[Rudolf]

Erling,
how about getting the foam fingers soaked with indian ink? Surely a mess to do, but that should not clog the foam. How well the plastic foam bonds the ink will have to be tested. 

[scorpion]

Hi Rudolf,

Interesting idea, worth a try just for the sake of it. I found out that Pelikan Fount Indian Ink is the best for writing (not clogging) and thus probably also the best for this purpose.

/Erling

[Graham Maynard]

Hi Erling,

When walking into the kitchen this morning the attention of my 'just out of bed' bleary eyes was significantly caught by those foam fingers 20ft away when I passed the open lounge door.
I had thought their light colour could be tolerated, but oh dear they do stand out, and in a way that triggers a need to double-take so to speak.

So I am keen to hear if you try the inking. 

It is possible to buy black conductive foam for the safe storing of loose integrated circuits.  I do have some, but I found it to be harder and more dense than the paint rollers, so I would not wish to use it here.

Yesterday my son turned up unexpectedly with some of his own CDs.
One that I had not heard before was by male vocalist Brian Kennedy and it has some amazing LF on it;  my son just stood there clapping his hands in applause.  He said he had heard that particular track on more than 20 systems before, but this was the tightest, cleanest and most naturally deep sounding bass reproduction that he had heard from it.   

He told me that I must have changed something from last week (the transformer, capacitor, chokes, relative levels etc.) and that it no longer needed a sub because a sub could not provide that same openness.

I reminded him how in the past that the other room doors down the hall would shake if bass was played loudly, and I told him to go and listen to them.  As soon as he walked out of the lounge door he turned round open mouthed - the bass almost literally stops at the door as you walk through it, and of course the other doors no longer shake.
There is some re-radiation from local floor/ceiling vibration, but  no longer the same pressure wave going down the hall as there has alwaysbeen from bass enclosure type reproducers.

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

[dweekie]

If you search on Head-fi, you'll get results on dying foam pads without clogging or changing the feel of the foam, which is very important in headphones.  "RIT dye" seems to be the common one used (http://www.ritdye.com/).

[nodiak]

It looks like a dye (Rit or other) should work out. Also possibly some grill cloth material ? Would need to be nicely cut or folded to lessen seams. Just a thought.
Don

[Graham Maynard]

Well these arrived and have passed testing for OB use.

http://www.eminence.com/proaudio_speaker_detail.asp?web_detail_link=DeltaPro-18A&speaker_size=18&SUB_CAT_ID=1

Considering that these are about the cheapest 18" driver available, with 28Hz Fs out of the box - which reduces after being driven hard - have a DC-R of 5.3 ohms - these do very well at LF on an OB.

I finger softened the rolled cloth cone surround and then blasted with pulses of 12Hz;  resonance now down to 22Hz which is what I want for parallel connection with one Beyma per channel (Fs 34Hz when last checked).

Also the DeltaPro-18A voice coil inductance is like having a free crossover choke, so they already start to self roll-off early when compared to other 15-18" drivers, hence their LF contribution needs less crossover filtering.

I do however consider these to be augmenters for 12-15" drivers, for they will sound 'slow' if used with anything of smaller diameter.  Hence they will be ideal for my intended purpose of extending LF with the Beymas, without increasing lower mid output.

A new 'Spirit' baffle now has to be cut.

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

PS.  Pity the magnet looks as if from a 1950's JukeBox LS, but at least they work well.

[Graham Maynard]

Further comment about the Eminence DeltaPro-18A  (the 18C is a 4 ohm version which could provide cheap increase in voltage driven SPL).

What is so notably missing from reproduction with this driver is the rise in output above 1kHz which is typical of most other large drivers, hence these are likely to need much less correction when augmenting other drivers.

Also, when compared with other drivers via XLBaffle, the DeltaPro18  appears to generate good maximum LF SPL.
Mine do not wheeze or rattle either, and where the Beyma 15" is also mechanically quiet it audibly frequency doubles at 20Hz (the manufacturer's datasheet shows 25dB down, but this is still audible!). The DeltaPro18 shifts much more air, more cleanly and without introducing a tonality either.

T-bass plus 2x DeltaPro on one OB ?  What a thought !  Yet this would work only if running with a 15", for these 18"ers really could not hand over to anything smaller, and then you would be into something really large, so I remain happy with what I now have available to rebuild with.

What I am now particularly looking forwards to is having two different driver mechanical Fs on the same baffle with voice coils connected electrically in parallel in order to smooth the electrical transduction impedance characteristic in a manner which two 'same' or very similar Fs simply cannot.  This 18" in parallel with a 15" will reproduce more cleanly circa 30Hz than could two similar 15" on the same OB.  Also, 'pretty' car based drivers with their heavier cones would be unable to contribute similarly unless separately powered.

Cheers .... Graham.

[gooberdude]

Hey Graham,

I currently own (4) Eminence 15" woofers in use for OB and am curious about how you softenend up
the surrounds.

Can you explain that process to me?     


I like how Eminence builds their drivers to last 40 years in any condition, but they are so tight out of the box.   


matt

[Graham Maynard]

Hi Matt,

I do need to warn that as soon as you soften a surround you not only reduce Fs, but reduce driver Q and increase the VAS as well, which is exactly what I wanted for OB dipole operation, and I am very pleased with the outcome.  However, reduced Q and increased VAS would affect tuned cabinet dimensions.

The DeltaPro18 has a double roll cloth surround - two up/ two down.
Where the roll is up I gently ran my finger nail along it pushing it down along the circumferential distance between two mounting holes.  I carefully repeated this ten times on that roll.
This was separately done ten times for each roll, including outwards finger nail runs from the rear to temporarily reverse the shape of surround rolls which protruded towards the rear of the driver.
Hence 40 finger nail runs per chassis leg spacing, repeated eight times for the number of frame gaps = 320 altogether.  I put screws through the holes to show which gaps had been processed.

I did not do this to stretch the fabric or coating, but to reverse the pressed fibre interlockings as if the driver had been well hammered.

After this the cone moves a greater distance with applied hand force or terminal voltage, and is much better sounding at deep LF, but has lost some upper bass/ lower mid efficiency due to its Q being reduced. 
This however is exactly what I wanted for low bass output because I have other drivers covering upper bass reproduction and higher.

Obviously this cannot be done with foam or rubber surround drivers, which I do not like anyway.  However their pressed centre spiders can also be loosened, though with much greater risks for the voice coil centre becoming offset.

A low Q DeltaPro18 resonance of 22Hz is exactly what I wanted for OB.  This is near inaudible when you tap the cone, whereas the original deep and satisfyingly low 28Hz had been recognisably 'tight' sounding.
If you can hear a cone resonance when you tap it, then you will also hear that same sound superimposed upon music waveforms - as with rubber surrounds, and especially via higher Qes coil/magnet/cone systems which are not capable of exerting as much amplifier damping control.

Cheers ........ Graham.

[dnewcomer]

Hi Graham,
 
 I was curious about your selection of the DeltaPro18, and was wondering if you considered the Beyma sm118n, if so, what criteria led you to select the Delta.
there had been some favorable talk of the Beyma, in Lynn's Beyond the Ariel Thread. would the difference of passive, or active crossover, alter your decision, mine would be active.

Thanks, Derrick

[Graham Maynard]

Hi Derrick.

Yes I did consider the Beyma SM118-N, but I already have the Beyma SM115-N.
These Beymas have the same magnet/voice coils with mechanical alignment X-max of 5.5mm. 

When I saw that the DeltaPro-18A has a thicker pole plate/larger X-max and lower Fs, I decided that it would be beneficial to parallel up so that the drivers have a greater difference in Fs.
Besides, one Beyma is more than sensitive enough above 100Hz and I did not want to run separate crossovers for each LF driver as Lynn suggested.

That Eminence driver can shift some air.

The Beymas seem to have gone for closely folded concertina edging and a very soft relatively small diameter centre spider.  The Eminence has an open double roll edge with stiffer large centre spider.  I am not an expert but I think it is the accordian edge limiting of cone motion which is leading to increased harmonic distortion with the Beyma.

I would still choose this particular Eminence for the low bass, even though there are more 'efficient' 18"ers out there.

Where else could you get an 18" driver which is so easy to take down to 22Hz Fs ?  (They cost the equivalent of $200US ea here in EU by the time we pay the carriage, so not a cheap choice!)

Whether active or passive crossover drive, it won't make any difference when it comes down to mechanical distortion above Fs due to input below Fs, and the 18" Eminence blows the 15" Beyma away in this regard.  And with there being little music below 22Hz ----

Cheers ....... Graham.

[Lazz]

Hi GM wondering from where you purchased the Eminence?$?

[Graham Maynard]

Hi Lazz,

http://www.bluearan.co.uk/index.php?id=EMIDELPRO18A

The DeltaPro18 might seriously disappoint a bass guitarist or musician due to the lack of mid range peak, this being exactly what we want for OB working.
And why Eminence claim only 37Hz as the LF limit is a mystery - maybe in a cabinet.  They go way lower than this on an OB.
Both mine had 28Hz Fs out of their boxes - and can only go lower with use.

Cheers ....... Graham.

[Graham Maynard]

The SPIRIT is being rebuilt.  Click on thumb, then click on image again.




Note the fan blowing outwards for the breathing of dust free air from indoors plus the glass of beer to wet my blow dried whistle; oh, and my 30+yr old (but still working well) jigsaw drill attachment.

Prior to getting to this stage of development there had been many aspects I attempted to tolerate, but which to my ears could not be overcome in the way which so many commercial LS manufacturers suggest are solutions; eg.-
1. Listening to a voice coming from ear level or below whilst seated does not sound as correct to me as one which is reproduced at the mouth height of a standing performer.

2. There is just no way that a kickdrum or the Pink Floyd heartbeat can sound right when reproduced by a small cone area no matter how the driver is enclosured, TLed etc.

Thus for me a fullrange/mid/high driver must be centred circa 4ft above the floor, whilst any OB LF augmentation must be via a large cone area close to the floor. 
My first SPIRIT was an attempt to position the widerange driver lower - as do so many other folk - but it did not work.  I've had the HF from 3-ways at 4ft for 35 yrs now, and hearing voices lower as if from small/short performers just does not sound realistic to me.

3. Separate electrical circuit correction for baffle size with a dipole fullrange driver can 'flatten' its sine measured response and make for a more comfortable listening experience, but this simultaneously modifies the dynamic nature of the sound being reproduced.

Where the wavelength related deficiency is countered with output from an augmenting driver(s) then the original fullrange is allowed to perform at its best - as if the entire composite system then sounds like a single extended range fullrange loudspeaker. 
The reproduction remains basically 'fullrange' though with main driver input no longer being separately modified from t=0 to counter loss effects which do not arise until after transduction and baffle size related wave propagation time periods have elapsed. 
Also, when fullrange drivers are run electrically flat (except for some LF cut to limit cone excursion) they can output to realistic sound levels when the baffle size related loss augmentation is provided by other drivers.
 
I have a box of small fullrange drivers (line and cluster experiments) as well as many conventional 'hi-fi' drivers sitting here unused, as proof of the construction path which led to me now offset fitting the B200s to 6ft tall x 18" wide OBs in order to utilise augmentation arrangements that make it sound better than anything we have had/heard here before.

There will be four other drivers on this panel; two in series connected D'Apollito to make up for reduced B200 output below circa 700Hz, and two parallel connected LF drivers below these for LF augmentation.  Of course the T-bass circuit will again be integral. 
Solid wings are not planned, though it is likely there will be carpet damping stapled inside 6 to 9 inch open side frames or fret cut side screens.  When the fundamental driver arrangements prove okay then decent thickness timber baffles will be cut/shaped, though I would probably need to borrow my son's proper jigsaw to make those, or get someone else to do it properly.

I need to say that when I set out on this fullrange journey I never wanted anything taller than 50" nor beyond 15" wide, but after hearing what this directly driven fullrange can do when augmented on a large panel, then there really cannot be any going back to small or boxed LSs. Originally WAF (A= acceptance) was a major concern for me, but now the sound itself has become so much a major part of WAF (A= appreciation) in a way I could not have forseen !!!

The D'Apollito arrangement for augmentation of a directly driven B200 at frequencies below its 1kHz lift was the only one which made it sound correct.  8" Peerless midbass drivers were not sensitive enough - even when connected in parallel - and so the Fane 10" (Eminence Alpha-10A equivalents) were purchased, and matched well.  Mounting the B200 above other drivers had previously led to an oddly changing vertical pattern with height, as indeed did the
 
10"
B200
15"
15"

on the first version of this Spirit design where I expected the Fane to cover for reduced B200 sensitivity <700Hz, and some low mid from the 2x 15" Beymas to balance baffle size related provision for the smaller drivers above.

I did not commit to buying four of the Beyma SM115-N without first trying them however, and this proved to be a very wise move because they do not have the low Fs as stated in the Beyma T-S specification sheet(!). 
Also two of these 15" mid-bass in parallel produced too much upper bass, and did not quite shift enough air below 30Hz to match B200 sensitivity.  One alone is more than adequate for upper bass provision, and, as an 18" driver could still be fitted to an 18" baffle (just), two Eminence DeltaPro-18A were purchased to further increase LF output.  ( I do not remember seeing these listed last year when I was looking through driver lists.)

The DeltaPro18 turns out to be a good OB woofer!  Both mine had an Fs of 28Hz out of their boxes, and this can only become lower through usage.  Finger working of the surround quickly had one of mine down to a low Q 22Hz resonance with its impedance peak reduced below 60 ohms.
I would not choose to use two of these together on one OB however unless they hand over to drivers which are clean at upper bass frequencies;  two could prove bass heavy with anything other than a good 15" or two 12", though of course we can all have different ideas about what is 'best'.  Hence I am looking forwards to running the 18" and 15" drivers in parallel, with their different resonant frequencies further reducing the resultant resonant Q.

Regarding tweeters - any addition might be rear mounted and possibly upward pointing for ambience - as Gilbert Briggs did some 50 odd years ago with the Wharfedale Baffle loudspeaker.  I still remain undecided about the value of separately front mounting (ie. interference beaming) a tweeter when a widerange driver already commands the integrated reproduction centre position.

....... the SPIRIT is being rekindled........ more shortly ......... Graham.

[Graham Maynard]

Already rockin' much better than before.



Click, then click again.


Cheers ...... Graham.

[Lynn Olson]

Hi Graham, I've been following your "T-Bass" circuit with great interest. I'm on friendly terms with two transformer builders (Bud Purvine and Dave Slagle), so I don't think it will be a big problem getting custom iron done - plus, I plan to use transformer attenuation for the HF horn anyway, and that will take a tapped transformer with 1 dB taps to do correctly anyway.

I'm having trouble puzzling out the schematic. Does the "T-Bass" circuit increase the source Z only at low frequencies, thus effectively raising the Qts of the driver without the negative sonics and efficiency penalty of a series resistor? If this circuit would make a separate bi-amplifier for the LF array unnecessary, I'd be very happy.

Looking over at the "20 Hz to 20 kHz Dipole" thread, the discussion of what is "enough" bass brings to mind the difference in Transatlantic construction techniques. Houses in the UK tend to be smaller and built of more rigid materials (brick facing, etc.), while US houses are bigger and almost always built with wood framing and sheetrock walls - and LF goes right through them to the outdoors.  Thus, the need for more bass with US construction, to offset the losses. A few troglodytes do their hifi thing in basements, but I don't, so I have to deal with the LF losses as they are.

[Graham Maynard]

Hi Lynn,

Not done any more on my LS recently.  Health prob is intruding so just trying to get the words down.

The T-bass works so well with easily obtainable 500VA 2x 40V toroid that I cannot see much advantage doing self winds.  Best to get an idea of whether the circuit is liked before doing specific designs, which can only be better when the primary is not there.
Large low resistance power transformers are more than adequate for the lows, though it does take experts to multiwind for the highs, where that is the best way to adjust for different sensitivity levels.

Driver impedance develops during waveform time.  It starts out as voice coil R but becomes the peak-flat-rise from low to high frequency during one, two or more half cycles.
The series tuned C+L on the 'earthy' end of the transformer does exactly the same, and it modifies the step-up in a similar manner in time, though is still in series with the unavoidable driver impedance.

It counters driver resonance induced tonalities.  Indeed I first tried it with a LF driver in a resistively vented box, and there was an amazing improvement, but there was no getting away from the sound of the box.  Then when I tried it on an unmounted driver it clearly contributed to overcoming the losses which make LF via an OB so hard to implement.

We might have 2x brick walls with cavity here in UK, but we also tend to have plasterboard ceilings and wooden board floors which are grossly susceptible to monopolar pressure energisation.  Radiation from large driver OB can still shake you, but does not set-up the same pressure energised ceiling/floor abberances, and makes for much less thump-thump bass both inside and outside of the listening room. 

Bet you already know most of this anyway.

You ask about impedance seen by the driver, but we can only measure this after one or more full cycles have elapsed.  We can *hear* the time related aspect even though we cannot measure it as a stable quantity whether real or imaginary because it changes with waveform change.

I once did a simulation of impedance as 'seen' by the driver, and it appeared higher around driver Fs, and so might reduce resonant development, but, what of the phase ? and does it actually counter driver resonance ? such that it is the T-bass which dominates the driver boost no matter what Qes it has;  Qms being a different matter.

Reproduction via the 'T'-bass increases LF beyond that which the driver is naturall capable, and actually counters the driver's tendency to resonate.  Thus LF reproduction really is very much improved - especially the transduction of dynamic waveforms where current and voltage amplitude or SPL simulations are completely useless, as I'm sure you also appreciate.

I did start another thread at -
http://www.diyaudio.com/forums/showthread.php?s=&threadid=130679

Maybe someone else will soon be able to explain better than I can, but it is the transduction dynamics which are important here, and I feel these sound as if more integrated than what might be illustrated via a steady sine plot.

Regarding driver impedance; I found the two different sized low Q drivers in parallel to perform much better than two of the same 15" drivers.

I had old and new (humidity resistant coating) 15" Beymas.  The coated one had increased mid output.  The plain driver was more 'comfortable' to listen to due to cone loss roll-off, but lacked detail.  The newer ones do need more careful Zobel correction though, this to prevent raised output level interference with the mids.  I prefered the coated cone.
Maybe the 18" Eminence I have could do with some coating too, though it has a higher coil impedance and is there for lower bass only.

Cheers ........ Graham.

[Graham Maynard]

Further to my last post I decided to check the transformer voltage waveform versus the input voltage drive waveform.  This at 170Hz.

http://www.diyaudio.com/forums/attachment.php?s=&postid=1626033&stamp=1223302276

This image shows clearly how the tuned L+C transformer 'grounding' node introduces phase change which results in delayed voltage reduction by the 'T'-bass circuit, such that the driver cone (mass) can better current driven during first cycles.
Currents in mA shown are;-
Blue, via a 9mH/68uF capacitor (slow bass),
Black, direct to LF driver (slight loss of driver force generation)
Red, output from 'T'-bass (slight increase of force to coherently overcome cone inertia)

The leading drive can actually be increased by too much - just need to get the tuning and damping correct by 'ear'.

This is why the 'T'-bass circuit is so good regenerating drum sounds via OB.

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

[jeffac]

Hi Graham,

I thought it best I move discussion over here to spare Painkiller any more OT and as it is your Spirit ideas that led me to try the Saba's with the B200.

Thanks for more ideas. I've been trying whatever I can to LEARN first hand.  I've had the small inductor for the B200 on my 'to try' list for a while based on your previous advice and with the Saba tweeter playing so strongly, the B200 HF is not needed so much.  You can't see it clearly in the pic, but the B200 has a phase plug my father cleverly turned of some old wallnut, very hard timber and dark so hard to see, hence the horizontal paint rollers. Even just taped in place, they do help tame the 'discussed to death' aggression in the B200 upper mids really well, great idea.

Currently the T'-bass is using a 300VA 25-0-25, 660 uf caps in series with 0.47 ohm resistor, low passed with 0.56 mH and 4.7 uf cap in parallel. So the Alpha's are playing up into the lower mids to assist the B200 here. I've tried using an extra 0.68 mH and really like this too as the Alpha's seem smoother and as it centres vocals higher and lessens the slight impression of being swallowed. I will try this inductor on the lower Alpha only. This might alter phase etc but fortunately I'm naive enough yet as to not to know better  and only experience to gain if it turns out not such a good idea.

Yep .. I've tested extending the baffle width using the Alpha boxes and as expected there's more bass, not sure about better bass? Currently I have some 25cm wide MDF taped on the sides and angled back ~45 degrees. Not worrying about this too much until get the second Charlize T-amp, all 11W into 4 ohm of it, hooked up for bi-amping. This little amp seems quite robust as I've been running the MTM all linked in parallel on one channel and the parallel Alpha's with T'-bass on the other. The Saba's are nominally 5 ohm so total impedance must be quite low. Not ideal, possibly stupid I know, but it has allowed me to quickly get a sense of the potential of this driver combination, which from early impressions I think is wonderful and hopefully will be even better once I get the tweeter better integrated and the bass weight/SPL sorted.    On that, I'm keen to try some more T-bass circuit values as I'd like to get a little more punch, suggestions?

And regarding your Spirit Mk11 design, how are you dealing with the B200 being positioned so high, do your critical listening standing.  I had my baffled slanted back slightly but vertical is much better and I've got a feeling that it would be even better still with the Saba mid at the top slanted slightly downwards. Lots to consided in baffle shape and construction and driver positioning and mounting and I'll try to incorporate some of these into my next baffle.

What do you think about this concept of a curved wing for the bass drivers that would be something between a baffle extension and a short U-baffle? Maybe made of something acoustically neutral like carpeted thin ply or even heavy backed carpet tiles. Just pondering and doodling as I like to do.



cheers.. jeffac

[Graham Maynard]

Hi Jeffac,

Apologies for my delay in replying.  So many aspects you raised here, I decided to wait until I could answer properly.

The B200 with phase plug might still be better with vertical foam fingers;  the lower one set very close to or even touching the phase plug;  this to disperse some of the normal axial output, also to degrade surface wave symmetry.

T-bass and Alphas.  This is where I would suggest reversing the lower Alpha and aim it into some soft damping material.
The more chokes we insert in series with a woofer the 'slower' the bass becomes, and the more difficult it is to integrate reproduction with that of the next driver.  You could then try with the Alphas connected in parallel ( with wires to the lower one reversed ) or connect the lower one directly to the T-bass transformer.  Reverse mounting a driver can be like inserting a choke, but without modifying the electrical phase of its drive waveform.

Angled wings I have always found to have less audible pressure reflection impact upon the forward output than any sides which are perpendicular to the baffle unless they are equal sized both back and front.  Better than solid sides I have found to be rolls of carpet offcuts which damp instead of reflect at the baffle edge, and which still extend the front to rear recombination path distance.  Even ordinary soft seat cushions held either side with the baffle on the cushion centre line can much improve LF, and this of course without introducing any unwanted frame reflection/ tuning peaks circa 200 to 400Hz.

Punch.  That needs power ! 
An amplifier to drive two 15" Alpha drivers really needs to be about 100W to realise dynamic capabilities.  Punch is the first cycle 'hit', not the sustained note, and whilst you won't be rattling the Alphas with 100W of continuous drive, this is what you need to achieve full bodied drive without running into clipping distortion.
The other thing you need for real punch is low impedance wound components and low Qes drivers.  The Alphas are not low Qes, which is another reason for keeping their electrical connection as phase coherent and direct as possible ( without series filter chokes ), as by reversing them and using damping material mounted on axis.

B200 height ?
You ask how I cope with this, whereas I could ask others how they cope with having their mids/tweeters so low !

From the pics you can see that the first 'Spirit' had the B200 in a lower position ........ it did not sound right !

Since the 60's my 'in-room' mids and highs have been coming from standing *mouth* height.  I've tried lower and higher, but to me performers, violins, trumpets etc. simply do not sound right unless their reproduction source is at the correct height within my room(s).  Maybe when the ceiling and floor reflections are incorrect my sub-concious recognises this, and tells me the resultant is wrong.

The new Spirit gets the 'sound source' back to where the artists and instruments would be if in my room in real life, and thus a subconcious 'something is wrong' reaction is not triggered.
 
Also the MB-WR-MB line up has a very gradual change with height between standing on tip-toe, or crouching lower than sitting, at the listening distance, which was more noticeable with the first test baffle due to the differences in low mid characteristics between the 10" and 15" drivers.

Curved wings ?  Yes a curved baffle would be fine to my ears -
http://www.kef.com/products/MUON/
check out the 'gallery'.
http://uk.youtube.com/watch?v=PsbXpQFemHw&NR=1
( Still sounds as if lacking T-bass 'punch' though !)
http://www.audiojunkies.com/blog/209/kef-muon-speakers

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