DIY-project: JBL Kebnekaise S4365

[pos]

2216nd variants as well as many current gen 1.5" drivers and horns are readily available tho.
You can also put Truextent Be diaphragms into a (possibly used) 1.5"/4" driver and get quite close to a 476Be.

[JeffW]

These luxury drivers are not for sale. You can not buy them from JBL.

Took over a year and a screw up on Harman's part, but I actually did buy a pair of 1501FEs from them. But this was back when they had drivers for sale by system on their website, there's not even any way to try to buy the drivers now.

[Ruediger]

It is over my head to do that interpolation.

Hi,

Here are the interpolations. Run them in a simulator software before cutting wood.

Linear interpolation in Thiele Table between alignments #2 and #3:

f3 = 57.9 Hz
fb = 45.4 Hz
Vb = 38.9 liter


We can as well make the driver fit the table. The driver's Qt depends on the amplifier's output impedance (equation 70 in the Thiele paper). The DC resistance of the speaker cables (back and forth!) and of crossover inductors which are in series with the driver gets lumped into that impedance as well. The net amplifier output impedance is ( 8 Ohms / amp's damping factor ). A typical value is 200. This yields ( 8 Ohms / 200 = 0.04 Ohms ).

1 / Qt = ( 1 / Qm ) + ( 1 / Qe ) * [ Re / ( Rg + Re ) ]

Alignment #3

Qt = .259 (our goal)
Re = 5.42 Ohms
Qm = 4.25
Qe = 0.25
Rg = 0.559 Ohms result: amp's output impedance + cables + inductors

f3 / fs = 1.77 -> f3 = 52.2 Hz
f3 / fb = 1.25 -> fb = 41.8 Hz
Vas / Vab = 4.46 ->Vab = 48.9 liter


Keele's Table:

Vb = 54.4 liter
f3 = 56.56 Hz
fb = 44.76 Hz

Keele, force 50 liter:

f3 = 61.6 Hz
fb = 47.3 Hz
R = 0.38 dB (R = Ripple)

Keele, force 50 Hz

Vb = 75.9 liter
fb = 41.3 Hz
R = -0.89 dB (R = Ripple)

There is a discrepancy between the interpolated values and the values from Keele's table.

Thou shalt build prototypes!

Ruediger

[Ian Mackenzie]

You may wish to acquire Bassbox simulator software

But all models are based on assumptions

The Qb3 is a maximally flat alignment that in fact does not exist once the loudspeaker is in the real environment

For this reason experienced engineers like GT account for typical placement in a room and then design the box tuning to take suit

As a result we have the "banana.

In the room the bass sound crisp and smooth

TheQB3 in comparison can sound fat and bloated

[Ruediger]

<snip>
The Qb3 is a maximally flat alignment that in fact does not exist once the loudspeaker is in the real environment
<snip>

Please explain

Ruediger

[Ian Mackenzie]

I will post a Bassbox simulation soon

[Ian Mackenzie]

Okay

I will try and give you context of using a particular jbl driver designed specifically for a jbl system.

How that was incorporated in the JBL 4365 and relate that to your project

Firstly, The whole TL alignment thing is a continuum of parameters based on electrical filter theory

However, all the simulators assume the driver is in free space ie not in a acoustic space with wall/floor boundaries.

Secondly the simulator does not consider the impact of placing the driver on abaffle of particular dimensions where the baffle will behave as a reflector up to a certain frequency and then a diffuser.

In the context of the JBL 4365 Jerry or who ever designed a driver with a stiff diaphragm and characteristics that would make it suitable for GT to adopt a systems engineered approach to a two way system

They were interested in a woofer that would work in a 2 way system and by manipulating the box tuning in a real room and the Voltage drive to the woofer from the passive crossover this would give a smooth response with reasonable bass extension

How is this possible given your simulator outcomes?

As you have discovered this driver when plugged into a simulator yield a small box and high tuning in the 40's.

In the real world however such a box is not going to deliver any degree of dynamic authority in the bass because it can't below the port tuning frequency

So we make the box bigger and tune the box lower in the region of 30 hertz

The box is now around 150 litres.
In the attached Bassbox simulator graphic you can see two curves

One is the driver as if no room gain existed and the other with addition of room gain

This is ball park of course

The key is the acoustic output in the 30 Hz zone and overall response flatness

In the box itself around 3 Db of sensitivity is shaved of the woofer midband response by the crossover to ensure a smooth response on the baffle

[Ian Mackenzie]

The up shot of this is you have two choices being either un assisted or assisted tuning with boost at the box tuning frequency fb.

Depending on the driver the assisted box volume may be smaller than the unassisted box volume

Assisted alignment are not easy to implement in practice without test equipment and experience

The other consideration the baffle step response of the woofer on your final baffle?

This could be as low as the 150 Hz region where the output of the woofer falls off.

Jbl of course tale this into consideration when they design and voice the system and determine the optimum voltage drives for the crossover network

You may therefore consider a dbx drive rack to enable setup and manipulate the final response of the woofer for your diy project

Good luck

[bubbleboy76]

Thanks Ian!

I have a DBX Driverack 4820 and measuring equipment, so I will try to get it tuned in including the room-gain, in the end. For 5 mic-positions placed in ear-height in a "5 on a dice" manner around the listening-position, averaged. Like G.T has suggested.

[Ian Mackenzie]

Good idea

The sims are fun to play with but don't rely on them in isolation

You want woofer motor delivering real output from the port in the 30/35 region

Going further with critical insights the human ear is acutely sensitive to bass amplitude flatness, extension and low 2nd distortion according to AES publications

If you can get the system outside in your driveway again the garage door it will be much easier to see what the box is doing without too much interference from room modes

Once you get base line eq saved then getting it right inside wil much easier

[bubbleboy76]

hlaari knows how to pack woofers safely

[JeffW]

Here's how Harman packs the same driver. There was nothing in the box besides the driver and that sheet of bubble wrap, you can see the outline of the magnet in the side of the box.

[Mr. Widget]

Here's how Harman packs the same driver. There was nothing in the box besides the driver and that sheet of bubble wrap, you can see the outline of the magnet in the side of the box.

Not to defend the outrageously poorly packed shipments that have left various warehouses contracted by Harman, they have moved their operations again and terminated their agreement with some of the past warehouse companies that they were subcontracting with.

There have been quite a number of shipping horror stories where Harman products have met terrible ends due to extremely irresponsible packing. That said when I received my DD67000 woofers last year they were very well packaged in custom packing with proper die cut inserts. I hope this is the new normal.

Of course Hlaari's example is not going to be matched by Harman anytime soon.


Widget

[hlaari]

I hope at all parts you got from me will work perfectly

My s9900 project does not go as good as I was hoping for, I have only got few parts of the s9900 horns
this have take a way to long time to get the parts I need for the project so I might sell the parts and the S9900 enclosures

there is possible at I can get a used pair of K2-S9900 for good price or I will go for the 4348 monitors from Japan

hlaari knows how to pack woofers safely

[bubbleboy76]

Here's how Harman packs the same driver. There was nothing in the box besides the driver and that sheet of bubble wrap, you can see the outline of the magnet in the side of the box.

Are you kidding me?!

When buying 2216nd:s from speakerexchange, how is the packing then? Good?