Positioning of ports and driver?

[Greg_M]

I have been using a speaker design program for my up coming speaker enclosure build. Although I am happy with the tuning and response figures it really says nothing about where in the enclosure to place the ports and driver.

Does it make a difference?

Front firing speaker, front firing ports .

How far to separate the driver from the ports. The driver and ports from the ends and sides of the cabinet. That sort of thing.

Is there reference material somewhere?

Also a question of material thickness....

I lean towards 1" mdf, but on a ported enclosure that is internally braced will it make much difference, other than the weight?
Would 3/4" work as well?


I am using WinISD

one JBL 2235H
8.657 cu ft
26.85 hz tune

Internal dimensions
17" w (driver/ports mounted on this side)
40" h
22" d

2 ports 4" ID 7.71" long giving .04 mach

This design dimension ratio(s) are sort of like the Klipsch KLF-10s I am using now except bigger.
The horn will be external (another project)


The design looks great on paper....


:confused:

[Robh3606]

You can put the ports where ever they make sense. I tend to put mine either on the back or on the side pointed away from where I sit. Helps keep any port noise inaudible. Just make sure the ports have room to breath.

Rob:)

[Woofer]

If you'd like to see an example of how inventive you can be, go here:

http://www.audioheritage.org/vbulletin/showthread.php?33697-JBL-4315A-with-custom-cabs-for-sale-UK (http://www.audioheritage.org/vbulletin/showthread.php?33697-JBL-4315A-with-custom-cabs-for-sale-UK)

It's an ingeniously 'modernised' 4315.

[Mr. Widget]

You can put the ports where ever they make sense. I tend to put mine either on the back or on the side pointed away from where I sit. Helps keep any port noise inaudible. Just make sure the ports have room to breath.

Rob:)True, but be aware that if you plan on pushing the speakers up against a side or rear wall you need a couple of times the port diameter distance to the walls.


Widget

[Mr. Widget]

Also a question of material thickness....

I lean towards 1" mdf, but on a ported enclosure that is internally braced will it make much difference, other than the weight?
Would 3/4" work as well?


I am using WinISD

one JBL 2235H
8.657 cu ft
26.85 hz tuneMaterials and thickness...

Materials, thickness, and bracing do matter. MDF is a soft and soggy material. It is well damped and therefore will not ring which is very good, but it is also rather lossy. Baltic birch plywood is stiff as hell and will not flex... I like to use a hybrid of these two materials, say a half inch layer of MDF bonded to a half inch layer of birch ply. I typically use 2x4s or 3/4" birch ply for bracing depending on the size of the box.


I think you will find the 2235H is better suited to a 5 cu ft cabinet tuned to 30Hz.


Widget

[Greg_M]

If you'd like to see an example of how inventive you can be, go here:

http://www.audioheritage.org/vbulletin/showthread.php?33697-JBL-4315A-with-custom-cabs-for-sale-UK (http://www.audioheritage.org/vbulletin/showthread.php?33697-JBL-4315A-with-custom-cabs-for-sale-UK)

It's an ingeniously 'modernised' 4315.


That's roughly the design I had in mind.
5 black bumpy sides and one "pretty wood" side. Haven't decided what "pretty wood" I will use, but I live a short way from "EdenSaw" which has a ton of exotic and domestic woods.

http://www.edensaw.com/MainSite/Store1/Store/CategoryHome/1092

The speakers will be in corners looking out at around 45° or so

Here's a couple screen dumps of the response curve. Not much difference in the two. The 8.657 cu ft has a very slight bump up (just above 40hz) before it starts to descend where the 8.245 cu ft does not.

I am leaning towards the 8.245?

Opinions?

[Greg_M]

Materials and thickness...

Materials, thickness, and bracing do matter. MDF is a soft and soggy material. It is well damped and therefore will not ring which is very good, but it is also rather lossy. Baltic birch plywood is stiff as hell and will not flex... I like to use a hybrid of these two materials, say a half inch layer of MDF bonded to a half inch layer of birch ply. I typically use 2x4s or 3/4" birch ply for bracing depending on the size of the box.


I think you will find the 2235H is better suited to a 5 cu ft cabinet tuned to 30Hz.


Widget


I used some Baltic Birch a while back when I made a Sunn 200S clone enclosure. It was good to work with but the piece I had was slightly curved which made making the enclosure square a bit of a challenge. The project took almost a year, but I was duplicating gathered measurements and back engineering the design plus gathering elusive parts.

[more10]

Driver and port placement is very important if you are using the speaker for midrange or building a large box. Since you are building a large box you will have standing waves between top and bottom, possibly ruining your response.

I built something very much like your box, Rebuilding a pair of srx (http://www.audioheritage.org/vbulletin/showthread.php?33259-Rebuilding-a-pair-of-srx). The box is basically a "-6 db bass shelf" as optimized by winisd. Placing the vent at the bottom allowed me to simulate in hornresp as an "offset driver". The box is a combination of BR and MLTL, the MLTL resonance filling in the dip in resonse curve around 100 Hz. Placing the driver in the middle of the box controls the first standing wave but the second at 300 Hz is impossible to avoid.

You can use Akabak if you have an old xp computer to simulate other driver and port placements.

If you want I can do a Hornresp simulation for you, since it is quite difficult to use before you learn it.

[Greg_M]

Driver and port placement is very important if you are using the speaker for midrange or building a large box. Since you are building a large box you will have standing waves between top and bottom, possibly ruining your response.

I built something very much like your box, Rebuilding a pair of srx (http://www.audioheritage.org/vbulletin/showthread.php?33259-Rebuilding-a-pair-of-srx). The box is basically a "-6 db bass shelf" as optimized by winisd. Placing the vent at the bottom allowed me to simulate in hornresp as an "offset driver". The box is a combination of BR and MLTL, the MLTL resonance filling in the dip in resonse curve around 100 Hz. Placing the driver in the middle of the box controls the first standing wave but the second at 300 Hz is impossible to avoid.

You can use Akabak if you have an old xp computer to simulate other driver and port placements.

If you want I can do a Hornresp simulation for you, since it is quite difficult to use before you learn it.


I do plan to do a cross over at 600hz.
It would be great to have your input using Akabak.

Thanks
G

[more10]

600 Hz will be difficult with such a large box.

I will make an Akabak script for you tonight, by exporting a Hornresp design. Then you can play around with the Akabak script placing the driver and port.

[more10]

Hornresp record (https://dl.dropbox.com/u/106944188/jbl2235br/jbl2235br.txt) (which you can import) and its simulation:

Input paramters (you will get this if importing the record)
https://dl.dropbox.com/u/106944188/jbl2235br/parameters.png

Schematic view of box:

https://dl.dropbox.com/u/106944188/jbl2235br/schematic.png

Port response:

https://dl.dropbox.com/u/106944188/jbl2235br/port response.png

The resonance at 180 Hz is minimized by changing length of L12 in loudspeaker wizard:

https://dl.dropbox.com/u/106944188/jbl2235br/wizard.png

Combined response (driver and port):

https://dl.dropbox.com/u/106944188/jbl2235br/combined response.png

Maximum spl:

https://dl.dropbox.com/u/106944188/jbl2235br/maximum spl.png

Red color indicates maximum excursion, black maximum power.

Impulse response:

https://dl.dropbox.com/u/106944188/jbl2235br/impulse.png

Simulation is done in corner (0.5 pi). The port peak at 350 Hz will be worse in real box, limiting this box configuration to about 200 Hz.

Hornresp cannot simulate other placements of the port, so I will make a simulation in Akabak later.

[Greg_M]

I'm totally lost

[more10]

Do you want me to remove the post or do some more explaining?

[Greg_M]

It's probably valuable to someone who knows the program, but I have no idea what all those boxes mean....

[Woofer]

Do you want me to remove the post or do some more explaining?

As someone that's just lurking in the shadows, please, don't remove it....
There's lots that I (for one), can learn from it.

[more10]

Well those boxes says that if you want to put the slot at the bottom of the box like this one:

http://www.audioheritage.org/vbulletin/attachment.php?attachmentid=56506&d=1343990664

you should place the driver in the center of the box. Also it says that there will be midrange frequencies coming out of the port, so the box only be used to about 250 Hz.

Hornresp is very limited when it comes to reflex boxes, you can only place the port at the end of the box. But it can simulate what happens inside the box (limited to one dimension). It is also useful because you can export an akabak script from such a design so you don't have to make one from start.

[Greg_M]

Perhaps I will build according to my original plan with the thought in mind that I may have to divide the enclosure into 2 parts for the addition of a mid-bass speaker should it not work out according to my ears.....?

[Ruediger]

that will give You enough bass. Somebody here in the forum should be able to give us a pointer to these plans.

If You want even lower bass You should consider a so-called B6 alignment. That is a 4th order box combined with a 2nd order elctronic filter. You will need an active crossover with provisions for that required 2nd order filter, such as the EV DX38 :) I can help You with the 4th order box.

But 1st try the 4430.

Regards
Ruediger

[Robh3606]

As Widget already pointed out the best general purpose box for a 2235 is 5 cubic ft tuned to 30 or so Hz.

The 4430 is tuned to high. The B6 box is the B380 at 4.5 cubic ft tuned to 26Hz.

Rob:)

[more10]

My modified Akabak (https://dl.dropbox.com/u/106944188/jbl2235br/jbl2235a.aks) script. I have tried to move the driver and port placements, but could not improve the response. This is the Akabak simulation (floor placement):

https://dl.dropbox.com/u/106944188/jbl2235br/akabak acoustic power.bmp

It will work in a large room, but in a small room there is too much output below 50 Hz or so. Midrange output through the port will be less of a problem with a smaller box.

Akabak will simulate only in one dimension, that is from top to bottom of box. Interference from driver to back of box must also be considered. The depth of the cabinet is 22" which is about 56 cm. 344/0.56 = full wave of 614Hz. The interfence is at a quarter wavelength, about 150 Hz. The box is a bit too deep.

The Akabak script has a few constants at the start, L12, L23, S1, ... Modify these constants and rerun the simulation. This drawing shows the constants in a box:

https://dl.dropbox.com/u/106944188/jbl2235br/Akabak constants.png

[Greg_M]

If you get a minute try a 6.7 cu-ft enclosure and see how your numbers look.

[more10]

If you give me the dimensions of the 6.7 cu-ft box (width, depth, height) and port (area, length) I will do a simulation tonight. Prefferably in cm.

Do you want traditional round port(s) or a slot port?

[Greg_M]

Internal dimensions
W 17"
D 23.25"
H 30"

Two round 4" ID ports 8.08" long ea

I also deducted some internal bracing to get to 6.7 cu-ft

Thanks!
G

[more10]

New akabak record (https://dl.dropbox.com/u/106944188/jbl2235br/jbl2235b.aks)

https://dl.dropbox.com/u/106944188/jbl2235br/akabak acoustic power 6_7 cu feet.bmp

Driver in middle of the box and center of ports 10 cm from bottom. I believe it is tuned a bit too high.

Compared to the larger box you have moved the problem frequency from 350 Hz to 450 Hz. If you make the box even lower, and maybe wider, you will push this problem even higher in frequency.

I have also simulated moving the port to the middle of the box, you will get a slightly larger dip/peak at 450 Hz.

[Greg_M]

Neat!
What would you recommend the tuning at?

Thanks for the quick answer!

[more10]

You better trust Mr Widget and the others for tuning recommendations. I have never constructed with this driver.

Do some more WinISD simulations, and when you are happy I'll do another Akabak simulation for a "second opinion". Don't forget to check port velocity in WinISD, if too high you will hear the port. When you do the velocity thing, add enough power to reach xmax somewhere, then you can plot port air velocity.

I still believe that the box is too deep. If you make it wider, and put the midrange in a equally wide box on top, you will push down the baffle step frequency as well. If you can get the baffle step frequency to the same frequency where you plan to divide, you can use that in you filter design.

[Greg_M]

I do appreciate the help.....
Thanks
G

[Greg_M]

I'm still alive.
Has it really been 7 years?
Still got the woofers and my boxes (shells).
A lot has happened since the beginning of this project.
Built a house in New Mexico and moved in. Everything else got put on hold during that time.
Just about ready to pick this project up again.
Glad I was able to find it.
G

[RMC]

Hi Greg,

Nice to see you're back in business.

I haven't followed the thread from the beginning, but here's a few notes that should help you decide.

RE "you should place the driver in the center of the box." (post # 16)

Weems (note 1) recommends that one does not place the woofer at the center of the box to reduce promoting standing waves in the cabinet. A little off center is recommended. Mounting a woofer a little higher or lower than box center should do the off center trick, plus it doesn't cost a penny more and its the same work baffle hole cutting wise, so why not if it helps. Though you say having your boxes (shells).

As for vent placement its usually not critical, more important is to avoid extreme vent proportions for slot/rectangular shape vent, and making sure the port has sufficient space to "breathe" inside the cab.

JBL's rule of thumb for port size (dia.) is one-third the woofer's diameter, so a 15" driver means 5" vent if only one is used. Two 4" dia. vents are equivalent to a 5.7" vent so that's OK. Two 3" ones = 4.25" so a little short. But then vent size also depends on the amount of power the speaker will be fed with. Practicalities usually determine if you'll use one, two or more (e.g. space on baffle or cost).

With regards to vent placement vs woofer location, Dickason (note 2) has looked into the driver/vent coupling/interactions with simulations for a 12" driver using 6", 4", 3" and double 3" vents. His conclusion:

"In the case of the 6" and 4" diameter vents, very close proximity to the driver seems to cause the least amount of disturbance. The 3" vent showed less problems when located at a distance from the driver. The dual 3" ports, Fig. 2.31 produce fewer problems when the two ports are separated by a reasonable distance."

In addition note the duo 3" vents were placed at the same distance from the woofer (so, for double 3" at some distance from each other but both being the same distance from woofer).

I think what Dickason refers to with "disturbance" may be vent air turbulence. However there's another aspect mentioned by Weems when vents are not 3"+ from woofer. He says that a port too close to driver is more susceptible to reflected sound from upper bass or mid frequencies. So you'll have to play this by ear...

Hopefully the above will be of some help to you. Regards,

Richard

(1) David B. Weems, Designing, Building, and Testing Your Own Speaker System, 4 th Edition, P.26 & al, 54.
(2) Vance Dickason, The Loudspeaker Design Cookbook, 5th Edition, P. 55

[Greg_M]

I really appreciate your thoughts on the topic.
Thanks!

I decided before I go any further to make sure the Thiele/Small parameters are correct on these "recones", so I bought a Dayton Audio DATS V3 Test System gizmo.
It's on the way and should be here at the end of the week.
Downloaded WinISD again, so I'm getting ready to go.

It's good to have projects.

Thanks again.
G

[Greg_M]

Here's where I am today.
After the big move, the enclosures wound up in the garage. That's 1-7/8" Bloodwood from South America, on the fronts and 1" particle board on the rest of the sides.
Backs and bracing pieces are sitting on top.
As you can see, I've got 5 sides glued together and some of the bracing inside. I just need to get out the belt sander and make all corners/edges flush and the route a radius on them.
Before we moved I still had the machine shop's CNC mills to play with so I cut out the speaker openings and drilled holes on the insides of the fronts for hold down screws.
So here we are.
85875

[RMC]

Hi Greg,

Good idea to get the Dayton gizmo so that you know exactly what kind of TS parameters you're dealing with. Sometimes folks get surprises, not always pleasant ones, after some "modification" to a driver. I remember one here that I helped he did same as you: he got the DATS 2 at the time, measured the woofers he bought used, when he saw the results the adjectives about the seller started flying, bastard, etc. If I recall right he abandoned the drivers. Knowing ahead of time, before further holes are cut in speaker baffle like in your case, is way better since one may be able to salvage a project with some work around (vary vol. and/or tuning).

Btw just learned from you DATS 3 is finally released, been waiting for it since V. 2 has been around for a while. Last time I checked it was still at version 2... Now V3 is on my near future shopping list.

You indicate having downloaded Win ISD again. Which version? 2002, 2004 or 2016, hopefully the last one. Note that, contrary to many other programs, Win ISD uses a box loss assumption of QL 10 which is NOT standard practice, often too optimistic, and this affects low frequency response modeled. The recognized standard is QL 7 so its better that you change in the software the default 10 for 7. Easy, here's how:

When you're in BOX tab look at bottom of screen and click on ADVANCED and a small window shows QL 10 and others, click on 10 and another small window shows 10 in blue, type 7 instead and close the window. Now you have QL7 for the moment. If you save a project with QL7 it should stay at 7 when you reopen that project (check it). HOWEVER, Win ISD always defaults back to QL10 when you open the software for a new project, so everytime you must not forget when in BOX tab to go to ADVANCED and change again QL to 7 instead of 10, a little pain in the ass...

Your box pic shows the woofer cut out is lower than middle of the boxes so that should help you reduce standing waves in the cabs. If I were you I would avoid making any other baffle holes until you get a good set of TS from DATS 3 and do some Win ISD simulations. Regards,

Richard

[Ian Mackenzie]

Hi Greg,

That looks like a nice project for your new home.

I am not clear if you plan to use these enclosures inverted (from the pic you posted)?.

My impression was your earlier plan was to use the woofer up to 600 hertz.

Your plans for the mid/high frequency driver of your choice will dictate how you orient the enclosures.
Ideally the woofer should not be far from the mid range driver or horn.

Depending on the size of the horn you may wish to mount it on top of the enclosure. Or if the enclosure has sufficient depth mount the horn inside the enclosure. This will impact on placement of the ports on the front baffle.

The 5 cu ft 3 tuning at 30 hertz is going to give you a smooth non drooping response with a -3 frequency around 35 hertz and will be -6 db about 32 hertz.

If you have a larger than the recommended 5 cu ft 3 enclosure up to around 8 cu ft3 and tune according to your software the response will droop down but will extend lower. The port will be more active at lower bass frequencies due to the lower tuning frequency. The ports will play a more audibly aggressive role in the bass response on most program material with the recommended 5 cu ft 3 enclosure tuning.

Placing the enclosure close to a wall boundary or a corner will compensate for the dropping response to some extent. Its entirely subjective.

Depending on how your room reacts with the system and the enclosure placement you will find out fairly quickly what enclosure size and tuning you prefer just by listening to the amount of bass you hear subjectively. The tuning differences are quite obvious with the 2235H.

[Greg_M]

I could put the woofer on top or on bottom at this point. It's just a matter of where to put the feet.
I had planned to put the woofer on the bottom, as they are sitting now.
When I last worked this project I bought a pair of JBL 2380A horns and 2446 drivers.
They are still waiting too.

I had the woofers reconed at a JBL fix it place in Lynnwood, WA so hopefully they are close to factory specs. We'll see.
I also have a pair of Altec 421As that I got for some reason, can't remember why now.

[Greg_M]

Hi Greg,

Good idea to get the Dayton gizmo so that you know exactly what kind of TS parameters you're dealing with. Sometimes folks get surprises, not always pleasant ones, after some "modification" to a driver. I remember one here that I helped he did same as you: he got the DATS 2 at the time, measured the woofers he bought used, when he saw the results the adjectives about the seller started flying, bastard, etc. If I recall right he abandoned the drivers. Knowing ahead of time, before further holes are cut in speaker baffle like in your case, is way better since one may be able to salvage a project with some work around (vary vol. and/or tuning).

Btw just learned from you DATS 3 is finally released, been waiting for it since V. 2 has been around for a while. Last time I checked it was still at version 2... Now V3 is on my near future shopping list.

You indicate having downloaded Win ISD again. Which version? 2002, 2004 or 2016, hopefully the last one. Note that, contrary to many other programs, Win ISD uses a box loss assumption of QL 10 which is NOT standard practice, often too optimistic, and this affects low frequency response modeled. The recognized standard is QL 7 so its better that you change in the software the default 10 for 7. Easy, here's how:

When you're in BOX tab look at bottom of screen and click on ADVANCED and a small window shows QL 10 and others, click on 10 and another small window shows 10 in blue, type 7 instead and close the window. Now you have QL7 for the moment. If you save a project with QL7 it should stay at 7 when you reopen that project (check it). HOWEVER, Win ISD always defaults back to QL10 when you open the software for a new project, so everytime you must not forget when in BOX tab to go to ADVANCED and change again QL to 7 instead of 10, a little pain in the ass...

Your box pic shows the woofer cut out is lower than middle of the boxes so that should help you reduce standing waves in the cabs. If I were you I would avoid making any other baffle holes until you get a good set of TS from DATS 3 and do some Win ISD simulations. Regards,

Richard


I'll check out the version today.
Is there a way to get version 7 somewhere and permanently install it?

[Ian Mackenzie]

I could put the woofer on top or on bottom at this point. It's just a matter of where to put the feet.
I had planned to put the woofer on the bottom, as they are sitting now.
When I last worked this project I bought a pair of JBL 2380A horns and 2446 drivers.
They are still waiting too.

I had the woofers reconed at a JBL fix it place in Lynnwood, WA so hopefully they are close to factory specs. We'll see.
I also have a pair of Altec 421As that I got for some reason, can't remember why now.

Well it sounds like your having fun.

When your ready send me a Pm and l will help you with with the crossover network. I have measurements of both the woofer and the horn / compression driver and can punch out either a passive to active solution.

That timber on the front panel is very nice. You can’t buy affordable loudspeakers with that finish these days.

[Greg_M]

The pretty wood came from Edensaw, by Port Townsend, WA.
http://www.edensaw.com/MainSite/Store1/Store/CategoryHome/1092

I used to live about a half hour from there.
It makes my mouth water to go to that place. They have exotic and domestic woods from all over the planet.
I got that chunk of wood in one big piece. Enough to do both box faces and some left over for a turntable base. (another project).
They do a lot of boat building up there in PT.

[Greg_M]

My WinISD program version is 0.7.0.950
I was able to change QL 10 to QL 7 as suggested and when restarted it came up as 7.

One question I have is what is 1st Port Resonance ?
By changing port size it changes. What should I be aiming for there as a value?

[RMC]

Hi Greg,

Your Win ISD version is the 2016 one and that's fine (most recent).

Even if QL7 came back that time, do check it periodically as QL10 WILL come back as default value (except in a saved project).

RE first port resonance:

From the help files of Win ISD ( open INFO at top, choose HELP, in the list of links then click Designing your box):

"Here, you can decide how many and of what shape you want your vents to be on your box. The ports, of course, serve to "tune" your box to the frequency that you have chosen while you were designing your box in the Plot Window. The size and -3dB (cutoff) frequency of your box determine the size of the ports.

1st port resonance shows frequency of first "organ-pipe" resonance. This helps you to estimate where port becomes undesired radiator.

Very important here is to make your ports large enough, so that peak air velocity won't exceed 5% of sound velocity in air. That is, about 17 m/s assuming normal environmental conditions. You can check it by using "Rear port - Air velocity" graph, by first setting simulation power to anticipated maximum power, see "signal"-tab. In this case, Pe is 125 Watts. "

In practice you don't really choose a specific value of 1rst port resonance, but that sophistication isn't a bad idea re woofer crossover frequency:

For example, I modeled quickly 2235H in 5 cu.ft. with Fb 30 Hz and the 1 rst resonances were: 1 X 5" tube gave 840.5 hz, and 2 X 4" tubes gave 561 hz. The 5" one having a higher 1 rst resonance would give increased XO frequency flexibility, without having port resonance in the way... Gotta go. Regards,

Richard

[Greg_M]

I got the DATS V3 test gizmo and did a test of my re-coned JBL 2235H (once I figured out how to run it). I also tested a Klipsch 10" K-1036.
Here is a dump from the V3 software.
I'll enter these values in to WinISD and see if it makes a difference.


* This data was exported from the Dayton Audio Test System: DATS
*
* Piston Diameter = 342.9 mm
* f(s)= 27.59 Hz
* R(e)= 6.107 Ohms
* Z(max)= 148.5 Ohms
* Q(ms)= 8.941
* Q(es)= 0.3834
* Q(ts)= 0.3676
* V(as)= 187.1 liters (6.606 cubic feet)
* L(e)= 1.6 mH
* n(0)= 0.9769 %
* SPL= 92 1W/1m
* M(ms)= 213.1 grams
* C(ms)= 0.156 mm/N
* BL= 24.26
* K(r)= 0.05604
* X(r)= 0.5535
* K(i)= 0.0164
* X(i)= 0.7138
*

[RMC]

Greg,

For comparison purpose the following are the parameters given in the official JBL table:

MODEL FS QTS QMS QES VAS EFF PE XMAX RE LE SD BL MMS FLUX
2235H 20 0.25 2.5 0.28 458.7 1.3 150 8.38 6 1.2 0.090 20.5 155 1.2

Some show quite sizeable differences between your actual numbers and those from JBL, doesn't necessarily mean the driver is worthless, but simply that its different from an original one, therefore may require some box adjustments.

Qts is pretty high which normally leads to larger appropriate box, however Vas is pretty low which normally leads to a smaller box size, so the latter compensates somewhat for the former (these two are appropriate box size influencers).

Double check your DATS measurement procedure just to be sure you got this right. Did DATS make you do a driver "conditioning" period before measurements? Usually driver parameters are taken after a couple of hours of driver suspension "loosening" for example ("Thiele/Small parameters are measured after a 2 hour exercise period...")

As I said before I haven't followed your project from the beginning but maybe your recones are not original ones with such differences? Or no driver conditioning was done? Regards,

Richard

[Greg_M]

Greg,
Some show quite sizeable differences between your actual numbers and those from JBL, doesn't necessarily mean the driver is worthless, but simply that its different from an original one, therefore may require some box adjustments.
Richard

Yes it is a re-coned driver. It was done by Morgan Sound in Lynnwood WA.



Greg,
Did DATS make you do a driver "conditioning" period before measurements? Usually driver parameters are taken after a couple of hours of driver suspension "loosening" for example ("Thiele/Small parameters are measured after a 2 hour exercise period...")
Richard

No DATS did not suggest conditioning, but I did do the measurement twice because I goofed it up the first time.
The measurements were slightly different the second time so probably there is something to your suggestion.



Greg,
As I said before I haven't followed your project from the beginning but maybe your recones are not original ones with such differences? Regards,
Richard

I bought the speakers used and took them in and had them re-coned myself.
When I opened the box today the receipt was still in there. They've just been sitting on the shelf since 2013. They are virgins.


I wonder what frequency they should be run at to condition them?

[Greg_M]

I have it hooked up at it is vibrating away at 40 cps.
It's not very loud (probably a good thing) but it's as loud as it will go.

[Ian Mackenzie]

It’s basically a new unused driver which can take forty hours of continuous use to break in.(as published by Jbl).

Mount it in a vertical position if you can so the cone mass is not weighing on the spider and surround.

[RMC]

Greg,

Who did the recones isn't really the issue here, but rather what recone kit was used, original JBL or aftermarket? Since you still have the receipt maybe its written on it the recone brand or part number? Even on the back of the cones there could be some indications like a part number?

Richard

[Greg_M]

C8R2235 JBL Recone kit $214

[Greg_M]

Greg,
Some show quite sizeable differences between your actual numbers and those from JBL.
Richard

Probably because it's a 2225H not a 2235H.
I just discovered that my memory was playing tricks on me.
:o:

It's different but in a good direction.

[RMC]

Greg,

At least you have genuine JBL cone kits which should be better than cheap aftermarket ones, yes different but in a good direction as you say.

I think the key for you will be giving these drivers a good break in period, then measure parameters again with DATS 3. At that point the numbers you will get are the deck of cards you'll have to play with in Win ISD. It shouldn't be bad.

Qts should drop and Vas increase somewhat with the workout. What seems a little more puzzling though is the cone mass measured at 213 gr vs an original 2235H cone having 155 gr. Even more so when Eargle mentions "Normal production variations in cone mass should be negligible, ..." (Loudspeaker Handbook, P.290). Here its a notable 37% more. Maybe test method or procedures involved may explain this? Check.

Since I'm planning on getting in the near future a DATS 3 what's your take on it up to now, easy user friendly to use or a maze with some pain in the...?

Richard

[Greg_M]

I used the DATS V3 to drive the woofer overnight at about 30hz. So it ran from 14:30 yesterday until about 09:00 this morning. Ran the test again and nothing had changed.

Did you know about the little ring weight that goes in the 2235 cone (maybe under the dust cover)? It's to add mass (obviously) so it can reach that low Fs. That might be where the added weight comes from?
Since the recone guy used the 2235 recone kit, it might be in there helping to push the Fs of the 2225 down from 40hz to it's current state.
Just guessing about all this.

Soon as I get one of the boxes done being braced, I'll mount up one of the 2225s and let it run for a while at 30hz or so under some power and see if anything changes.

As far as DATS V3, it's pretty easy to use. You'll need something like Blu-Tack to stick some weight to the cone for the second part of the test. When I did the 10" Klipsch woofer the 75 grams of the Blu-Tack was enough by itself. When I did the 2225 I had to use the Blu-Tack and a tape rule to get enough weight to make DATS happy. There a few YouTube videos that cover it.

Using the measurements from the DATS gizmo, I fed WinISD all the new data. and finally got a call out for the ports (for the 6.7144 cu ft box) of 2ea 3" tubes 4.15" long. Put in the figures in to "Port Length Calculator" and got 2ea 3" tubes 4.18" long, then "Box Port Design" and got 2ea 3" tubes 4.54" long.

I'll wait until I run these speakers in some more before I actually cut holes in the back of the boxes. Meanwhile I can just use the box to hold the woofer vertically and run it for a while.

I think I'm making progress.

[grumpy]

For anyone trying to follow this thread: if a 2225 has its soft parts replaced with a JBL factory 2235H recone kit, it is no longer a 2225, but a proper 2235, unless the mass ring is left out, in which case it is a 2234.

[RMC]

Greg,

Yes I knew about the mass ring, this thing is mentioned on the forum regularly since many folks have or want 2235. Barry here has shown twice recently with his measurements that the gain in lower Fs with the mass ring on is about 3-4 hz if I recall right. Not a huge gain bass wise.

RE "Ran the test again and nothing had changed." Not even a bit yet is surprising. Good idea to continue the break in, preferably as long as it takes to see numbers change at least a little and then some. The drivers waited 7 years so they can wait a few days more, in workout, before getting installed for sound.

I'm not overly concerned about Fs at this point. But more about Qts and Vas for your box shell size (6.7 cu.ft. is that your calculated net or gross internal volume? Bracing, driver, etc. accounted for?). If the driver suspension system (surround and spyder) can start getting more compliant ("loosening") then those two parameters should start to move.

Keep in mind what Grumpy wrote to avoid confusion. Sure you're making progress, designing boxes is science based and you have to go through many steps, including the nitty-gritty stuff. You may feel entangled in details now but its certainly an investment in knowledge/learning for this and the next pair...

Thanks for the feedback on your DATS 3 experience up to now.

RE port size and length, don't be surprised if you get some small differences between one software and another. The Editors of Win ISD themselves wrote a 20 page tech paper on this using 11 different softwares to explain why results vary from one to another. Shortly, it depends on the assumptions made (e.g. temperature, speed of sound, etc.) and equations used.

For your boxes 2 X 3" vents is bare minimum specially if you'll feed those with a fair amount of watts. I'd be leaning more towards 2 X 4" or even 1 X 5" if practical (remember the 1rst port resonance thing?). Regards,

Richard

[Greg_M]

The cu ft box measurement is gross, not subtracting 1x1" corner bracing or the speaker itself. The box is 30" tall, 17" wide 22.75" deep inside.
I'll be pushing the speakers with a McIntosh MC275 so only 75 watts per channel.

I probably should mention that I live on the Continental Divide at 7,350'. The air is quite thin and very low in humidity. Maybe this is altering some of the figures?
I moved here to do astrophotography. Still building the observatory (yet another project) but it's too cold to work outside so I'm doing inside projects now.

Also, I am no professional using the DATS system so I could be screwing something up in its use. If anyone has a check list or suggestions I sure wouldn't take offense at a little critiquing in its use.

If anyone would look at my WinISD figures, I'd be glad to make my project file available.

If all this is for naught I can still use the KLF-20s.
:)

[Greg_M]

For anyone trying to follow this thread: if a 2225 has its soft parts replaced with a JBL factory 2235H recone kit, it is no longer a 2225, but a proper 2235, unless the mass ring is left out, in which case it is a 2234.

I'm remembering now a conversation with the speaker reconing guy (because I couldn't find any 2235s that were not crazy expensive) where he told me about reconing the 2225s with 2235 kits saying what you are saying. He said the frames are the same it's just the soft parts that differ.

The 2225s are quite a bit more plentiful as they were/are used a lot in theaters.

I had forgotten all this until you mentioned it.
Thanks
G

[RMC]

Greg,

Looking at your Win ISD figures now is a little premature pending you get more final numbers after the drivers' good workout. Also too early to say if all this is for naught. Lets stay positive and keep hope, before switching to salvage mode.

Those 1X1" are joints reinforcement mainly. Not really cabinet panel bracing which you will need with such box size. You may still use these.

Your 6.7 cu. ft. gross box is basically a 5 cu.ft. net box with almost 25% overvolume for losses, bracing, drivers, vent, etc. Seems acceptable.

The minimum panel bracing I would use, in view of the low 75 W /channel, is 2X2" lenghts for a front to back panel brace midway and same for a side to side panel brace midway or so to reduce panel vibrations, while taking little space inside the box. So a cross type bracing job, simple to do and it won't take away much box volume. You need to keep bracing in mind specially on your front panel layout before making more holes.

The air volume displaced by a 15" driver is given at 0.2 cu.ft. (6L) by JBL, useful to know for gross/net box calculations.

The altitude you live at may have some effect but never lived that high so not really familiar with the impact & details. e.g. I know in my snowblower manuals there's references to using it at higher altitude, I think 4,ooo ft.+ (carb air/fuel mix special adjustment?), but never looked closely at altitude aspect since I live at low altitude. Probably more for Colorado folks who also get a fair amount of snow.

Then atmospheric pressure must be different too? Doesn't seem to impair mustache growing though, that should make you smile. Nice.

NM and Colorado members here living on higher grounds may be in a better position than me to help you with the altitude aspect. Regards,

Richard

[Greg_M]

After doing some subtracting I came up with 6.3529 cu ft. net

[RMC]

what have you subtracted? 6.36 cu.ft. net seems pretty high to me when all in

[Ian Mackenzie]

Hi Greg,

Richard has you on track for a nice bass reflex system.

For years l used to do lots of calculations to work out the net volume. I felt like l knew every inch of the design but still the result was not always quite right.

That said once you try it out it’s fairly obvious if the tuning is off. The bass will obviously sound restrained or boomy.

[Greg_M]

what have you subtracted? 6.36 cu.ft. net seems pretty high to me when all in

0.2 cu ft for the spkr and 279 cu in for the corner bracing.

[RMC]

Greg,

The typical 2235 box is 5 cu.ft. net tuned to 30 hz. Doesn't mean others are forbidden but then you need to model such to see the expected result with different tunings and volumes. Yes in salvage mode volume can be modified somewhat.

Richard

Mod to volume: to reduce, putting pieces of wood or books even bricks (latter too heavy) to take excess space in cab. To increase, much thicker glass wool or similar to create virtual volume.

[RMC]

And you haven't counted good bracing yet...nor vent and the other drivers in the cab

Horns & drivers space taken .05 to 0.1 cubic foot (JBL)

[RMC]

When all is factored in volume calculation and if still too large then you get say a cubic foot of cheap old books and put them at the bottom of the box, therefore taking some excess volume.

That's an example of box salvage mode.

[Greg_M]

no other drivers in the boxes.
if I use 2x2s that would be another 279 cu in.

No idea how to figure port displacement.

We're down to 6.1915 cu ft not counting the ports.

[RMC]

Greg,

RE Port Displacement

Volume of a cylinder: V= Pi X radius squared X height or length of cylinder


Start with area of a circle: Pi X radius of circle squared, so 3.1416 X the circle's radius squared, so you get square inches

From that result multiply by vent length in inches which will give you cubic inches

e.g. Assuming your down to 6 cu.ft. then put a cubic foot of books in the speaker box and you got 5 cu.ft. net, the beauty of books or wood pieces is you can play with volume by removing some or adding some if need be.

Or you may decide to stick to a 6 cu.ft. box, then have to model it in Win ISD to see if the results can be made to your satisfaction while playing with tuning frequency.

[Greg_M]

Ok
Down to 6.145 cu ft

[RMC]

then you decide what net volume you want to use 5 or 6 cu.ft., read post # 63

[Ian Mackenzie]

Hi Greg,

If you are not familiar with calculating the bracing here is an example using 2 x 2 lumber

This bracing represents shelf bracing that is typical of the enclosure bracing used in the vintage systems. Having built a number of enclosures for these drivers over the past 30 years l can attest to the structural integrity. Adding too much bracing is bad and can cause more problems than it solves.

Qty 10 x 17 inch braces for sides, top and bottom. 170 x 4 = 680
Qty 5 x 22.75 inch braces for front and rear 113.74 x 4 = 455

Total 1135 cu inches 3
Cu Ft3 0.656

Add
Driver 0.2
Ports 0.145
Total 1.00 cu ft3

Gross enclosure volume 6.7 cu ft3
Less 1.00 cu ft3

Sub total 5.70 cu ft 3

Add
Volume gained with 1 inch fibre glass on 5 sides
0.57 cu ft3 (10%)

Net Volume 6.27

QL = 7 (assuming no box leaks and a well braced box. Note a QL = 5 is typical for large enclosures due to panel flexing, in this instance there enclosure is over volumed from 5-10%. So you are in the ballpark. By doing some modelling of the port tuning and comparing this to the actual measured port tuning you can then start to fine tuning the design. )

You could make 6 cu ft 3 your target volume and see how it goes. The tuning Fb will be a but lower than the 5 cu ft 3 but you might like it. If you don't like it you can always go for smaller target volume by asking some solid mass inside the enclosure.

Edit

Attached in top image is a snap shot of the jbl 2235H spec driver in Leap Enclosure Shop
Blue is the 6.27 cu ft 3 enclosure tuned to 28 hertz.
The f3 is 29.75 hertz and is -6 dB at 26 hertz.

Red is the 5 cu ft3 enclosure tuned to 30 hertz.
The F3 is 32.5 hertz and is -6 dB at 28 hertz.

These results are models only and demonstrate the differences in one tuning to another.

As you can see there is no free lunch but if you want low bass the 6.27 cu ft3 enclosure is an interesting design. In the bottom image the original Jbl 4343 was 5.5 cu ft3 in red was tuned empirically for the best overall subjective results. The f3 is 31 hertz with a 29.5 hertz tuning. The blue curve is the recommended 5cu ft3 enclosure tuning.

The scale is 2 dB per division so you can see the effect of different tunings and enclosure size.

What is interesting is that the human ear is quite sensitive to low frequency amplitude variations. Given your larger than recommended enclosure size you can work out with some experimentation what works best for you in your listening environment.

The science as Richard refers to it is well documented in terms of working out an enclosure and tuning from scratch. Given JBL recommendations l felt it was appropriate to
Illustration the effect of some different tunings and enclosure volumes to give you an idea of what to expect.

The graph scale is 2 dB per division.

[Greg_M]

Thank you for your time and effort Ian Mackenzie.

I was worrying about making stupid arithmetic errors, so this morning I did a simple spread sheet.
There are several variables like internal dimensions and bracing dimensions etc. (Note two different sizes of corner bracing)
Here's where I am as of this morning (while the glue is drying from yesterday).

85919

[Greg_M]

Hi Greg,

Add
Volume gained with 1 inch fibre glass on 5 sides
0.57 cu ft3 (10%)

Net Volume 6.27


I think it's extremely interesting that volume can be gained with insulation.
If there is a formula or way to estimate this, I'd love to include it in my little spreadsheet.
Thanks again

[Ian Mackenzie]

http://www.cieri.net/Documenti/JBL/Documenti%20tecnici/JBL%20-%20Enclosure%20Guide.pdf




If you refer to the links below you will pretty well find everything you need to get a good results.

Jbl use and recommend fibre glass with guidelines on where and how much to use. If you decide to do something else your on your own. My recommendation is to follow the Jbl recommendations. Attempting to come up with your design using your own methods could easily result in a situation where you end up being not on the playing field.

You can comfortably go up or down one cubic foot in volume from the recommendation of 5 cu feet as long as you adhere to the tuning process outlined by Jbl. The trade off is enclosure size versus system LF output and other performance criteria outlined by Jbl. There are articles on the lossy properties of fibre glass. Just google.


http://www.lansingheritage.org/images/jbl/plans/1979-manual/page21.jpg


http://www.lansingheritage.org/images/jbl/plans/1979-manual/page29.jpg


http://www.lansingheritage.org/images/jbl/plans/1979-manual/page30.jpg


http://www.lansingheritage.org/images/jbl/plans/1979-manual/page31.jpg


http://www.cieri.net/Documenti/JBL/Documenti%20tecnici/JBL%20-%20Enclosure%20Guide.pdf

[Greg_M]

Interesting reading.
Thank you

[Greg_M]

Greg,

RE Port Displacement

Volume of a cylinder: V= Pi X radius squared X height or length of cylinder


Start with area of a circle: Pi X radius of circle squared, so 3.1416 X the circle's radius squared, so you get square inches

From that result multiply by vent length in inches which will give you cubic inches

e.g. Assuming your down to 6 cu.ft. then put a cubic foot of books in the speaker box and you got 5 cu.ft. net, the beauty of books or wood pieces is you can play with volume by removing some or adding some if need be.

Or you may decide to stick to a 6 cu.ft. box, then have to model it in Win ISD to see if the results can be made to your satisfaction while playing with tuning frequency.

I was expecting different calculations other than a cylinder since it's hollow.
Easy enough.

[Ian Mackenzie]

No worries.

Have fun and enjoy your project.

[RMC]

RE "I was expecting different calculations other than a cylinder since it's hollow. Easy enough."

Since its a cylinder volume calculation, i.e. calculation of space taken or air displaced in the box by the cylinder, then it goes without saying, but maybe I should have said, its the cylinder's outside diameter that's considered here, so the radius used isn't based on tube's inside diameter, but rather on outside diameter.

When an effective port calculation is made i.e. for a device actively passing air in and out when in use, the relevant radius to use would be based on tube's inside dia. However, here we're only concerned about space occupied by the tube inside the cab (btw excluding front panel thickness), regardless if tube size is appropriate or not acoustically speaking. Hope this provides the nuances.

[Ian Mackenzie]

That’s a good point.

Back in the day used 4 inch cardboard tube.

If the port was made of 3/4 inch mdf there would be an error with the external volume versus internal.

I would go bigger than 3.5 inch internal diameter. If the rear of the port is 4 inches or less from the rear put a bend in it. Jbl did this as routine in a number of systems. It’s easy to do with 4 inch pvc pipe fittings.

Likewise keep the ports at least one diameter from the sidewalls. The effect messes with the port function and you will end up having to compensate with the port length (as Jbl did with the 4343 and 4345 monitors.

[Greg_M]

Likewise keep the ports at least one diameter from the sidewalls.

Does this include one diameter minimum from the top or bottom?

[Horn Fanatic]

Does this include one diameter minimum from the top or bottom?

Greetings -

Large ports will provide more articulate low frequency at the point of resonance. In other words, a tight bass. Large meaning, 5-8 inches in diameter depending on the volume of the cabinet. On a pair of reproduction ALTEC Lansing Capistrano I built, I used two 6" diameter ducts exiting the bottom panel. The down side; the larger diameter port, the longer it must be. The objective it to keep the vent velocity (mach number ) down. At the point of resonance, it doesn't matter where the port is as long as it is not obstructed. A tube port can even be outside of the enclosure, as it does not have to be inside. When calculating port parameters, consider the tube port as a separate physical element in the circuit. For example; if a circuit board layout has been finalized, but later determined that a choke must be added, then make the circuit board larger to accommodate it. The cabinet volume must be increase to accommodate a tube port.

H.F.

[Horn Fanatic]

I think it's extremely interesting that volume can be gained with insulation.
If there is a formula or way to estimate this, I'd love to include it in my little spreadsheet.
Thanks again

Greetings, again -

The typical volume gained by adding damping material would be 1.4 times the gross volume of the enclosure. Personally, I use damping material sparingly, as too much will suck the live out of the cone. The impression that a highly dampened enclosure adds bass is a misconception. Damping material does two things; attenuate high frequencies, and decrease the velocity of sound at low frequencies. Essentially, damping material fools the speaker into behaving as though it is in a larger enclosure. Adding bass? More like allowing the speaker to produce the low frequencies it already does, unfettered by a small enclosure.

H.F.

[RMC]

Hi Horn Fanatic,

RE "The typical volume gained by adding damping material would be 1.4 times the gross volume of the enclosure."

Just to avoid having too high expectations. Its not evident that one can add that much virtual box volume with damping material. It sure was the belief among experts and audiophiles for some time, however:

"Previously it was thought that the maximum possible increase in effective volume was in the range of 1.4, or 40%, but Leach (1989) provided a more accurate analysis indicating that the maximum ratio is 1.31. In normal practice most loudspeaker engineers observe an increase of perhaps 1.2." (Or 20%) John Eargle, Loudspeaker Handbook, P. 63, Regards,

Richard

[Greg_M]

What an expensive book.
I thought of buying it for reference, but they want a ton of money for it.

[Greg_M]

Meanwhile.....
While I'm building on the cabinets, I have one of the speakers being "conditioned".
Right to left.....
Astron 12v power supply
Tenma Audio Generator.
Boss car amp (usually for sub woofers etc)
It's been exercising the speaker for almost 24hrs now at 20hz.
I think I'll run it at the next higher step (28hz) for a while.
I pushed it at 20hz until it started "clacking" and then backed off enough to make it stop.
The JBL metal tag fell off the back of the spkr. Easy enough to reglue.

85950

[RMC]

Unfortunately good engineering books are generally expensive. These don't have a large audience nor printing run, making cost per unit much higher. That kind of manual won't appear on best-sellers lists. If you can find a good used copy at reasonable price it may well be worth it.

JBL's John Eargle wrote a number of respected audio books: Microphone Handbook, Handbook of Recording Engineering, Electroacoustical Reference Data, Loudspeaker Handbook, Handbook of Sound System Design, etc., but the last two cover a good part of it. Those are often used as University level references for Audio Engineering courses. Which explains why sometimes it gets heavy math wise but most of it is ok.

The Sound System Design book was updated in 2002 by John Eargle and Chris Foreman with a new title: JBL Audio Engineering for Sound Reinforcement. The Loudspeaker Handbook was also updated in 2003. A fellow having both editions of it reviewed the new one and said it was mostly the same text as first edition except with more recent driver model examples and technologies.

You can get for free SOME of the Sound System Design book's stuff in JBL's own "Sound System Design Reference Manual", 1999 also from Eargle. Just Google JBL's manual and get a pdf file of it from their site.

Its unlikely there will be any further updates to these books since he passed away some years ago (RIP).

[Greg_M]

I tried a few used places and of course ebay and all those that I could find are even more expensive than the new ones off Amazon?
I did see that the library in Las Cruces has a copy so maybe I can get the Datil library to bring it in for me.

[RMC]

Another interesting one, and much cheaper, is Vance Dickason, The Loudspeaker Design Cookbook.

I have 5th edition but its at the 7th edition now I think. More home theater audio stuff was added which I'm not into. Price on my edition was $35. so quite affordable, a good used one could be a bargain.

Robert Bullock, Bullock on Boxes, 1991 also interesting

David B. Weems, Designing, Building and Testing your own speker system, 4th edition, 1997, says $20. on mime (best one for beginner)

All the above depend on knowledge level you're at or want to be or can stand. What are you looking for?

[Greg_M]

I'd like a book that I can learn in to, but is something I can understand now.
Usually I have to open the book and read a page.

[Greg_M]

My first 2225H is done being conditioned and I did a DATS test on it again.
This is after 24hrs at 20hz and 24hrs at 32hz

This is what the software had to say...

* This data was exported from the Dayton Audio Test System: DATS
*
* Piston Diameter = 342.9 mm
* f(s)= 19.65 Hz
* R(e)= 6.838 Ohms
* Z(max)= 179.2 Ohms
* Q(ms)= 7.299
* Q(es)= 0.2896
* Q(ts)= 0.2785
* V(as)= 272.8 liters (9.635 cubic feet)
* L(e)= 1.715 mH
* n(0)= 0.6815 %
* SPL= 90.44 1W/1m
* M(ms)= 288 grams
* C(ms)= 0.228 mm/N
* BL= 28.98
* K(r)= 0.04122
* X(r)= 0.5916
* K(i)= 0.02001
* X(i)= 0.7072

PS
I did put my Fluke meter on it and got about 6.2 ohms.

[RMC]

Your first 2235 done not 2225 as Grumpy indicated...

I updated a little the info on Weems' book, year and price.

Since its cheap, then I would suggest you start with Weems book to get a good understanding of the concepts and how they work, its pretty well made and I still refer to it sometimes for specific things. These type of books are not all made the same way with same info. Somethings you'll find in one and other things in another. They're not at the same level either.

So start with Weems', then move on to Dickason's Cookbook, and when you have pockets full of money you get Eargle's Loudspeaker and Sound System Design handbooks.

Back later about your latest parameters...

[RMC]

Compared to last measurements, this time box relevant parameters (Fs, Qts, Vas) have improved notably, though Vas still somewhat short, but its been shown that Vas is less critical (error has less impact) than some others. Qms is still too high to have Qts on par. RE went up but you measured 6.2

You lost some efficiency/sensitivity in this second set. Cone mass is even further than it was before (also from spec) as well as Bl increasing again. So there remains a number of mysterious things here.

I don't have a ready made explanation for all the deteriorations. For example one may risk for efficiency that ever higher cone mass has reduced it, however motor strenght (BL) is shown to have increased also...

No (efficiency) calculation does consider Fs, Vas and Qes so it'll be interesting to see what the following says:

Try Dats up to first para. list without doing the Vas measurement, then input the DATS para. you have at that point in Win ISD Driver Editor (para. tab) and let "Auto calculate unknowns" do the calculations for the missing ones and see how they compare with todays para. list as measured by Dats.

I would input the numbers slowly and look as you go to the ISD calculated numbers, and also to changes made by the auto calculator. It may give you some clues.

[Greg_M]

Could be I am doing something wrong operating the software also.
It would be interesting to have an original 2235 to measure as reference.

On the enclosure front, I am at 10,702 cu in (6.2 cu ft).
If you count the magical 20% from insulation it is 12,842.5 cu in (7.432 cu ft.) however I am not using the latter in WinISD.
Maybe I should?

Looks like one 4" vent 4"± long at this point.

[RMC]

I'm assuming DATS wise that you are following the instructions and procedures from Dayton Audio.

RE having an original 2235 to measure as reference, well I've posted (post #41) the parameters of the original 2235H. This should be used as reference or for comparison purposes. There can be small differences from unit to unit but it should generally look like that (not big differences).

On the enclosure front, I personnaly see no interest in going with a too large enclosure (6.2-7.4 cu.ft), except in a specific purpose which I won't get into here. I don't necessarily agree with, nor endorse, some of the things mentioned in this thread.

I've already indicated the typical 2235H enclosure is 5 cu.ft. tuned to 30 hz. The other one I know is JBL's B380 which is a little smaller at 4.5 cu.ft. and uses the 2235H. Usage of larger volumes than these aren't from me, so I'll let their proponent(s) deal with the specifics about those and the issues involved.

There's nothing wrong with going a different route than an already proven design. However one has to know why making that choice, what are the consequences, and how to mitigate the issues. Important to know.

A 5" vs 4" vent is about the same work and little more cost, so why not go for the 5" and have peace of mind, instead of using a just sufficient 4" one. I'm aware JBL's B380 has a 4 1/8" vent. B380 is 1983. Since then vent size recommendations have increased, and increased again with the newer very high power drivers (e.g. 2KW music).

In four more recent JBL Speaker building documents I checked (not counting 1983 Audio article), two recommend a 5 cu.ft. box for 2235H, Fb 30 hz (other two don't mention driver). As for vent size for 5 cu.ft./Fb 30 hz box, two mention a 5" tube, one a 6" tube and one 2 X 4" tubes. None for a single 4" dia. vent.

One day or another guys always end up putting more power in their boxes than initially thought, at the very least to test the beast... You may regret a smaller vent down the road for you or if you ever sell the cabs.

The "magic" of Fiberglass insulation isn't that magic btw. You need a lot, and my mention of it was in a box salvage mode context, which doesn't really apply to you, having larger than optimal volume already. So if you need to do something it may well be to reduce it, not enlarge it. Read item # 20 on pic, its from JBL.

Driver limitations in a box have to be recognized and dealt with accordingly. The sky isn't the limit.

[RMC]

Sorry forgot the pic...

85981

[Greg_M]

If you believe WinISD, there is nothing to be gained by changing the vent size from 4" to 5"
My enclosure net size is 10,702.1 cubic inches and if using the insulation boosts it by 20%, it is 12,842.5 cubic inches.

Tuning this enclosure to 28hz will give me pretty much flat response to 40hz and a slight bump up after that. If the insulation trick works it will be a slightly bigger bump but not much.
I have nothing to lose if it does not work and not much to gain if it does.

Why worry?
A 4-string 20-fret bass (like a standard Fender Precision) goes from E1 (41.203 Hz, second E below middle C) to Eb5 (622.25 Hz, second Eb above middle C).
https://www.quora.com/Whats-the-range-of-a-bass-guitar?share=1

The lowest note of a double bass is an E1 (on standard four-string basses) at approximately 41 Hz or a C1 (≈33 Hz), or sometimes B0 (≈31 Hz), when five strings are used.
https://en.wikipedia.org/wiki/Double_bass


This all hinges on WinISD being accurate.
I have uploaded my WinISD file, so if you want to play with the figures you certainly can.
It is here
http://www.datilcam.com/GregsBox.wpr

Let me know if you find some figures that are in error and I will run them on my version of WinISD.
Thanks
G

[RMC]

When dealing with vent size one must remember speaker sound level impact on this, therefore keep an eye on power input data. Many softwares use 1 watt level for speaker modeling . At that level vent size doesn't matter much. But as power increases markedly then it does matter.

I have played with this software power level thing in the past and saw the impact on vent size, it was eye opening. The software suggested vent size did vary based on input power number fed in the program. I don't remember if it was with Winspeakerz or Win ISD or both.

For the last few years I've been using mostly Winspeakez, though I still have Win ISD and use it once in a while only because its driver database is pretty poor compared to Winspeakerz.

Some softwares allow you to change driver input power level to model at higher or lower SPL, and some use as default the driver's power capacity number for suggested vent. Can't remember which power number ISD uses and if it can be modified since I use it less often.

In any case, at the end of the day its still your project, not mine, so you decide what vent suits you best. Amen.

[Greg_M]

What method did you use to obtain the TL parameters?


Not sure what TL parameters are.

[Greg_M]

I think it is time to put the speakers back on the shelf and terminate this project again.
This just isn't fun.

[Ian Mackenzie]

For anyone reading this thread l can confirm from my DATSv3 measurements the issue with the data is a user error.

When setting up the VAS measurements the piston diameter must be measured correctly. With a foam surround the piston diameter needs include half or more of the surround.

I used three or four 36 gram lead sinkers and got a VAS of 459 LT.

My T/L data was close the Jbl data. The driver must be rigidly clamped with clear space under the driver.