PDA

View Full Version : 2245 port sizes



dennis j leisz
09-12-2003, 08:08 AM
I'm starting a new winter speaker project. It gets cold here in Minnesota. I am designing an 8 cu. ft. enclosure for a 2245 . I can't get any of the design programs to correlate on port size lenght etc. I need some suggestions. I plan to cross over at 300 hz to a 10in. I would like to tune the cabinet to 28hz and use (3 )., four inch I.D ports. What do your programs tell you the length of each port to be. I use this info as a starting point and use a test tone with a scope to fine tune the lenght. Any other suggestions would be great. I looked at the Solen site for help on the Ultimate experience. I have never used the 2245 before , looks like fun.

4313B
09-12-2003, 08:14 AM
Well, the old JBL method of tuning comes up with 11.125" as the required duct length.

Bass Box 6 Pro comes up with 10.96" as the required duct length with "normal fill", 11.58" with minimal fill and 12.26" with "no fill".

And yes, the 2245H is fun :)

boputnam
09-12-2003, 10:01 AM
Hey...

Here's a 4345 monitor plans - Link (http://www.audioheritage.org/vbulletin/showthread.php?s=&threadid=314) the first page of which shows, and gives the port dimensions for the 4345 cabinet as 4-1/8" diam by 8-1/4" duct (x 3).

Fun, fun, fun... :nutz:

4313B
09-12-2003, 11:24 AM
Well, if you are going with stock stuff then you might also want to know that the B460 used three 4.125" ID ports each having a 14.875" duct.

I think you're right on though going for a 28 Hz target tuning frequency. The 4345 might be a bit too high (plus you don't have a big 5.4 mH inductor in your signal path do you?) and the B460 might be a bit too low unassisted.

Use the port length I suggested and prepare to cut some length off as you measure the actual tuning frequency :)

BTW - What lengths are your programs giving you?

Guido
09-12-2003, 02:06 PM
Originally posted by Giskard

BTW - What lengths are your programs giving you? [/B]


My Program state 10 qubic foot with three 4 inch ID Ports each 8,11 inch duct as ideal for the 2245H.

Tuning frequency is 28 HZ.
- 3 dB at 36 Hz

4313B
09-12-2003, 02:15 PM
For ten cubic feet tuned to 28 Hz I get three 4" ID ports with 7.71" ducts from BB6P with "normal" fill, 8.21" ducts with "minimal" fill, and 8.75" with no fill. The JBL formula says three 4" ID with 7.875" ducts :)

Now do you see why I advocate just tuning by the proper methods :)

dennis j leisz
09-16-2003, 12:44 PM
Thankyou for your input and suggestions. What I am basicaly doing is designing and building an updated 4345 system. This new speaker will sit in my new dedicated sound room. No wife acceptance factor in this space! I am using the 2245 , a 2123h crossed over at 300hz, a 2450h on the 2380a for the mids and a 2405 for the high frequencies. I am designing a 2nd order Linkwitz-Riley crossover. I have a pair of 077 s sitting in my shop that I may use instead of the 2405s. Any opinions? One other question. 9.2 cu.ft cabinet gross volume with 3- 4 1/8 " dia X 8 1/4 long ports= 28hz. Will 2 ports work and if so what dia. and lenght. As soon as I'm a bit further along I'll post some progress photos. I will drive these speakers with the McIntosh MC1201 monoblocks.

4313B
09-16-2003, 01:02 PM
Originally posted by dennis j leisz
Will 2 ports work and if so what dia. and lenght.
I use dual 6" ID ports for 2242 and 2245 subs but you can use dual 5.25" ID ports with 10" ducts in your 9.2 cubic foot volume.

Mr. Widget
09-16-2003, 01:40 PM
"I have a pair of 077 s sitting in my shop that I may use instead of the 2405s. Any opinions?"

They are identical, don't worry about the 8 ohm, 16 ohm thing. All 077s and 2405s use the same diaphragm. You can find several posts on this subject here on the forum. I use the 077 clear acrylic horn plugs in my 2405Hs because they look cooler.;)

John Y.
09-16-2003, 02:22 PM
Originally posted by Giskard
I use dual 6" ID ports for 2242 and 2245 subs but you can use dual 5.25" ID ports with 10" ducts in your 9.2 cubic foot volume.

Giskard,

Why do you suppose JBL went to one (rather large) port on their latest pro subwoofs, the 4645C and 4641?

These enclosures are the 4518A, I believe. Do you suppose the enclosures are identical for the two systems mentioned above (same tuning)? Do you further suppose a 2245 sub would go in nicely, without mods, into this enclosure?

John Y.

4313B
09-16-2003, 02:46 PM
Originally posted by John Y.
Giskard,

Why do you suppose JBL went to one (rather large) port on their latest pro subwoofs, the 4645C and 4641?

These enclosures are the 4518A, I believe. Do you suppose the enclosures are identical for the two systems mentioned above (same tuning)? Do you further suppose a 2245 sub would go in nicely, without mods, into this enclosure?

John Y.
I suppose because one single port is cheaper and easier and they have the baffle room to use it.

They are tuned the same. Fb is 25 Hz. The original 4518A was tuned to 30 Hz for the 2245H. If you put a 2245H into the 4645C enclosure it will perform roughly the same as a B460 without the BX63 boost. The original S1S used the 2245H and was tuned to 30 Hz as well. JBL has since moved to the 2242H and will be retuning the S1S to 25 Hz if they haven't already done so.

Ian Mackenzie
08-22-2005, 11:23 PM
Inspifred by Dennis Leisz original question I thought it would be fun to revisit this old thread being a 2245H and 4345 fan.

Having built a clone of the JBL 4345 I calculated the internal Vb from the following data with the assistance of JBL (courtesy Greg Timbers)

Vol from cabinet internal dimensions = 10.1 cuft 3
Subtract
Vol of Dog Box = 0.71 cuft 3
Vol of Horn and driver = 0.1 cuft 3
Vol of Ports = 0.2 cuft 3

Gross internal volume = 8.99 cuft 3

Subtract

Vol of woofer =.0.3 cuft 3
Vol of cleats = 0.4 cuft3

Add effect of 3/4 inch fibreglass fill = Vol of woofer and cleats
(courtesy of Greg Timbers)

Net VB is 9.00 cuft3

Port dimensions 4 1/8 diameter x 8 1/4 inches long (3) ports

Tuning FB = 29.7 hertz (Bass Box simulation-normal fill QL 7)

The impedance curve suggests a Fb of 26 hertz but in practise the Fb in terms of real movement closer to 28.0 hertz from my observations.

In the original design the tuning was somewhat empirically done, based on the results of listening tests. One can change the tuning frequency of that system by +/- 2 Hz and still get a very presentable low frequency curve. It really came down to the preferred sound character.


Ian

Ian Mackenzie
08-23-2005, 01:21 AM
The sound character of the bass is different in bi amp mode and I will do some experimenting with small variations in Fb and advise the results.

Ian

Lancer
08-23-2005, 05:23 AM
The sound character of the bass is different in bi amp mode and I will do some experimenting with small variations in Fb and advise the results.

Once again, reference Keele's paper - Sensitivity of Thiele's Vented Loudspeaker Enclosure Alignments to Parameter Variations.

It's going to be tough for you to advise the results as they are somewhat subjective. Sure, you will affect power handling and can advise on that objectively but with a driver that robust it shouldn't really matter for home use. Some people think a 26 Hz tuning doesn't have enough "fill", other's think it's fine. Some prefer it tuned a bit higher. Any room/boundary reinforcement will have an effect.

You can play around with QL too, the stock 4345 should be somewhere around 10 as opposed to the "typical" 7.

In the original design the tuning was somewhat empirically done, based on the results of listening tests. One can change the tuning frequency of that system by +/- 2 Hz and still get a very presentable low frequency curve. It really came down to the preferred sound character.

Ian Mackenzie
08-23-2005, 05:52 AM
Hey your up early!

I agree but I was planning only a +-1 hertz variation just to see what gives.

Once again, reference Keele's paper - Sensitivity of Thiele's Vented Loudspeaker Enclosure Alignments to Parameter Variations.

That is a great read, I recall measuring about 10-11 QL, the problem is the 2245 is like a big mic and a devil to measure.

I figure the system was either tuned for bi amp as best result or tweaked for good result in passive/active with that 0.5 ohm coil and 90 uf capacitor.

Odd they did not make it 10 cuft 3 and be done with it. The previous iteration I had was around 10.5 cuft tuned to 27.2 hz with Keele's equations

Vb= 15Vas ( Qts) 2.87

I suppose of I was serious I would buy a woofer tester and start with clean data.

Ian

spkrman57
08-23-2005, 07:07 AM
I wonder why the 2245 exhibits more problems with room placement(ground planes and such) vs the 2242 that seems less affected by its surroundings?


Just curious!

Ron

Lancer
08-23-2005, 07:37 AM
Hey your up early!

I went to bed? :rotfl:

Odd they did not make it 10 cuft 3 and be done with it. The previous iteration I had was around 10.5 cuft tuned to 27.2 hz with Keele's equations

I have tales to tell...

I wonder why the 2245 exhibits more problems with room placement(ground planes and such) vs the 2242 that seems less affected by its surroundings?

The answer is actually in the Product Brochure. The 2242 is designed to be augmented by boundary reinforcement while the 2245 was designed to yield rock solid response within it's bandwidth without boundary reinforcement.

spkrman57
08-23-2005, 07:44 AM
Why is there so much difference in response(ie: 10 cu ft vs 10.5 cu ft)?

I would think there would be more variances with the room loading and speaker position than with a difference of a 1/2 cu ft in cabinet size.

Ron

Ian Mackenzie
08-23-2005, 04:38 PM
Hey your up early!

I went to bed? :rotfl:

Odd they did not make it 10 cuft 3 and be done with it. The previous iteration I had was around 10.5 cuft tuned to 27.2 hz with Keele's equations

I have tales to tell...

.

Well do tell.....:hmm:

I went over all this stuff ages ago with Witold Waldron, the inventer of Calsod and owner of Audiosoft..and he laughted and said it all depended on the assumptions made with a particular model........this was well before I found out it was empirically tuned. I mean you can doodle till the cows come home but what the Calsod guy did explain was the ear is very susceptable to variations in LF flatness and rolloff.........and group delay.

By running a series of simulations and by deduction and I therefore reasoned that the 9 cuft 3 tuning at 28 hz was to attain a flat response with maximum extension ie 32 hertz and reasonable group delay.(transient response).
They also had to deal with a 0.5 ohm 5.4 hm choke in series with the woofer.
Plugging that into the simulation is not that pretty...0.7 db ripple in 10cu ft 3 with an fb of 30, the 9 cu ft 3 tuning at 28 herts contols the ripple more, being only .2 db. Perhaps this was borne out in the empirical trials JBL did.......

This is all by the by when one actually builds and starts tuning a real box as there is QL and all the other attendant issues that can make the model less predictive.

There are also alignments by others of the JBL fold incl Gary Margulius that offer tunings on a make to fit box size.

Ian

Lancer
08-23-2005, 08:19 PM
There are also alignments by others of the JBL fold incl Gary Margulius that offer tunings on a make to fit box size.

Yeah, check out Drew's legacy stuff.

http://www.drewdaniels.com/Audiotech.htm

Joe Etrick (Barath Acoustics years and years ago) assisted on the spreadsheet.

Ian Mackenzie
01-25-2007, 08:28 PM
Yes I know,

I am back looking at this again!!

But there is a reason. I wanted to fine tune for my 0.2 ohm 5.4 mh inductor. Up until now I have used 0.33 ohm resister in series to mirror the 0.5 ohm JBL choke in the stock 4345 vol enclosure (9 cu ft3 net)

As I recall Greg suggested that the box Fb might need to be put up a notch.

I tried this and chopped an inch off the ports.

Its highly subjective, initially I thought it was better, more punch but after playing a number of cd's I came to the conclusion it was not as good as my original tuning per the stock 4345 8 1/4 inch ports. I then realised after looking at a BassBox simulation that the ideal box should reduce in volume and the fb should go up accordingly.

Rather than alter the volume at this point I decided to go the other and keep the volume as is and reduce the Fb, so I blocked one of the ports.
Very interesting result, somewhat overdamped.

The thing as mentioned earlier is everyone simulator has adifferent set of assumptions. Drew's calculator gives 11.27 ft cu3 , Fb 26.4 with a QTS of 0.28. I use that value to allow for the added DCR of my choke 0.2 ohms.

Bassbox gives as sub 8 cuft volume without ripple? Drews simulator says there will be ripple at this volume. I just wonder if there are varying interpretations of what maximum flat response means.

My feeling is that I will need to manipulate both the volume and fB to nail this.

Ian

Ian Mackenzie
01-25-2007, 10:36 PM
Page 46 is very intuitive

http://www.drewdaniels.com/BINDER.pdf

John W
01-26-2007, 08:50 AM
Ian, have you measured the actual Fb of the box with the different port configurations? Also, are you sure the t/s parameters of your pair of 2245s match the spreadsheets?

Ian Mackenzie
01-26-2007, 09:41 AM
This is probably of more interest to would be/are JBL 4345 builders.

Bi amping these monitors is the most popular mode of operation. Its is intuitive however to understand the contraints placed on the system in internal passive mode and how this effects performance.

Those interested in using this monitor in passive mode with a high quality amplifer may find this information useful.

I contest that unless your active crossover is of the highest quality you will loose resolution and transparency and this will mitigate the benefits of an advanced passive crossover. The full range amplifier should be a high current type, preferably with class A operation at least at low power levels.

After carefully review the JBL data and some design details of the 4345 I was able to model the 2245 in the stock 4345 enclosure reasonably accurately in Bassbox 5.1

There are a couple of things that are not entirley obvious. The placement of the ports adjacent to the side wall effects the effective length as does the use of mutiple ports, hence the actual physical length is shorter than the designed Lv for a Fb of 28 hertz. The dcr of the 5.4 choke used in the stock system is about 0.5 ohms. The manner in which the biamp switch is arranged causes additonal Dcr in the order of 0.3-4 ohms. The stock enclosure has QL of 10.

The effect of the total dcr added to the system creates a modified QTs of 0.29- 0.3 (JBL 2245H QT =0.27) . The Ql of 10 and the modified QTs gives rise to ripple in the woofer passband, starting at about 0.1 db 200 hz, 0.2 db at 98 and a maximum of 0.4 db in the 60 hz zone. I would contest this results in a lack of definition and clouds the midrange based on my own evaluation using the stock 3145 networks. While there is a known effect on the midrange due to interaction of the motional impediance of the drivers, the ripple in the woofer passband contributes to a less than ideal performance in internal passive mode.

To compensate for the modified Qts in the system design would call for a larger enclosure and a different tuning frequency.

Over the weekend I explored some interesting approaches to improving the bass and hence the midrange performance in passive mode.

The choke in my diy 3145 crossover has a DCR of 0.2 ohms. My choke is hard wired without any biamp switch to ensure minimal additional losses.

The lower DCR choke results in a modified QTs of 0.28 on the 2245H. This is a more ideal value to minimise ripple in the woofer passband. The Bassbox simulations showed no ripple for this alignment.

Using Table X of Small's T/L data (from Bullock on Boxes 4th order QB3 alignments) for a value of Qts for 0.28. The Fb given by H =1.3747 = 27.5 hertz, Alpha =3.4971= Vb 8.29 cu ft3 net and F3/fs =1.6826 = 33.6 hertz.

I implemented these updates into my cloned 4345 system for assessment. This required a reduction of the net enclosure volume and an adjustment to the port length.

The bass balance with the midrange which is a fairly subjective aspect of loudspeaker performance was similar to the stock alignment. The clouded almost woofey character was far less pronouned, there being less overhang and better definition. I tried a slightly lower Fb of 26 hertz as a comparison and found this made the low bass more pronounced but the mid and upper bass was less aggessive.

I will continue to assess these variations on the tuning and update this thread in due course.

Ian

grumpy
01-26-2007, 10:32 AM
Hi Ian,

You've probably noted this elsewhere, but it might be useful, toward understanding your
tuning/development process in this thread, to briefly describe your listening environment
(room, placement including any stands, and listening position), as this would be integral
to what you are hearing/reporting. Cheers, -grumpy

John W
01-26-2007, 10:58 AM
I would be interested to hear what actual Fb and box volume you think sounds best.

I guess my point is that even if you follow the nice set of 4345 plans exactly, there will still be slight variations in the final volume of the cabinet. You may have a thicker front baffle, a little more stuffing, slightly wider bracing, etc. Also, I guess I am a little suspicious about the 9.0 cubic feet in the published specs.

Also, how much of a constant is the QTs of a 2445? Once the driver is initially broke in, does this value change much throughout the life of the driver, or with environmental factors like temperature, altitude and humidity. Maybe the actual QTs of your drivers started off at .29 and now they have been changed to .32.

Ian Mackenzie
01-26-2007, 04:14 PM
Hi Ian,

You've probably noted this elsewhere, but it might be useful, toward understanding your
tuning/development process in this thread, to briefly describe your listening environment
(room, placement including any stands, and listening position), as this would be integral
to what you are hearing/reporting. Cheers, -grumpy

Grumpy,

I have them in a relatively small at the moment( by Yank standards..its not a Red Neck barn!) that measures about 4 x 5 metres. (apartment).

The loudspeakers are 2.2 metres apart (from the horns) , about 0.25 metes from the rear wall and 0.45 metres from the side walls. They were elevated 0.2 metres off the floor on wooden stands but for safety reasons I have them on 50 mm alumuminum cone tip toes while I work on them. This provides a degree of de-coupling from the floor. The listening position 2.8 metres from each speaker and is 1 metre from the rear wall and within arms reach of the Bar tap! They are toed in at a slight angle, not much. I find this arrangement works well in terms of imaging.

grumpy
01-26-2007, 04:23 PM
Excellent. Thank you kindly! -grumpy

Ian Mackenzie
01-26-2007, 05:12 PM
I would be interested to hear what actual Fb and box volume you think sounds best.

I guess my point is that even if you follow the nice set of 4345 plans exactly, there will still be slight variations in the final volume of the cabinet. You may have a thicker front baffle, a little more stuffing, slightly wider bracing, etc. Also, I guess I am a little suspicious about the 9.0 cubic feet in the published specs.

Also, how much of a constant is the QTs of a 2445? Once the driver is initially broke in, does this value change much throughout the life of the driver, or with environmental factors like temperature, altitude and humidity. Maybe the actual QTs of your drivers started off at .29 and now they have been changed to .32.

John,

I am coming around to the opinion that there is no best Fb and Vb and this is explained in Drew's binder pdf ..the server is down at the moment. There are two extremes on paper...maximumly flat ....overdamped and extended bass...underdamped. Somewhere in between will be happy medium to suit the individual application.

The 4345 box tuning was somewhat empirically, based on listening tests.
If you follow the plan (in the tech reference area) you can't go wrong, just do everything they say. If you stray from the plan..look out. The vol is respect to the net volume after accounting for the woofer and cleats from the gross volume in the woofer chamber and the effect of the fibreglass fill which adds real volume. I spent hours labouring on this and this was Greg's explanation. I have checked it and its correct.

The design was tuned on the real world, things were done a certain way (like the port length) with the actual box to get a desired result.

'Both of my 2245H are relatively new (less than 5 years old). To the best of my knowledge the parameters (electrical/mechanical) don't change, the 2245 is such a robust assembly its unlikely to vary much.


The point of what I have been doing here is highlighting and addressing the woofer performance in internal crossover mode. If you are going to bi amp don't loose sleep over it. But the coil (the choke ) the capacitors shunting the LP filter on the woofer effect the bass, noteably at FB but as I have indicated they can influence the whole bass range. Basically the way its set up for internal passive mode is a best fit without having to re tune the box but its obviously not as good as biamping where that coil anc capacitor are out of the picture and there is no motional impediance interaction of the drivers.

What I am attempting to do is tailor the tuning for internal passive only mode. Yes the room has an influence, but that is aside from the woofer sounding correct with the mid cone and as a system as a whole.

If you put them right in a corner it will obviously have an effect.

I am talking about 4th order QB3 alignments here, other alignments may suit certain rooms and situations better, ie 4th order SBB4 and BB4 aligments yield H=1=20 hertz, Alpha=3.0127=Vb 9.62 cuft3 net, F3/Fs=1.9805=39.6 hertz.

This alignment would be an interesting exercise as the Vb is no excessive in terms of real estate and as it has the best transient response and the gradual rolloff may suit the room gain of a domestic living environment. There being less output in the 40-60 hertz rang than the QB3 alignment

The problem with this alignment is the vent constraint as long ports are required, limiting 2 4 inch ports to about 15 inches..too long for the stock 4345 , perhap an elbow port, or one 4 inch port only 6 inches long. Side mounting the ports might be an option for a diy project.

I hope all this makes some sense.

Ian

Ian Mackenzie
01-26-2007, 11:25 PM
I attempted the above BB4 alignment.

Very interesting! Almost a dry studio sound.

It certainly has the transient response, but lacks some of the punch of the typical JBL sound.

Ian

Ian Mackenzie
01-28-2007, 03:40 AM
John,


The 4345 box tuning was somewhat empirically, based on listening tests.
If you follow the plan (in the tech reference area) you can't go wrong, just do everything they say. The design was tuned on the real world, things were done a certain way (like the port length) with the actual box to get a desired result.

Ian

Well I have gone full circle here with one small but rather important realisation.

I was reading Bullock on Boxes again today and reviewing the chapter on fine tuning practical designs. QL (losses that effect the effectiveness of proper alignments) is often governed by leaks according to Robert Bullock.

Having messed all weekend I decided it was time to revert to the stock tuning. As weird as it may seem following the installation of some new casters I decided to tape the joints at the mouth of the ports.

Having already played a few tunes to check the positioning I could hear a noticeable change and improvement in the bass definition following the taping. The fit was firm but obviously not air tight.

This is a very pleasing result and I will sit with the stock tuning.

Ian