Isobaric 2245

[more10]

HT BB6 Pro reports that an 8.757 cu ft box (net effective) with "typical" fiberglass fill yields a Qtc of 0.5

Oh well, the transform should "fix" the difference.

Could be wrong T/S parameters in WinISD, or a bug. This is what I have:


From JBL:

MODEL FS QTS QMS QES VAS EFF PE XMAX RE LE SD BI MMS FLUX
2245H 20 0.27 2.2 0.27 821.2 2.1 300 9.65 5.8 1.4 0.130 21 185 1.22

There is a big difference in QES. WinISD calculates it.

[more10]

Oh well, the transform should "fix" any difference.

In frequency domain yes. But not in time domain. It might sound very much alike, but the engineer on my shoulder tells me to build to Q 0.500000 :-)

[4313B]

In frequency domain yes. But not in time domain. It might sound very much alike, but the engineer on my shoulder tells me to build to Q 0.500000 :-)

I hear you!

This is what BB6 Pro has:

[4313B]

Disabling the damping option.

[more10]

Here we go. Disabling the whole "fill" nonsense:

Still one cubic feet missing. Can you try with my value for QES?

Since QES is calculated in WinISD there is probably something wrong with the JBL datasheet.

[4313B]

It didn't like the printed JBL value of Qes, volume suggestion went down a bit.

We went through this years ago with the E145 too. That is how we discovered some parameters had been documented by JBL incorrectly. In that case it looked like someone swapped a 4 and a 7 in the Vas value if I remember correctly. I think it was supposed to be something like 470 instead of 740, something like that.

[more10]

WinISD QES is 0.306.

So there is one cubic feet difference. If you simulate a 12 cubic feet enclosure, how much does that affect Q?

[4313B]

WinISD QES is 0.306.

Right, sorry, I see that I used the 0.27 value you posted from JBL.

So there is one cubic feet difference. If you simulate a 12 cubic feet enclosure, how much does that affect Q?

One with "typical" fill (0.449) and one with the "fill", aka "Damping" tab, disabled.

This is the net effective volume, nothing is taken into consideration, no bracing, no fill/damping, no transducer displacement, no series inductor, nada. Just an effective volume.

Usually what I do is build the box without anything BUT fiberglass fill taken into consideration via BB6 and the net effective volume works out just about right.

For example, I build a 9 cubic foot box, brace the hell out of it, bolt in the 2245H, put in the dual six inch diameter ports and I come up with a 7.8 cu ft effective volume. Then I line every panel except the baffle with 2" fiberglass from home depot and voila! 9 cubic feet effective volume. Sometimes I will just use 1" fiberglass to get a smaller net effective volume. If you don't use fiberglass, and everyone pretty much hates the stuff, then all bets are off with my little process.

[pos]

Oh well, the transform should "fix" any difference. Just build some boxes and let the transform sort them out.

In frequency domain yes. But not in time domain.

A proper Linkwitz Transform (ie not a simple EQ, but a real asymmetrical shelving well matched to the actual parameters of the high pass, and correcting the response down to the VLF) will correct both magnitude and phase all in once (ie minimum-phase).
You can actually turn any 2nd order response into another one for both magnitude and phase using this technique, without any drawback beside a different power consumption curve.
WinISD Pro will let you test all that and show the resulting VA(f) curves.

[more10]

The difference in Q in a box without fill is 3%. Nothing to worry about.

Fill will increase the box size as seen by the driver, lowering the Q.

So anything between 9 and 11 cubic feet is ok for a closed box. Making the box bigger will give you plenty of room for vents if you want to add them later.

[4313B]

A proper Linkwitz Transform (ie not a simple EQ, but a real asymmetrical shelving well matched to the actual parameters of the high pass, and correcting the response down to the VLF) will correct both magnitude and phase all in once (ie minimum-phase).
You can actually turn any 2nd order response into another one for both magnitude and phase using this technique, without any drawback beside a different power consumption curve.
WinISD Pro will let you test all that and show the resulting VA(f) curves.

I like the Crown as opposed to the miniDSP if only because the Crown makes it all so much simpler to me, plus, that all in one box thing is nice.
The miniDSP is very fun to play with though. I used it on my little Tannoy dual 6" subs which are linkwitz transforms.

In any case, I think we agree that isobaric isn't what we would do in our houses?

[Ian Mackenzie]

Interesting discussion

Have you evaluated them in your room in a standard enclosure yet?

I think there are two schools of thought or two roads.

Are you wanting subs or an integrated system?

If the former look at GTs original sub article. It recommends an equalised 8cu ft box vented.
Subjectively these woofers perform better dynamically in vented boxes.

The alternative is a larger un eq box about 12 cu ft
Crossed over below 80 hertz as recommended you don't have to worry about group delay or resonances from a large box.

I run one 2245 in 8 cu ft in my HT room with 800 watts and it's fine. Doubling the cone area will reduce the distortion below 50 hertz but thus depends on the room size and how loud you need it. I personally think for HT amp power is subjectively more important.

If you want stand alone boxes (like in the French link) it's a different ball game.

Integrating to the 2206 correctly in an important design goal as this will impact on out outcome subjectively particularly if you plan to run normal program material.

Attempting to crossover below 150 hertz will be challenging without sophisticated analysis.

Some people end up running them below the cut off F3 of the bass mif woofer and that is around 40 hertz with the belief that is the only way to do it.

You can do that but you will loose punch of the cone area and displacement of the 2245 in the hope on bass extension.

By crossing over from 150 hetz and above up to 250-300 hetz I think you will free up a lot of design goals and find it subjectively superior.

I have found between 9-10 cu feet optimum in a well brace enclosure and experient will the amount of fill and boundary layer damping on the walls ( too suppress standing waves 300-500 hertz).

Try and get the woofer at least 6 inches off the floor or better still use a graphic eq to correct any boundary reinforcement in tbe 100 hertz range.

You really need to do this or it will sound like a dogs breakfast no matter how well thought out the design is.

This is what gives the impression of lack of bass clarity. The woofer is either in piston range or it's not. Speed and all that is crap provided you not inviting significant group delay by trying to stretch the f3 that is swamped by room gain in most
cases anyway.

Hearing an integrated system well set up with a Smart Live using the 2245 (like Bo's 4345) is an eye opener.

This ties in with 4313's comments in the earlier post.

You can attempt to design for a balance of driver, box and room response Or you can modify the response externally with active EQ (like the M2).

[gasfan]

I guess you just can't help it :-) Im still here.


You will have room gain below 29 Hz. Any of the designs made with the 2245 will need eq, so if you can change it in the filter you can adjust after you have built the boxes.

Regarding the 2245 you have, there is plenty of good advice in this thread. You just have to decide what you want to implement. Isobaric is not a good solution. If quality is more important than output go for closed enclosure. If output is more important than quality go for vented.

If placing two 2245 together, mount one driver with the magnet facing out in order to get symmetri. If placing the boxes in different corners there is no point doing this.

Yet another alternative is to build Onken enclosures. Play around with the online calculator: http://www.mh-audio.nl/onken_br.asp. It will sound more like a horn or a closed box, than a vented.

I can't remember the last time my wife said that. Thanks.

Lee mentioned the Onkens also. Yes, lots of good advice. Thanks so much.

[gasfan]

Interesting discussion

Have you evaluated them in your room in a standard enclosure yet?

I think there are two schools of thought or two roads.

Are you wanting subs or an integrated system?

If the former look at GTs original sub article. It recommends an equalised 8cu ft box vented.
Subjectively these woofers perform better dynamically in vented boxes.

The alternative is a larger un eq box about 12 cu ft
Crossed over below 80 hertz as recommended you don't have to worry about group delay or resonances from a large box.

I run one 2245 in 8 cu ft in my HT room with 800 watts and it's fine. Doubling the cone area will reduce the distortion below 50 hertz but thus depends on the room size and how loud you need it. I personally think for HT amp power is subjectively more important.

If you want stand alone boxes (like in the French link) it's a different ball game.

Integrating to the 2206 correctly in an important design goal as this will impact on out outcome subjectively particularly if you plan to run normal program material.

Attempting to crossover below 150 hertz will be challenging without sophisticated analysis.

Some people end up running them below the cut off F3 of the bass mif woofer and that is around 40 hertz with the belief that is the only way to do it.

You can do that but you will loose punch of the cone area and displacement of the 2245 in the hope on bass extension.

By crossing over from 150 hetz and above up to 250-300 hetz I think you will free up a lot of design goals and find it subjectively superior.

I have found between 9-10 cu feet optimum in a well brace enclosure and experient will the amount of fill and boundary layer damping on the walls ( too suppress standing waves 300-500 hertz).

Try and get the woofer at least 6 inches off the floor or better still use a graphic eq to correct any boundary reinforcement in tbe 100 hertz range.

You really need to do this or it will sound like a dogs breakfast no matter how well thought out the design is.

This is what gives the impression of lack of bass clarity. The woofer is either in piston range or it's not. Speed and all that is crap provided you not inviting significant group delay by trying to stretch the f3 that is swamped by room gain in most
cases anyway.

Hearing an integrated system well set up with a Smart Live using the 2245 (like Bo's 4345) is an eye opener.

Thanks again Ian. I will stay with vented.

[pos]

I like the Crown as opposed to the miniDSP if only because the Crown makes it all so much simpler to me, plus, that all in one box thing is nice.
The miniDSP is very fun to play with though. I used it on my little Tannoy dual 6" subs which are linkwitz transforms.

Can you do L-T with the Crown? When I checked the architect software I don't remember seeing any L-T.
Hypex has asymmetrical shelvings which are the same thing.
Too bad the Crown does not let the user load biquad IIR coefficients, and even FIR ones for that matter...
Being able to freely use the FIR in the crown (instead of those textbook linear phase filters that are of no use if the acoustical slopes cannot be taken care of first) would let M2 users linearize the phase of the 800Hz crossover, and even possibly the BR...

In any case, I think we agree that isobaric isn't what we would do in our houses?

I have never tried it myself, but I cannot see any really good theoretical advantage to it (if EQ is a given...). T can so many disadvantages and pitfalls when implementing it tho, especially with 18" drivers that are supposed to play up to the midbass range