My Dream Project: Planning a TL with Fostex Full Range Driver

[Hari Iyer]

This is what i get with ML the line.

Length = 36.5 in
Cross-section = 100 sq.in
Off-set = 3.5"
Terminus Area = 7 sq.in
Terminua length = 1.8 in
Tuned @ 49 Hz

Unstuffed:

Stuffed: 0.5 lb/ft3

Looking at the graphs makes me lean more towards the ML design because of the following,
- The wrinkles are very linear in the unstuffed line
- Tuning is close to the driver resonance
- The unstuffed impedance peak is well tuned and without distortion
- The TL Acoustic impedance is well balanced but all the harmonics are too much excited
- Allows low stuffing to reduce the odd harmonics
- Above all the line length has a good WAF
- Allows golden ratio cross-section

Waiting for Kapvin views too.

I am feeling a bit uncomfortable with the double hump of the stuffed response. Also the enclosure is tuned below resonance. This being a low Qtc driver needs to be tuned around 5 Hz to 15Hz above resonance for a well damped cone motion. If not controlled could interfere with the mid-freq response that could ruin the overall linearity.
Not sure if i should be designing with these values.

Today did a new simulation with a Straight Line TL and focused my design on the system phase. There is no ML & no taper too. This gives me a single impedance peak with a little more stuffing. A little offset gives me a better control on the unstuffed wrinkles too. Agreed the gain at low frequency response is not much (around +2dB in the LA) but the overall system linearity is great(terminus, phase, TL acoustic Z, Electrical Z, Group delay, and terminus velocity).

More over the above is a tried and tested design and tweaking will involve only adjusting the stuffing to taste. I am close now to finalize this design shortly.

Did some more simulations with the expanding taper geometry and tuning around 12Hz above resonance and with a taper ratio of 2. Also altered the closed and open area, offsets to arrive at a more acceptable 3rd harmonic wrinkle, tuning and group delay spikes. This gave me a much wanted low freq lift at 50Hz and a roll-off much controlled @ 3.3dB/ octave from 100Hz to 50Hz and a very linear 12dB/octave @ 33Hz. The SPL phase is also well balanced with the suggested stuffing density @ 0 deg.
Above all also got approval for WAF parameter
So this is now going to be my final box. Have called my carpenter tomorrow. Hope he turns up tomorrow so that my box will be ready by the time i am back from Bangalore after completing my office project.
Will post simulations if anyone wants to have a look.

This is how the box is going to look from the front without the rounding.

 

[jls001]

Please show driver location too. Is the cabinet tapering at the top? Also, 50 Hz hump could possibly lead to bass boom as most RCC houses exhibit about 6 dB room gain.

 

[Hari Iyer]

Please show driver location too. Is the cabinet tapering at the top? Also, 50 Hz hump could possibly lead to bass boom as most RCC houses exhibit about 6 dB room gain.

SL = 64 sq.in
So = 128 sq.in
TR = 2
OS = 8.3 in
LL = 55 in
tuned @ 65Hz

the -3dB is at 50Hz, -12 dB is at 33Hz implying critical damping. So adding the 6dB room gain will bring the 33Hz to -6dB giving a very well balanced low end lift and balance imo.

 

[Hari Iyer]

I was looking at many BLH designs available at the The Frugal-Horns Site -- High Performance, Low Cost DIY Horn Designs website and found many interesting design. IMO Most BLH designs available are not horn speakers in true sense as the Mouth area is very small giving a low freq cutoff of around 200Hz to 250Hz. To have a low freq cut-off of say around 50Hz will require a mouth area as big as your room which is not practical.

Hold your breath, this is the size of the BLH if you want to get a decent 60Hz cut-off frequency designed by Nelson Pass.

https://www.passdiy.com/gallery/speakers/the-kleinhorn-part-1

These BLH behave more like a TL at low frequency and horn at frequencies higher than the cut-off frequency dictated by the mouth area. Also simulations are difficult and are mostly are through trial and error giving high chances of mis-match.

Given the above situation, i am finding the TQWT design to be a low risk candidate but still requires many hours of fine tuning before the first wood is cut. Maybe this weekend i will be able to finally conclude my design.

Finalizing my box with a TR of 10 and its going to be a ML-TQWT tuned at 57Hz.

 

[philipreji]

Hari, please post photos during the construction stage I too have a pair of the same drivers, unused yet..will follow your design if the end results are good

 

[Hari Iyer]

Hari, please post photos during the construction stage I too have a pair of the same drivers, unused yet..will follow your design if the end results are good

I am not sure if i will be able to post photos at the construction stage because the carpenter will be building the box in his workshop. I can share the design diagram on an email later if that works for you.

 

[philipreji]

I am not sure if i will be able to post photos at the construction stage because the carpenter will be building the box in his workshop. I can share the design diagram on an email later if that works for you.

Thanks, will PM you my email ID later.

Reji

 

[Hari Iyer]

Designing the TL enclosure has already started giving me nightmares now. I am not happy with the wrinkles of the unstuffed line in the ML-TQWT for the System and Terminus SPL response as below,

But the ML-SL gives a much better linear wrinkels for the higher harmonics (both System & Terminus SPL) as below,

I have done numerous amount of dirver offsets and ML-area and length adjust and the above is the best i could manage with these box geometry.

Also the group delay in the SL has much less spikes than in the TQWT.

This again confuses me about the geometry of the box. Should it be TQWT or SL. The only major advantage i can see with the TQWT is the looks and WAF factor but a much larger and taller box - Height 64"

The SL on the other hand can have a much smaller box (H-45") and can be built with a golden ratio cross-section. This will look like a traditional box with no fancy curves or contours but will be a well designed and engineered box imo.

Yet keeping my fingers crossed on this.

So this is how the final box is finally going to look

The ML terminus is in the bottom front and there will be an additional base for this speaker. Can now finally ask the carpenter to cut the wood and now i can have a good night sleep till the box is made.

 

[jls001]

Don't change your mind again

Keep away your laptop now

 

[philipreji]

So this is how the final box is finally going to look

Looks sleek and nice!

 

[Hari Iyer]

Don't change your mind again

Keep away your laptop now

Looks sleek and nice!

By looking at the box its a very simple box.

This is what Einstien has to say about simplicity - Einstein said, "Everything should be made as simple as possible, but no simpler."

This is what Nelson Pass has to say about simplicity -

Simplicity is a common element of the best and most subtle designs. It is preferred for purely aesthetic reasons, but also because fewer elements color the sound less, and lose less information. Many audiophiles, including myself, are willing to sacrifice other areas of performance to achieve the intimacy with the sound available through a simple circuit.

 

[philipreji]

Just being curious: why didn't you consider making the "back loaded horn type enclosure" recommended by Fostex for these drivers in its datasheet?

 

[Hari Iyer]

Just being curious: why didn't you consider making the "back loaded horn type enclosure" recommended by Fostex for these drivers in its datasheet?

IMO and others BLH designs are not true horn speakers and their low freq cut-off freq is somewhere around 200Hz to 250Hz by looking at the mouth area. Hence they too behave like a TL at lower freq and like a horn at higher freq. More over the depth of this published design is almost 18" and width around 11" and unfortunately i do not have floor space in my home for such a large speaker. More over simulation of a BL horn speaker is very difficult and most designs are usually trial and error.

Even the current design is not very small but i have the flexibility of simulating and adjusting to my requirement without much compromise on performance which i will not be able to do in any published designs.

 

[Hari Iyer]

Some more tweaking and simulation results in a different enclosure geometry. Now i will not disclose the geometry and loading but share only the simulation results. Will directly show the finished speaker once they are built so that FMs are kept guessing what i will finally built.

 

[Hari Iyer]

The wood is getting cut this week and below is the simulated response finalized by me.

This is the Characteristic Impedance at the Boundary conditions

 

[philipreji]

The finalised response looks quite good. Waiting for the hearing impressions

 

[Hari Iyer]

Would be getting only one box this weekend for testing and measurements. Will try and post images.

 

[Hari Iyer]

Finally deciding to go with varying stuffing density. Still am waiting for confirmation from FM Kapvin with his suggestions.

This is what the varying stuffing density gives me

As you can notice that the wrinkles are much controlled and the low end extension has not detoriated much. Also the TL acoustic impedance are very linear amplitude.

This enclosure behaves like an TL at the low frequency region and like an oscillating horn from the 3rd harmonic frequency onward.

I am expecting the enclosure probably this weekend if everthin works okay. It also coincides with my return back to Mumbai from Bangalore.

Am more than eager to review the outcome.