[Graham Maynard]

The Spirit...
................of Music...
................of Open Baffle...
................of Do It Yourself...



Shown leaning against the wall, this was thrown together in just 3 hours, however the thought processes took months! 
Height is ~5ft (1.5m), width 18" (~46cm).  A base and low support/wings are to be added.

The top driver is a 10" Fane Sovereign 10-125 (an Eminence Alpha-10A equivalent) used as BSC for the 8" B200.

The Visaton B200 is fitted with a conical dust cap mod and foam fingers as mentioned here (post#17);-
http://www.audiocircle.com/index.php?topic=51673.0

Low frequencies are transduced by 2x Beyma SM-115N driven by the 'T'-bass circuit;-
http://www.audiocircle.com/index.php?topic=53322.0

The SS NFB amplifier is of my own construction, a 100W/4R/choke version which is not upset by low impedance loads;-
http://www.gmweb2.net/

Note that the LS crossover/wiring has merely been 'thrown together' to see if the ideas gel acceptably, and yes - they do.  (The only complaint so far from my wife is that the bass was so loud and firm - in the kitchen - that it affected her concentration when cooking;  this with flat AF drive - not EQed.) 

So far the crossover is;-


Essentially the B200 is driven 'wide-range' with a series 'C' for bass protection.

BSC and level adjustment to compensate for the rising with frequency B200 sensitivity is added by the other drivers, not by using components in series with the B200 which would otherwise interfere with its dynamics and natural clarity.

All other drivers have high cut via the single series choke, though when time permits the individual upper and lower driver responses must be individually defined.
 
The Fane has its own series 'C' bass cut to protect it from LF.

The Beymas are driven in parallel by the 'T'-bass step-up transformer arrangement. This increases voltage drive around Fs where resonant impedance is high.  However, the capacitor removes that boost above driver Fs, thus effectively there is increased LF boost only below driver resonance on the OB. 

The first B200 I tried with the other drivers on this baffle had a normal dust-cap, however the high frequency response was too 'in-your-ears' peaky (yeough!).  Here the conicular dust-cap plus foam finger modification replaces that 'shrillness' with 'charm'.

I also tried the B200 without any low-mid augmentation from the 10"; this made the reproduction sound 'thin' on the likes of saxophone and piano, with their 'detail' remaining 'clear' but their sound as if lacking in 'body' and thus not presenting with sufficient physical 'size'.  The Beymas have a reasonable mid-range response yet were incapable of providing the necessary low mid augmentation on their own; I put this down to the electrical phase change introduced upon the Betmas' electrical input as a side effect to the amplitude change necessary to provide the appropriate mid frequency SPL response.  In short, they could not be made to blend in inconspicuously.

I also tried the low Q Beymas without the transformer boost, whereupon reproduction lacked both depth and punch.  This circuit behaves like a user adjustable driver converter, where the LF response can be optimised instead of having to make do with a driver manufacturer's higher Q specification, as is often chosen to provide a 'flatter' SPL response but with a lack of direct and 'dynamic' air motion drive.  This transformer/choke arrangement effectively increases output below driver Fs, instead of above where it is SO NOT wanted due to OB and room interactions already providing much more boost than is tolerable!
(Probably why OB bass has not been successful without EQ or separate amplification in the past.)

I might yet add an Eminence APT (with series 220 to 470nF capacitor and parallel 22 ohm damping resistor) as a supertweeter, or rear HF projector.

Other drivers could be used - nothing critical.  Maybe even a Fostex where I have the B200, and for the 15"ers - Madisound EM1550 or from Eminence say 15" Beta, DeltaLF, Kappa or KappaLF types.

For me this is only a start, but it is a very good start. 
I shall be refining it further, however I am not in a position to do so as quickly as others might, thus any follow-ups are likely to be widely spaced.

I wanted good OB reproduction from a single amplifier and without EQ:  Job almost done !

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

[ttan98]

Questions please,

1. Any freq. response curve?
2. How does it compare to your previous OB speakers?
3. Are the sound up to your expectations?

cheers

[panomaniac]

Very Nice Graham!

I notice that you are not using a resistor in the LR leg of yout T-Bass circuit here.  Any reason for that?

[Graham Maynard]

Hi ttan,

No response curve.  Very early on here.
Also what I am looking for is a LS with good dynamics;  the FR does not show that.  You can have a flat SPL, but lifeless reproduction.

I have not had any other LS compare, also loud for voltage.

On its own the underhung B200 is a bit strident, but with crossover underlap from the other drivers to provide the baffle step and sub 1kHz amplitude compensation, it can be transformed.  Thus the driver responses gradually overlap below 1kHz until the B200 sensitivity wins out and holds the image detail as the others provide the body.
I believe very similar to what Richard reported here;-
http://www.audiocircle.com/index.php?topic=56536.0

with the sensitivity of Richard's single Alpha at 50Hz probably being similar to both of my Beymas.

Expectations ?  Yes, I'm very pleased, and not even optimised yet.
Drums sound like drums.  (If only I'd thought of this years ago before I sold all my 18" Goodmans.)
Will be some time before I can develop this more though.


Hi Michael.

No resistors ? 
Two reasons. 
The big drivers are not at all run in yet, so I'm providing extra drive.
The Beymas are in parallel so choke and transformer resistances have a much greater impact - as if resistors are already fitted.

Must look at other transformers or wind my own to minimise losses because it is better to have adjustment with external components than to be limited by winding resistance.

Actually when driving hard I noticed the small capacitors I was using became quite warm.  So must use bigger for the final construction.

When I try a simple crossover choke in series with LF drivers, say 3 to 10mH, the bass does not integrate with the sound of the wide-range driver.  This is what I have always noted and bass seems 'slow'.  The capacitor in the T-bass circuit brings the bass up to speed - though part of the matching here is due to the high sensitivity of the B200.


Cheers ....... Graham.

[panomaniac]

Thanks Graham.

Keep us up to date on the rig!

[Graham Maynard]

I think this project also represents the Spirit ... of 'Fullrange', for it is based upon a widerange driver with augmentation, rather than a typical woofer-mid-tweeter arrangement with crossover as for 'studio' monitor etc.

Over at diyAudio, Lynn Olsen has been working towards an OB with drivers working in distinct low-mid-high frequency ranges.  My aim has been to use a good wide-range driver, and then augment it in order to extend the LF range. 

Here the use of simple filters ensures that there are minimal audible 'Q' effects due to a crossover.  Of course this can lead to interferences in the radiation pattern, but with drivers which are clean beyond their desired range there is a reduced need for high slopes. 
To me shared output is more 'musical' than when specific frequency ranges are reproduced by drivers having seriously different transduction and radiation characteristics which can quite literally 'chop up' the sound we are supposed to enjoy.

Cheers ........ Graham.

[scorpion]

Hi Graham,

Very interesting and impressive looks ! Will be nice to follow this development.
In the picture what is the collars around the foam fingers ?

/Erling

[Graham Maynard]

Hi Eling.

That is just masking tape used to hold the fingers to the drivers before I screwed the drivers to the baffle.  Surface tape alone would have fallen off.
_______________________________________

Some further explanation of my findings and thoughts are necessary here.

With a crossovered bass-mid-high driver arrangement then 'tweeter at the top' mounting is normally better as it then has an uninterfered radiation at ear height through its working range.

With the gradual underlap used here, I found that the widerange B200 could not be at the top because a sudden mid-high radiation cut-off developed above the axis of the B200 due to cancellation interference induced by the drivers below.  (As with a line of drivers.)
For a smoother response in the vertical axis the underlapped B200 needed to be between the drivers which were augmenting it below 1kHz.  As here.

I also intend to mount the 15"ers sligthly forwards on the baffle, and with a backwards tilt to the final assembly in order to 'raise' the B200 HF and optimise time alignment.

Cheers ........ Graham.

[JoshK]

Cool design. 

I imagine my OB project is going to have a similar baffle layout but with very different drivers.. 

[hurdy_gurdyman]

You can have a flat SPL, but lifeless reproduction.

I have been saying this for years! However, judging from the way those that have heard me seem to think, it's sort of like pissing into the wind... Many (most?) people rely far to much on a frequency response graph over anything else.

Oh yeah, interesting design. 

Dave

[Graham Maynard]

Hi Josh,

Looking forwards to seeing your design.


Hi Dave,

Yup !
So we need the wind to change direction.

If a wavefront is distorted (EQed) before it hits the driver(s) in order for the driver(s) to develop a flat SPL/frequency steady sine response which is modified by secondary driver/baffle/enclosure/line/room responses, then the initial dynamic wavefront timing (which is directly radiated before those secondary driver/baffle/enclosure/line/room responses can develop) will also be distorted.

OBs are not immune to problems, but allow wide-range drivers to  reproduce more 'freely'.

Cheers ...... Graham.

[corloc]

Graham,

I'm glad to see all your hard work with the B200 is finally coming to fruition.    I haven't been playing in the DIY world for several months now, and you have been a very busy boy.  

Chris

[Graham Maynard]

Hi Chris,

I guess it is that never ending 'listen' - 'think' - 'improve' continuum we tend to get sucked into as loudspeakers impose their own characteristics upon our music.

This design developed out of the Full-Range or Single Driver route, and it is that single main widerange reproducer which holds the overall reproduction of this project together.

Perfection is not possible, so I must see whether I can get this sizable effort to *adequately* disappear from my mind's eye. 
It is large, but on a flat (maybe gapped) base these could be pushed right back into a corner when not in use. 
My Ditton 66s, on imovable 12" stands to decouple their bass from the floor/room, are large too, and it is their typically unintegrated 3-way enclosured sound which must be replaced.

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

[Graham Maynard]

I guess doing more to this LS is going to take time - I enjoy it so much I can't stop going through CDs to give them another spin.
The ones which have the best improvement in reproduction are those which were recorded live at a concert, as opposed to in a studio.  These can sound lacking when played through conventional bass-mid-tweeter systems - but so really 'alive' when the B200 has augmentation down into full bass.

Other driver feeds have been separated for individual functions.
The 10" covers B200 reduction in sensitivity below 1kHz.
The 15"s cover BSC.

Values may yet be adjusted due to the balance of reproduction changing with each modification.  The bass drivers still need to be repositioned and some supporting wings organised.

My son says it needs a sub.  He likes boxed bass and monopole pressurisation which makes the room to shake. 
I don't any more, and this is one main reason for me choosing to go OB with larger drivers.

Here the clean LF body thwack and vibe from drums whilst still being able to hear separated notes from LF instruments is cleaner than I've heard from enclosure bass.  (Makes me think of the differences in LF tracking capabilities when comparing elliptical and spherical stylii on the same cartridge bac in the '60s.)

Also I am determined to use a single amplifier only, without sacrificing main driver integrity.

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

[Graham Maynard]

The transformer is now a 500VA 2x 40V component.
The capacitor presently 1,000uF in series with 1 ohm.

Cheers ...... Graham.

[panomaniac]

That's a beast!   That transfo will act as balast for the speaker, at lease. 

[-Richard-]

Hi Graham ~

Nice work!!!!! As always I like your intrepid continuous investigation into what you need to do... and what you have found out... about and from the materials you are working with... in order to release the music from the objects you are using to create it. I like your careful listening habits. That is the ground of this important work we are all doing.

Your new baffles are impressive. Perhaps our work consists of adding to and then taking away until we can arrive at something that blends everything we like about all the various stages of our exploration.

"On its own the underhung B200 is a bit strident, but with crossover underlap from the other drivers to provide the baffle step and sub 1kHz amplitude compensation, it can be transformed.  Thus the driver responses gradually overlap below 1kHz until the B200 sensitivity wins out and holds the image detail as the others provide the body." Beautifully written description of what you think is going on... I was quite affected by the clarity of your thinking here... and I will use your sense of what you think is going on to further inquire into the integration of the Alpha-15A and B200's.

Thanks for sharing, Graham!!! Please continue to inform us of your findings and experiments.

Warmest Regards ~ Richard

[Graham Maynard]

At the weekend I also tried holding a few inches of wood against the baffle edge. 
Used as baffle width extenders the LF improved, but I don't have the space to do this.
As wings this also lifted the LF level a fraction, but tended to 'close in' and sort of 'nullify' the open-ness of reproduction; like compressing the bass 'depth of field'.

This is something I have noted many times before when trying OB mounting - the differential pressurisation at the rear affects forward reproduction, and the only way I found of relieving that pressure with solid wings was to gap them from the baffle.  Given that I am trying for WAF too, I shall need to organise sides having more open construction with the possibility of being covered by material or carpet.


Hi Michael,

Well it is going to take a fair bit of weight in the base for this one - that's my excuse anyway, LOL.

I'll probably make an old fashioned 'breadboard' style base panel layout to split input drive for each driver.  Everything together is getting heavy, so I will make output terminals for each driver to ease separate board removal for carrying.


Hi Richard.

Thanks for your comments.  I have found your descriptions quite revealing too, if not in ways which the objectivists might accept.

Hence I feel that your recent 'Insights' thread reveals 'hands-on' something else not considered by those who think about and aim for a flat SPL/frequency response, as if this is the ultimate goal.

I have always found that if you introduce more that a single inductor or a single capacitor to cross-over between drivers, that reproduction becomes modified not only by the desired crossover response but also the by characteristics of the crossover network itself, because that network independently acts reactively within its own secondary time frame. 
However, you cannot quantify this affectation by measuring with steady sines because the response modification arises and settles out again before any sine wave becomes steady enough for its amplitude to be measured.  Thus an original waveform beomes dynamically modified during music time in a manner which cannot be determined by measuring the resultant (steady sine based) SPL response.
( Right Dave ? )

Your recent direct amplifier connection to the Alpha is removing some of that 'time' shift which causes the lack of reproduction integration with the B200, because the shifts in time are more unacceptable than is more coherent amplitude variation with frequency.  Unfortunately I do not have an Alpha to check this out.

MJK suggests using L+C components for both upper and lower range drivers.
For OB especially, my feeling is that a 'C' ought not be directly connected across a LF driver voice coil when driven by series 'L' (produces a 'box' like sound), nor an 'L' connected across a HF voice coil when driven by a series 'C' (produces a 'horn' like sound). 
At LF a resistor of voice coil DC resistance could be placed in series with the capacitor, or the capacitor not used at all, with the inductor value then being re-adjusted to optimise 'reproduction' for the LS/baffle/ room situation, rather than aiming for a flat SPL response.  The series connected 'C' + 'R' components then being more like a Zobel used to counter driver impedance rise with frequency.
Similarly a resistor in series with the parallel connected 'L' for an HF or wide-range driver will damp any secondary tuning effects.

The efficiency of a Visaton B200 main driver provides scope for a LF augmentation driver to be rolled off much less sharply;  maybe with a single choke plus Zobel working well for the Alpha-15A.

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

[tubamark]

. . . I have always found that if you introduce more that a single inductor or a single capacitor to cross-over between drivers, that reproduction becomes modified not only by the desired crossover response but also the by characteristics of the crossover network itself, because that network independently acts reactively within its own secondary time frame. 
 . . . For OB especially, my feeling is that a 'C' ought not be directly connected across a LF driver voice coil when driven by series 'L' (produces a 'box' like sound), nor an 'L' connected across a HF voice coil when driven by a series 'C' (produces a 'horn' like sound). 
At LF a resistor of voice coil DC resistance could be placed in series with the capacitor, or the capacitor not used at all, with the inductor value then being re-adjusted to optimise 'reproduction' for the LS/baffle/ room situation, rather than aiming for a flat SPL response.  The series connected 'C' + 'R' components then being more like a Zobel used to counter driver impedance rise with frequency.
Similarly a resistor in series with the parallel connected 'L' for an HF or wide-range driver will damp any secondary tuning effects.

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

Graham,

I'm admittedly weak when it comes to understanding non-textbook xo's.

Could you possibly provide a/some simple schematic(s) of what you are describing above, with actual values?  --I'm not sure I'm following all this correctly  . . .  Is the result still 6db (electrical) slopes, or?  Is this essentially the same integration approach described by the Lampizator?
I'm not sure how avoiding phase (time) issues is possible, once any L or C is introduced.

I'm very interested, as I've also been doing Augmentor + small fullrange, with the goal of low slopes and generous overlap for all the reasons described.  Plus it's far less maddening (or costly) to tweak when fewer components are involved.

Thanks for sharing. 
-- Mark

[Mr Content]

Interesting Graham, nice to see others doing some work with the B200. I find the B200 a great driver, been useing it for 2 years now. Good work


Mr C