[Defo]

Im toying with the idea of using my 2226H woofers in an open baffle covering around 60hz to 500hz, and using a dedicated subwoofer for the lowest octaves.

Any thoughts of how this might work out? Im aware that its preferable to use a high qts driver in OB designs, but if one is aiming on only achieving good output down to ≈60hz, is it doable with my current 2226H woofer?

[JoshK]

Its a good question.  Essentially I am doing similar using a pair or quad of Lambda TD15X.  I am targeting 50hz - 300hz, but just on paper, not really firmed up.

[MJK]

How are you planning on getting low Qts drivers to reach below 100 Hz? EQ? What baffle size? What kind of effciency do you need to achieve to mate with the midrange driver?

[Defo]

How are you planning on getting low Qts drivers to reach below 100 Hz? EQ? What baffle size? What kind of effciency do you need to achieve to mate with the midrange driver?

Not too keen on mounting them in a large baffle, so I guess some eq will be necessary? I havnt done any calculations as I have no experience with dipoles. They are being mated with an Altec compression drivers by the way.

Im just thinking; a little while ago I used the 2226H in free air (!) as I had to improvise a little while making my enclosures. I measured frequenzy response to have good output untill about 100hz, where output dropped off rapidly. Can it then be so hard to acheive ≈60hz in a small/medium sized baffle?

[MJK]

In my opinion, putting low Qts drivers in a moderately sized baffle and trying to get decent bass using EQ and amp power is forcing a solution. That type of driver is just not naturally meant for the OB applications, it is designed to be used in a box. I really think you would be better off with a lower cost higher Qts driver which only covers the frequencues below a few hundred Hz. But that is just my opinion.

[Magnetar]

MJK is right, wrong woofer.

I have a bunch of 2226's - even in a box they are a bit weak below 80 cycles.

Maybe use it as a midrange with several OB 18's for bass and some subs below the 18's to do the very bottom.

[Graham Maynard]

Hi Defo,

As far as I know neither MJK nor Magnetar have tried my T-bass circuit.

High Q drivers increase LF output electro-mechanically, which means with resonant *delay*.

My T-bass circuit increases LF output electrically and not only without that mechanical delay but with minimal electrical phase change, so it does not cause the same loss of LF coherence. 
Those who have not tried this circuit will not not know what I am writing about.

There is not any reason why you cannot try your JBL on an OB, and if you run the T-bass circuit as well I suggest that you will never look back.

http://www.audiocircle.com/index.php?topic=53322.100

Thus I could say that using a low Q driver ( and not higher Q ) on an OB is the correct way forwards for quality reproduction, BUT, it will need boosted LF input, and to my ears EQ is every bit as bad as driver Q. 
Hence my considerable efforts which eventually resulted in the T-bass circuit.

I suggest that buying the parts to construct a T-bass circuit for use with your current very nice JBL drivers will be money better spent than purchasing higher Q drivers which cannot ever reproduce as accurately.

Cheers ....... Graham.

[JoshK]

Graham,

Have you taken woofer in question's FR and impedance curve with and without the circuit in place for comparison?

[Graham Maynard]

Hi Josh,

Clearly you have something in mind relating to your question, but I do not know what it is.

The low frequency response of all LF drivers is dominated by open baffle mounting, esp. the front/back edge cancellation, this being the aspect which the T-bass circuit partially addresses.  Then there are other characteristics which influence Fs and VAS.
The transformer circuit cannot do more than a higher Q driver can, but it most audibly allows a low Q driver cone to much more accurately transduce LF electrical waveforms.

Adjusting the capacitor and inductor values can then optimise for other driver characteristics.

What makes a driver good for LF on an OB is displacement, and there is no reason why Defo should be advised to buy another when he already owns an excellent product.

Cheers ......... Graham.

[JoshK]

I guess I am trying to get at what it does more analytically.  You've described what it does subjectively, but I am trying to understand the circuit and what it does electrically. 

I was thinking that if you had nearfield FR response before and after then you could model then change with the predicted baffle responses.  In this way, it would be understood what the circuit what actually doing. 

I am not above trying it.  I have plenty of parts and likely a suitable transformer on hand.  It will taken me some time before I can try it though (probably in the new year) with my current schedule.

Josh

[Graham Maynard]

Not a trick question - but what is your NF FR response ?

Answer - amplitude only.  Not coherent LF wave reproduction.

EQ introduces lag/lead depending on type.
So does LP at say Fs on OB to flatten the overall response.  This especially introduces additional lag at the crossover frequency, causing a 'slow' bass which simply cannot be overcome !
Anyone already used to a compromise relationing to this 'slow bass' is in for a treat when they try the T-bass.

The T-bass amplitude gain is not phase shifted below Fs, and with a low Qes driver the transduction is also more phase coherent.

I can design any number of passive or active circuits which will flatter FR, but none will allow a driver to reproduce as coherently and with lower levels of stored energy than will the T-bass. 
Phase at 20 and 200Hz are flat.  What EQ or crossover used to provide OB EQ can do that ?

About 10 years ago I designed an e-bass circuit which provided adjustable LF lift (0 to 20dB @ 20Hz wrt 100Hz) and holds the phase flat throughout, but it runs at line level and does not prevent LF driver storage muddying reproduction when directly connected to a power amplifier.

Flat FR is useless if drums and LF percussion lose their dynamic impact !

Needs to be heard, and cannot be simulated via computer !!!

The transformer step-up doubles amplifier output voltage.  This is in series with a choke so increase in LF is instant but at increasing impedance with frequency.
Below Fs driver Z has still not dropped to nominal, so the amplifier power requirement is not quadrupled, and with OB mounting the load is not as with an enclosure anyway, so amplifiers are not overloaded.
The increasing series drive impedance around Fs prevents the driver electro-mechanically storing as much energy, so the reproduction does not become as muddy as with line EQ and direct amplifier connection. 
Above Fs the output is not doubled and can even be reduced via the series C+R working in combination with the driver and its parallel connected Zobel to reduce higher frequencies.

A thread here trying out the T-bass.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=128015&perpage=25&pagenumber=1

Cheers ......... Graham.

[JoshK]

Fine, maybe I am not asking the right questions.  But, if you designed this circuit, you had to have some idea in mind of what you were trying to acheive electrically, right?  Or was it a 'suck it and see' procedure?  If the former, then what electrical "features" were you trying to get at?

With all due respect, I am trying to look through the 'marketing claims' to what it actually "does". 

[Graham Maynard]

We all know what the aim is with LF - to increase voice coil/cone amplitube below cabinet/OB/driver roll off, but to do it with minimum phase shift and minimum Q, because Q introduces recognisable phase shift too.

You can't 'look at' the 'marketing claims' because the dynamic coherence of music reproduction cannot be visually displayed using ordinary equipment, you need to hear it to appeciate the concept and the compromises you have for so long been accepting.

Suck and see ?  Took me far too long for it to be that.
There are even other ways of using the same components I do to acheive the same response, but for each different way the dynamic current flows with amplifier voltage will be different, and thus the reproduction is too.

Maybe it is more simple than what you are enquiring about - for it is certainly far more simple than I could ever have believed possible after 40yrs of working with amps and spkrs.
It was like a 'Eureka' situation, where several aspects just clicked together in one simple circuit.

Cheers ......... Graham.

[JoshK]

Thanks.  FWIW, I wasn't trying to be combatitive or dismissive.  I am just trying to understand what it is doing. 

It just seems to me that one should be able to measure the additional LF boost, if you will, and the change in Q, phase or whatever else was the constrained on that problem.

[Graham Maynard]

Can I suggest that you imagine sine voltage waveforms passing through the circuit.
Say 20Hz where the capacitor is approx OC and the choke SC, which gives almost 2x voltage output.

Say 200Hz where the choke is approx OC and the capacitor SC, which gives capacitor series R in series with driver DC R // Zobel or additional series choke here depending on the woofer.

Say series L-C resonance where the tuned circuit on the primary side is used in opposition to the driver/baffle lift so that the driver cannot absorb more energy than from initial step-up due to the L-C acting as a parallel tuned circuit on the step-up tap.

The L an C are tuned to help reduce the driver/baffle peak.  First half cycle dynamics are plain impedance ratios in series with the driver, beyond then the L and C with series damping resistors additionally reduce drive where continuous waveforms would lead to (loaded) cone mass/suspension Q induced peaking with phase shifted energy storage.

Unless you have another question that makes me answer differently I don't know how much more I can explain.

Of course the gain and Q changes could be measured, there have been some who have measured the improvement already.

However it always comes down to what we *HEAR*.

Hope I'm getting through here ?  - The amplitude response is irrelevent if it is not in phase !!!!

EQed to flat output which is out of phase sounds more distorted than naturally rolled off output which is coherent.


Cheers ...... Graham.

[Defo]

Very interessting reading Graham! Although I dont get all the technical stuff.. 

I tried my 2226H in free air today, and I just loved the sound! (above 100hz of course - since thats were it rolls off) I'm destined to put em in OB as soon as I get a subwoofer to take care of the rest. That being said; I may send you a PM when the time arrives, if thats okey with you?

Cheers

[Graham Maynard]

Feel free Defo.

Glad you are moving forwards with what you already have.

You could try running them in by driving then to maximum excursion with LF drive.
This can improve LF output.

Cheers .......... Graham.

[JANDG]

I also am considering using the 2226H drivers out of my JBL SR4725's for OB project. I like the idea of quality Mid/midbass driver as the heart of OB system. My experience in OB says having a above average large OB driver as mid/mid bass is excellent start. Much easier to integrate the 40-80/160hz. I would also like to use the 2426H CD & 2370 horn I allready have. Anyone see a problem using this CD & horn in a OB rig..? Only thing I would need is a very high quality OB bass driver. around $300. a pop for those should cover that. I can actively xo everything since have enough amps, or just active on the OB bass driver & passive on the 2226H & 2426H cd & 2370 horn combo. Someone please stop me if this is NOT a good idea..
Thanks J & G

[D OB G]

Hi JANDG,

One school of thought suggests picking a crossover freq from mid to high where the directivity is equal, if possible.  There is more chance with a compression driver.  i.e. need a lowish freq when the mid is large.  You'd need to know the coverage pattern of the 2370 horn.

David

[D OB G]

Hi JoshK,

http://www.audiocircle.com/index.php?topic=53322.20

When it comes to trying to give one (only one of many parameters) objective measurement, Reply 37 shows measured FR down below resonant frequency of driver (23Hz) using Graham's T-Bass circuit.

This was an exercise, just to see what could be done, but through the thread Graham gives details of the interrelationships of the components, and the possibility (virtue?) of using lower Q drivers.

David