extra -6dB/octave cut after 80Hz. I did some measurements only for the H-frame as below,





The sealed enclosure are measured as below:





All the above measurements are at a distance of 9 feet from the speakers at the listening position centered between the 2 speakers.

What i have realized using with an H-frame OB is that, it can never be a plug and play speaker. You will need some serious EQ and notch for it to power and level match with the rest of the band-width. Hence most OB setup can work only with bi-amping very well unless you have an excellent full range driver that covers from 20Hz till 20KHz. Even with such a driver the dipole peak and dipole roll-off will need some equalization for it to be flat and linear in the listening range.

Thanks for looking.

 

[Hari Iyer]

This is how the speaker currently looks,

I am planning to replace the circular wooden waveguide with an exponential cast aluminum horn. Just wanted to check if i am able to get more center image focus with that compared to the waveguide.

The above setup measures as below,



Thanks for looking.


I

 

[Hari Iyer]

Last week i replaced the wooden waveguide with an exponential horn and also changed my speaker placement for a more practical approach that doesn't block my moving space, obstructing viewing TV from my sofa and does not block the main door. This placement gives me pin-point imaging, channel separation and staging without loosing on speed, dynamics and articulation. I may 100% lock this as my final speaker placement now. Only remaining is to put a bracket for mounting the horn and anodizing the cast aluminum bracket and fixing them. I have already figured out the best position for the horn on top of the sealed box by measuring the step response for correct driver integration. This how they will look,

 

[mahaaprasaad]

Last week i replaced the wooden waveguide with an exponential horn and also changed my speaker placement for a more practical approach that doesn't block my moving space, obstructing viewing TV from my sofa and does not block the main door. This placement gives me pin-point imaging, channel separation and staging without loosing on speed, dynamics and articulation. I may 100% lock this as my final speaker placement now. Only remaining is to put a bracket for mounting the horn and anodizing the cast aluminum bracket and fixing them. I have already figured out the best position for the horn on top of the sealed box by measuring the step response for correct driver integration. This how they will look,

View attachment 88523

Beautifull build and looks very classis. Where you found that Grill Cloth... ,mixed well with the cabinet texture and color.

 

[Hari Iyer]

Beautifull build and looks very classis. Where you found that Grill Cloth... ,mixed well with the cabinet texture and color.

It's actually leftover cloth of my sofa cover which I have reutilzed here as grill cloth

 

[sbg]

It's actually leftover cloth of my sofa cover which I have reutilzed here as grill cloth

Now!, where can I find these Acoustically Transparent Sofa Covers.

 

[Hari Iyer]

Now!, where can I find these Acoustically Transparent Sofa Covers.

There is a 5dB gain in my sealed box from around 330hz onwards till 7khz due to the baffle step. This grill cloth surprisingly attenuates this gain exactly from 350hz onwards by -5dB till 7khz. So the speakers needs to be listened with the grill put on them. I could have added a BSC ckt to reduce the gain accordingly, but this would mean I would need to remove the grill every time I am listening which would be inconvenient. So I preferred an acoustic solution that worked.

I had actually tried 3 grill cloth,
- acoustic transparent standard spkr grill cloth
- Raymond pant piece - this reduced by -12dB which was way too much
- the sofa cover cloth was very lossy and did the trick.

Unfortunately this particular cloth won't be available online as I bought them from my vendor below my apartment. If you need PM me.

 

[sbg]

There is a 5dB gain in my sealed box from around 330hz onwards till 7khz due to the baffle step. This grill cloth surprisingly attenuates this gain exactly from 350hz onwards by -5dB till 7khz. So the speakers needs to be listened with the grill put on them. I could have added a BSC ckt to reduce the gain accordingly, but this would mean I would need to remove the grill every time I am listening which would be inconvenient. So I preferred an acoustic solution that worked.

I had actually tried 3 grill cloth,
- acoustic transparent standard spkr grill cloth
- Raymond pant piece - this reduced by -12dB which was way too much
- the sofa cover cloth was very lossy and did the trick.

Unfortunately this particular cloth won't be available online as I bought them from my vendor below my apartment. If you need PM me.

No, I was surprised that you could just randomly use surplus cloth for the speaker.

 

[Hari Iyer]

No, I was surprised that you could just randomly use surplus cloth for the speaker.

I randomly tried cloth and finally settle for the one that worked for me.

 

[rajan-]

interesting to see someone's journey from the land of OB to a box + horns (a la vintage klipsch kind of sound) .

once you hear a good horn, its very difficult to get it out of head, and self denying it, just makes it progressively worse

not sure Mr. Viren Bakshi still sells the solid milled teakwood LeCleach profile horns separately.
Get them while you can if you are going to stay with horns. Not hearsay, I do have them & they are going to stay with me. they travelled with me when I moved continents.

a while back , when I built my first set of horns, someone here on the forum who is an avid OB enthusiast mentioned that "OBs are mirrors & horns are joy" . that statement seems to sum it up well.

 

[Hari Iyer]

interesting to see someone's journey from the land of OB to a box + horns (a la vintage klipsch kind of sound) .

once you hear a good horn, its very difficult to get it out of head, and self denying it, just makes it progressively worse

not sure Mr. Viren Bakshi still sells the solid milled teakwood LeCleach profile horns separately.
Get them while you can if you are going to stay with horns. Not hearsay, I do have them & they are going to stay with me. they travelled with me when I moved continents.

a while back , when I built my first set of horns, someone here on the forum who is an avid OB enthusiast mentioned that "OBs are mirrors & horns are joy" . that statement seems to sum it up well.

Thanks for the information about Viren's horn. I actually have a wooden waveguide made of walnut wood that I was using before the cast aluminium horn. That too was quite good. The horn gave me some gain and a wider horizontal dispersion which was missing in the waveguide. I have now tilted 30 deg for both speakers to face me while seated on the sofa and both speakers are almost 8 feet wide from each other. This gives me better staging, separation and imaging without compromise on speed or dynamics. This setup is quite addictive and brings out the best in rest of my setup.