[Graham Maynard]

Hi ttan,

Oh I just don't know !

I'm sure David above would agree that for proper dipole bass you need to shift a decent amount of air (much more than with an enclosure), which means larger diameter plus good X.max.
I started out thinking I could maybe get by with 2x 12" or 4x 10", but I now see I simply cannot get the real LF without going for something like 2x 15".

My son just installed an all black Samsung vertical DVD system in his lounge with four tall speakers about 4" wide, a centre and a woofer.  His comment - the bass sounds 'bongy'. 
He already had a folded 8ft line under his bed for the old bedroom video player, and a 15" in woofer in his car, so he knows what bass is meant to sound like !

And what does my dear wife say - those would be nice in silver !
Me - there's no way I'm listening to anything like those.
She - well I don't want anything big !

I went through the entire Eminence range but was not happy with any of the 12" drivers for *dipole* bass. 

A Beta-12LT could be used, but it does not have especially low Fs or large Vd=170cc. 

The Acoustinator-N20.12 is better, but more expensive and still only Vd=203cc.
http://www.eminence.com/guitar_speaker_detail.asp?model=ACOUSTINATORN2012&speaker_size=12&SUB_CAT_ID=6


whereas a single Delta-15LFA offers a Vd=419cc.
http://www.eminence.com/proaudio_speaker_detail.asp?web_detail_link=DELTA-15LFA&speaker_size=15&SUB_CAT_ID=2


and a single Kappalite-30.15LF offers a Vd=846cc (=5x Beta-12LT)
http://www.eminence.com/proaudio_speaker_detail.asp?web_detail_link=KAPPALITE3015LF&speaker_size=15&SUB_CAT_ID=3

Some folks suggest I use a 12" car sub.  But these tend to be low 'Q', inefficient and have high inductance which makes them respond less speedily.  They often have heavy cones, whereas a dipole cone should be light and easy to drive.

Another aspect is roll surrounds.  To me the rubber ones always make the driver quite literally sound 'rubbery', and even the foam ones don't seem as clean as those with plain pleated/treated paper or cloth.

I can't offer any solution just yet;  still looking myself !

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

[D OB G]

I'm using drivers by Eminence built for Madisound as subwoofer drivers- EMI 1550. They are not high tech drivers.

My requirements were:
good efficiency- specified as 96 dB @2.83V.
low Fs- 23 as I've said earlier.

The Mms is good at 84g.
They do not exhibit a cone break-up peak, but roll off naturally at 1600 Hz.

They use a cloth accordion surround which I believe from accounts (e.g. your's Graham and Zaph's) is better than rubber or foam if it is used over a modest travel, and modest it is with only Xmax 7mm peak.  Multiple drivers really would be best, and the project I'm working on will use two per side- more would be better!

The only spec that could be better is the high Le of 4mH.

They are about US$90 each, which I consider to be very good value.

Graham, you are probably right about the 50-80 Hz anomaly, but the speakers sit (toed out about 20 degrees, but with flat response at the listening position with the current state of play) in front of a six panel Japanese screen, and I'm lucky to get away with the WAF as it is!

The DEQX filter is literally linear phase, and certainly 0 degrees over the bass freqs. so I'm relying on that to ameliorate the response deviations.  I'm yet to get Praxis out to measure the overall response.

Yes, I'm using a lot of boost- probably over 25 dB (small baffle, low Qt).

I certainly agree that your circuit has great versatility, and I wish there would be numerous accounts in this thread of many DIYers showing the various results they could get if only they tried it!

David

[BrunoB]

I'm using drivers by Eminence built for Madisound as subwoofer drivers- EMI 1550. They are not high tech drivers.

My requirements were:
good efficiency- specified as 96 dB @2.83V.
low Fs- 23 as I've said earlier.

The Mms is good at 84g.
They do not exhibit a cone break-up peak, but roll off naturally at 1600 Hz.

I wonder if a Mms of 84 g is low enough?

I have built a Linkwitz dipole subwoofer using a pair a PA Audio WN-15R:
http://www.intertechnik.de/index.html/JTI2bmF2aWQlM0QxNzY4JTI2bGFuZyUzRGRlJTI2c2lkJTNEbjQ4MWRmMjAxOTdjYzMlMjZiJTNE.html?detail=45955

I choose this driver because it has a relatively low moving mass (66 g) and  a good BL/Mms ratio (about 0.24). I am using the subwoofer for nearfield listening for frequencies below 80 Hz.
 I have measured the  FR  at the listening position: flat down to 15 Hz with equalization (Reckhorn B1). I am not happy with the sound: bass is  too slow, not detailed, not involving.   I am using a digital amp (Sharp SMSX1) and very short thick speaker cables (R <= 0.1 Ohms).

My plan is to build a new Linkwitz dipole subwoofer using 2x 9 Mission CP-138 (moving mass: 8 gr, BL/Mms is about  4 times higher than the PA Audio WN-15R). This driver costs only 19 Euros a piece! It got a very good review in the German magazine Hobby-Hifi. I will then build   Graham's LF boost circuit (from memory, the Fs of the Mission is about 55 Hz). I will also use a Reckhorn B1 to equalize the LF response.

Any advice or suggestions are welcome.

Bruno

[Graham Maynard]

Hi Bruno,

When it comes to dipole drivers we need to make do with what reasonable cone mass we can get, and 84g for a 15" cone that will probably flap about 1" peak to peak when driven hard is fair enough.

? CP-138 ?  This has an 82mm dia cone, so it simply cannot displace enough air volume to produce real LF even when being used in a line or resonant cabinet

Your WN-15R has an X.max of only 3.5mm, but it should still be very useful.

The main aspect relating to your description of its sound is actually the direct amplifier connection.

If the LF part of your loudspeaker is an enclosed box, it will always sound less open than when the same driver is in a u-frame, or even better, when on a completely open baffle.

Stick with the 15" driver and give this some time. 
If it is boxed then try with the back off.
If better, then try the transformer circuit.

Cheers ........ Graham.

[Rudolf]

I have built a Linkwitz dipole subwoofer using a pair a PA Audio WN-15R:
http://www.intertechnik.de/index.html/JTI2bmF2aWQlM0QxNzY4JTI2bGFuZyUzRGRlJTI2c2lkJTNEbjQ4MWRmMjAxOTdjYzMlMjZiJTNE.html?detail=45955

... I am not happy with the sound: bass is  too slow, not detailed, not involving. ...
Any advice or suggestions are welcome.
Bruno

What kind of "Linkwitz dipole subwoofer" is this? W-frame or H-frame? If first, try the latter.

[BrunoB]

I have built a Linkwitz dipole subwoofer using a pair a PA Audio WN-15R:
http://www.intertechnik.de/index.html/JTI2bmF2aWQlM0QxNzY4JTI2bGFuZyUzRGRlJTI2c2lkJTNEbjQ4MWRmMjAxOTdjYzMlMjZiJTNE.html?detail=45955

... I am not happy with the sound: bass is  too slow, not detailed, not involving. ...
Any advice or suggestions are welcome.
Bruno

What kind of "Linkwitz dipole subwoofer" is this? W-frame or H-frame? If first, try the latter.

W-frame.

My sub looks like this one:

[BrunoB]

Hi Bruno,

When it comes to dipole drivers we need to make do with what reasonable cone mass we can get, and 84g for a 15" cone that will probably flap about 1" peak to peak when driven hard is fair enough.

? CP-138 ?  This has an 82mm dia cone, so it simply cannot displace enough air volume to produce real LF even when being used in a line or resonant cabinet

Hi Graham,

thanks for  your comments.


I calculated that the  surface iof the Mission CP-138 is about one tenth of the 15'' driver. I plan to use 18 of those little drivers. I already bought 4 of them to make a test. My main concern with smaller drivers is that they can get noisy when reaching Xmax. The good news is that I don't need a lot of power because the  dipole sub is placed very close to the listening position (about 30 cm). This setup has two advantages: no room interactions and the neighbors can't hear it.

The main aspect relating to your description of its sound is actually the direct amplifier connection.

Could you elaborate?



Bruno

[Graham Maynard]

Hi Bruno,

Aha, a 'W' ripole, where the amplifier powers opposing drivers.  A ripole cabinet might output the driver's full displacement, but not as efficiently as a plain dipole because the drivers do quite literally work against each other.
http://www.lautsprechershop.de/hifi/index_en.htm?/hifi/ripol_en.htm

Also, I'm not so sure that your WN-15R drivers are best suited to this kind of loading;  more like this;-
http://www.forum.poweraudio.ro/viewtopic.php?t=2136

If they were mine I would have them on a flat baffle, one above the other with choke and/or line EQed drive, plus the transformer circuit giving additional lift below Fs.
An arrangement looking like the Emerald Physics CS-2;-
http://www.emeraldphysics.com/intro.htm
http://cls.audiogon.com/cgi-bin/cls.pl?spkrfull&1215130213

Or similar to Post#7 in both these threads (wings optional);-
http://www.diyaudio.com/forums/showthread.php?s=&threadid=122152
though here the lower drivers are framed to increase efficiency.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=122361

Cheers ....... Graham.

[Graham Maynard]

Hi Bruno,

Came across some useful feedback here;-
http://www.diyaudio.com/forums/showthread.php?s=&threadid=108893&highlight=

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

[panomaniac]

Hey Graham

Thanks for turning me on to this circuit.   Just now banged one together with clip leads and spare parts.
At first, I thought it didn't do anything.  But after a few cuts of random jazz, I hear the extended low end.  Maybe the 1st few cuts didn't have any real low end.

Here's the question.  How to tune it?  I just slapped together what parts were on hand.  24V CT transfo, 12mH choke, 750uF cap, 1.5R.  Followed by a 2nd order passive LP.  Now how do you tune to match the driver?  How to figure the L/C values?  E.G., my drivers have an Fs (measured) of 33.6Hz.  My L/C values have to be way off, but how do you figure them?

[Graham Maynard]

Hi Michael,

Glad you were able to try the circuit.

I can't think of any way of 'calculating' the component values necessary because so much happens between the voltage waveform being fed to an amplifier's input and the transduced reproduction being experienced by a listener.

I'm stuck for time at the moment, but I will write some notes about optimising the driver response and modify this reply asap.

The circuit works best when directly connected to NFB (low impedance voltage output) amplifiers, so, if a LS is not being driven by LS sited monoblocs, the transformer circuit will work best by having a direct connection to the amplifier output terminals (co-sited) with separate wires then being taken out to the LF driver/crossover section alone.

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

[panomaniac]

Thanks Graham

I can pretty much figure out how to hit a needed frequency range ( I think)  but wonder where it should be.  Related to the driver Fs?  Are you doing anything unusual with the phase?

Thanks for the tip on placement of the circuit.

BTW, what are you guys using for caps?  I don't have any non-polar that big, so have used 2 polarized caps back to back.  Is that what you're doing?

[Graham Maynard]

Just electrolytics here for now.  Low ESR types. 
To do properly would bias up a back to back pair with say a permanently connected PP3 if good non-polarised not available.

Values will depend on the driver Qes as well as Fs. 
Also baffle size and room position.

Yes phase is modified, and series impedance optimised about resonant frequency so that the driver stores less energy.

More tomorrow.

Cheers ...... Graham.

[Graham Maynard]

This T-bass circuit was designed for OB and H/N/U/W etc. mounted drivers having sufficient X.max capability for extended LF response.  It may also be used with aperiodic (IB) enclosures, though is not recommended for any systems which electro-mechanically generates a double humped impedance characteristic, as with inadequately damped transmission lines and vented enclosures.

Adjusting the component values.

The LF choke controls the frequency below which an increase in LF drive is generated;  it is electrically in series with the transformer. 
The transformer steps-up the drive voltage, but also steps-down the loudspeaker driver impedance which loads the amplifier.
The choke value cannot be ascribed to a turn-over frequency (as when being used in crossover) due to combined transformer action and the specific LF driver impedance resulting from its individual mounting.

Below Fs a LF driver's impedance seldom falls back to its nominal resistive value at low AF, and this is the reason for an amplifier driving this circuit not 'seeing' an 8 ohm loudspeaker as a 2 ohm load via the 1:2 transformer ratio.

The resistor in series with the choke helps to limit the peak current at infra bass frequencies, and reduces the Q of the series tuned circuit simultaneously formed with the capacitor connected across the transformer primary. 
Without this resistor the voltage drive would be almost doubled at frequencies below those responsible for music content, which would mean almost four times the power.  The LF amplitude could then exceed a driver's X.max and affect reproduction linearity at normal bass frequencies. 
If you have a good amplifier and multiple bass drivers then a lower value or no resistor at all might be possible, peak amplifier current drive then being limited by the combined series resistance/impedance of the transformer primary and choke windings in series with the transformed LS impedance divided by four.

So the choke value needs to match an individual driver, its mounting (size and type) and its X.max capabilities, rather than simplistically be calculated to turn-over at Fs.  Also, optimum reproduction will arise with slightly different choke values if the series resistor value is changed, or indeed if the capacitor and its series resistor value is changed as well.

The capacitor value is chosen to ....

To follow.

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

[D OB G]

As Graham says, there is no point in chosing an inductor below the capablities of the driver, or indeed the lowest freq the room can support.
For the exercise, I demonstrated that I could get to a measured 15 Hz, but this is a waste of time.
30 Hz is a more reasonable goal with typical bass drivers having an Fs of around 40 Hz.

David

[BrunoB]

I wonder if a Mms of 84 g is low enough?

I have built a Linkwitz dipole subwoofer using a pair a PA Audio WN-15R:
http://www.intertechnik.de/index.html/JTI2bmF2aWQlM0QxNzY4JTI2bGFuZyUzRGRlJTI2c2lkJTNEbjQ4MWRmMjAxOTdjYzMlMjZiJTNE.html?detail=45955

I choose this driver because it has a relatively low moving mass (66 g) and  a good BL/Mms ratio (about 0.24). I am using the subwoofer for nearfield listening for frequencies below 80 Hz.
 I have measured the  FR  at the listening position: flat down to 15 Hz with equalization (Reckhorn B1). I am not happy with the sound: bass is  too slow, not detailed, not involving.   I am using a digital amp (Sharp SMSX1) and very short thick speaker cables (R <= 0.1 Ohms).

I have now to take back what I wrote above. The problem was the low pass filter I used. It was not steep enough and let too much sound above 100  Hz pass through. The Reckorn B1 has a poor low pass filter. I solved the problem by using the subwoofer output of a HT receiver.

This is the frequency response of the W sub at the listening position with boost applied (sub is placed just behind the sofa):


The -3 db point is 21 Hz.  The FR is almost perfectly flat, but that is just luck.  The rapid drop below 20 Hz is partially due to the subsonic filter of the B1.  Bass is very tight and extended and dynamic. It sounds very good. There is no room interactions. The sub does not vibrate. The woofers barely move.

I don't plan to add Graham's circuit at this point because the frequency of resonance of the woofers in this W frame is too low (26-27 Hz).

Bruno

[Graham Maynard]

Hi David,

I have not yet got round to writing more about the components.

I was surprised at how low you had the circuit running, but as you say it is almost a waste of time because you run out of X-max and clip the other bass frequencies.

I wish your EMI-1550 drivers were available here.  I have spent hours going through manufacturer's specifications and can't find anything like it in the UK or in Europe.
Have you come across anything better ?

If an equaliser was not being used the transformer circuit could work on its own to optimise reproduction - the transient response and the clarity.
So David, I wonder if you were repeatedly alternating between tuning the transformer circuit and adjusting the EQ until you were happy with the composite response, or did you do one before the other and just leave it at that ?

Personally I would not look for a response which was flat and suddenly rolled off at 20Hz, due to the LF phase change which would then arise over the low and mid bass range of frequencies.  I would still be looking for some output down to 15Hz, though at a reducing level which with room gain would not lead to driver over-excursion.

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

[Graham Maynard]

Hi Bruno,

Good LF there, and you say a system resonance of 26~27Hz with drivers having a specified Fs of 39Hz.

David's driver has a Qes of 0.23;  yours 0.26;  both low and thus preventing unwanted driver resonance when the driver is not loaded by cabinet air.

The transformer circuit has an advantage other than simply boosting LF amplitude, which anyone can already do with EQ.  It improves the quality of the bass beyond what a directly coupled amplifier can achieve, whether separately EQed or not.

It is looking like the only 15" OB suitable driver I will be able to order in the UK is a SM-115N or SM-115K at
http://profesional.beyma.com/ENGLISH/producto.php

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

[panomaniac]

Hi guys--

I can some quick circuit simulations on the T-Bass.
Will post images of the graphs as soon as I figure out how to post them on this forum.

Basically, the circuit always gives the same amount of boost +6dB.  That's fixed by the transformer windings.
The circuit Q seems to be set by the inductor, the bass peak frequency set by the cap.  The resistor changes the Fpeak a bit, but it seems better done with the cap.

The circuit with no inductor has a low Q and just makes a nice bass boost with Fpeak set by the cap.

In the sims I used the inductance values for my transformer.  30mH each side of the secondary and 350mH for the primary.  Speaker model was very simple.  Just 6.5R in series with 0.9mH which corresponds to the values of my driver.  Using a more complex speaker model, things can get strange, fast.

Someone who is better at Spice simulations should be able to show us a lot more.  Modeling the driver is going to be the tricky part.

EDIT:  Well, for some reaason I can't seem to get the graphs on here.

You can see them here at the gallery:  http://www.audiocircle.com/index.php?action=gallery;area=browse;album=1452&pos=0

[Graham Maynard]

Hi Michael.

+6dB would be the maximum gain possible with simulated 'lossless' components.  However a real choke and transformer will have winding resistance,and there is a resistor in series with the choke too, so the maximum gain is not likely to be +6dB. 
Increasing the R in series with the choke decreases the maximum LF boost and the effect of its value is increased due to the effective LS impedance transformation.

The primary is not used.

An equivalent LS load would be necessary having an impedance peak which matches driver characteristics to properly study the circuit.  Even then I'm not so sure that the waveforms generated could express the sound being heard.  Remember that the voltage waveforms do not tell the whole story and there is a need to observe driver current as back-EMFs are reactively generated.

If I was trying to design this circuit from simulation I don't think I would have got anywhere.  This was checked out in real life first.  There might be the possibility of further improvement by placing a Zobel over the entire winding, but I have not tried this and will not be able to for some time.

Real drivers have a phase leading characteristic as they go through resonance, this being different to their electrical phase characteristic through resonance.  I see you have captured the lag introduced by the circuit.  This flattens the overall phase response by effectively shifting the lead down to a lower frequency.

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