Ok... here's the latest development. I am going to build a new prototype baffle in dual-layer 30 mm MDF, as show in the attached drawing. To increase weight and stability, two 50 mm steel plates are placed at the bottom, adding approx 50 kg of weight.
We will look into adding "Orion-style" wings on the sides, if we can make it look properly.
We will look into adding "Orion-style" wings on the sides, if we can make it look properly.
Attachments
Hi Stig
Wow - 50mm steel plate - sounds cool -
From my experience though - you dont need a steel plate - given you have a decend floor in your room your speaker can couple to.
With some success I try to go the other route and minimise masses that ain't - by definition - compact.
Meaning the magnet structure itself and any masses that you eventually would like to directly couple - without any distance / construction in between.
One other thought to your design and its latest variants.
I don't mind having lost the competition for the "magicians hat" - the differencies would be subtile anyway I guess.
But what I really would like to suggest you try the 21" *without* the side wings.
You will have plenty of SPL anyway above room mode terretory and below - ....
I use a double 15" for example and my OB (two for each side actually) is only roughly half a meter / 20" with a *single* sidewing as I use them in a 90 degree arrangement - so its comparable to your OB dimension.
I found they sound considerably better (kind of lucid dancing through the lower notes) if I do lift them from the floor as well for some cm / 1" as to avoid volume forming behind the speaker.
This you could try by slightly lifting your OB completele from the floor and remove the side wings for the experiment - maybe barce it with a rod instead to keep stability.
This too is not common - as everybody else is trying to get long runs from front to back.
Give it a try - a 21" might do even without a baffle but you loose the positive effect of non circular shapes with regard to edge diffraction.
Michael
Wow - 50mm steel plate - sounds cool -
From my experience though - you dont need a steel plate - given you have a decend floor in your room your speaker can couple to.
With some success I try to go the other route and minimise masses that ain't - by definition - compact.
Meaning the magnet structure itself and any masses that you eventually would like to directly couple - without any distance / construction in between.
One other thought to your design and its latest variants.
I don't mind having lost the competition for the "magicians hat" - the differencies would be subtile anyway I guess.
But what I really would like to suggest you try the 21" *without* the side wings.
You will have plenty of SPL anyway above room mode terretory and below - ....
I use a double 15" for example and my OB (two for each side actually) is only roughly half a meter / 20" with a *single* sidewing as I use them in a 90 degree arrangement - so its comparable to your OB dimension.
I found they sound considerably better (kind of lucid dancing through the lower notes) if I do lift them from the floor as well for some cm / 1" as to avoid volume forming behind the speaker.
This you could try by slightly lifting your OB completele from the floor and remove the side wings for the experiment - maybe barce it with a rod instead to keep stability.
This too is not common - as everybody else is trying to get long runs from front to back.
Give it a try - a 21" might do even without a baffle but you loose the positive effect of non circular shapes with regard to edge diffraction.
Michael
StigErik
Of course I did not want to talk you out of the TPL with my diagram. Just wanted to show where the compromise begins.
I believe it needs a four way system to get a consistantly true constant directivity dipole. Three ways will be good enough but will have a "weak spot" somewhere in the design.
Since stereo is a "horizontal effect" I always concentrate on the hiorizontal pattern first. Controlling and adjusting the vertical radiation pattern is a definite second order affair to me. My EDGE simulation was done for the mikrofone positioned at 1 m height and perpendicular to the baffle, so no effects of changing vertical or horizontal radiation pattern show at all.
And - as Michael already said - EDGE is a bit hysterical, making everything look worse than it will come out in the practical application.
When looking at your latest drawing of the baffle, you may consider to make the rear opening taper for the W22 wider than now. I don´t have any dipole experience for 8" midrange drivers, but with a simple 45° taper smaller drivers tend to develop rear cavity resonances in the 500- 2k region, which show up even in the frequency response at 90°.
Rudolf
CLS said:If you have some spare time, try the T-bass for yourself, please. All parts can be recycled if regret, nothing is wasted. And you'll have a wonderful experience for the time you spend.
Although some doubts out there, I've seen NO unpleasant response from the real T-bass users. Isn't that interesting?
All tech talks and experience sharings have been posted in that thread, so no repeating here.
Just try it. It's a free and wonderful gift.
Hi, always curious - a gave maynards circuit a quick try last night.
Didn't do any tweaking though as I had already laying aroung
- 5mH 0.5Ohm air core hex wound
- a decent 250VA torroid
- some high voltage capacitors that I bypasse with a 220uF mundorf foil
and all the other parts.
My impression was that there in fact is kind of bass enhancment with that circuit - it certainly emphasisies on certain aspects and also giving kind of "box" sound impression to OB.
It would be worth to do some investigation but don't have time - and in the end - personally I preferr the lean and extremly "dancing" sound of pure OB in the lower department (without the circuit).
Just my thoughts and subjective impressions - not to open any "in depth" discussion here (I'd come over to Maynards thread once I'd would like to do)
Michael
It may look like an obvious decision to put the steel plates down to the floor. But if you think about it a bit more, the added mass should be positioned as high on the baffle as possible to keep it from rocking. Best compromise would be to integrate the steel into the cross bar above the 21" IMHO.
I´m with Michael regarding the side wings. Better try with some provisional wings, how much is really needed, before building the real thing.
I´m with Michael regarding the side wings. Better try with some provisional wings, how much is really needed, before building the real thing.
Rudolf said:
The mechanisms involved here are manifold if we go deeper into this subject.
That accelerating the "cone mass" needs some other mass to build up the force needed is obvious.
Lets label that other mass the "reacting mass" to make clear that its movement is necessary but not of our desire in the first place.
Under free suspension condition (hang the speaker on a thread or do our experiments in outer space) obviously only the speaker motor itself can be (more or less) considered to be the compact "reaction mass" – meaning no flimsy structures involved here.
If we would be able to bolt down the magnet structure – not the basket ! - to mother earth – we might consider mother earth's mass - plus the speaker motor's mass – as to be (more or less) the compact "reaction mass"
But now – if we bolt down the basket (lets assume it to be of marginal mass for the moment) to mother earth - we have to deal with the compact "reaction mass" of the motor and the (to some extent) "detached mass" of mother earth as the basket is not ultimately stiff.
This rules are involved with *any* mass attached via some structure.
The resulting "reaction mass" behaviour is complex – depending on the rigidity of the structure the mass attached and multiplied by the count of mass-spring-systems you can differentiate.
Furthermore there is at different SPL and at different frequencies different energy leakage due to internal material friction involved.
Dampening isn't necessarily a pure positive thing for us either.
Basically dampening (transforming energy flow from mechanical material bending / stretching / compressing into heat) dumps part of the energy we would like to radiate as SPL.
To what amount and at what time exactly these energy losses (with respect to the desired cone movement) occur is highly unpredictable given the complex interaction of masses springs and materials usually involved.
(at least this is my guts feeling
– haven't done full investigation on material dampening effects yet)Now - back to the OB at hand.
It's not necessarily our goal to prevent any speakers from rocking.
More precisely – it should be our goal to prevent large radiating areas from *unpredictable* rocking.
Remember, its just a mere and very linear loss in SPL if the speaker (as "reaction mass") *in phase* moves into the opposite direction than the cone !
This is true for all frequencies and also does *not* depend on the amount of SPL !
As the OB at hand is not *very* large, if we compare the OB area to Sd of the 21" driver I'd suggest – if you are willing to increase complexity of your design – to do a cut below the mid and let the 21" rock its part of the OB as much as it happens.
In the freqeuncy band of interest there shouldn't be much of "chaotic" plate movement of the OB either - so no problems form this side as well
This is one of the benefits of OB designs - things can be handled such simple.
If you dont want to do the cut i'd be with Rudolf
Michael
mige0 said:
...
Dampening isn't necessarily a pure positive thing for us either.
..
(at least this is my guts feeling
where did I have my head ? – certainly there is something more precisely and revealing to say about the very nature of dampening from former investigation into that subject
Dampening through material kneading is related to internal friction – as already stated.
Now – there is a distinct difference between different forms of dampening a mass spring system.
- first we have the possibility to transform energy proportional related to the distance (of movement) – which is the case of friction related dampening in general
- second we have the possibility to transform energy proportional related to velocity – which is the case of a cars shock absorber and for the dampening effect to the cone by the motor.
The second case we *don't* have (usually) in speaker (box / structure) building.
The first case has the gross disadvantage to add inherent waveform distortion to the original force over time stimulus.
This is overlooked constantly when people think of dampening as a cure for each and everything.
Michael
The idea with the steel plates was to avoid rocking of the entire baffle. I use a 25 kg sandbag right now. Without it, the top of the baffle sways as much as 1 cm back and forth when the woofer is working hard. With the sandbag, the movement is greatly reduced. When they load a large tower crane with weight to stabilize it, they to this at the base, not far up in the air.... its the same idea here.
The "wings" are there for two things - stiffen the baffle, and creating a U-frame to increase LF sensitivity. And no - the 18" is far from enough at low frequencies, and we dont expect the 21" to be able to output much SPL below 35 Hz either. The speaker is NOT capable of playing Yello or Kraftwerk at more than background listening levels!
I've sold all TPL150H's I had in stock, but more of them are coming in early in May I hope. I could of course try them as well.... I'm a bit unsure if we can cross any lower, but its an interesting idea for sure!
The "wings" are there for two things - stiffen the baffle, and creating a U-frame to increase LF sensitivity. And no - the 18" is far from enough at low frequencies, and we dont expect the 21" to be able to output much SPL below 35 Hz either. The speaker is NOT capable of playing Yello or Kraftwerk at more than background listening levels!
I've sold all TPL150H's I had in stock, but more of them are coming in early in May I hope. I could of course try them as well.... I'm a bit unsure if we can cross any lower, but its an interesting idea for sure!
StigErik said:
There is a distinct difference between balancing static forces (as for the crane example) and balancing dynamic forces (as for vibrations / rocking / movement).
The aim is different as well as I have tried to outline.
A sandbag *will* help - as will the steel plate in your case – but more efficient with respect to more predictable movement of the upper part of your OB would be the position in between of the woofer and the mid.
Think of the OB as of being a weak part - bending at several modes – than it becomes more clear that you would like to place a "calming" down reaction mass at the point between the drivers.
Stiffening by side wings is good but otherwise it would be better (though not so practical I admit)...
Though – anything goes ...
StigErik said:
Having enough SPL in the room mode department is hardly anything you will get from *any* OB.
John Kreskovsky has done measurements that clearly show the effects involved.
As I cant find his page right now – have a look at Rudolf's page where the plot is linked (the very last one)
http://www.dipolplus.de/frameset.htm
Michael
pixelpusher said:
I'm working in the same direction as your thoughts go.
Though wave guide operation (in the sense of directivity control) isn't exactly what you need with an AMT in dipole operation having directivity control "built in" already
*But* besides a possibly slight gain increase, one promising aim *I* am looking for is towards baffle edge treatment in the sense of making the tweeter close to "invisible" with a optimised contour for exactly that.
I'd rather call it "OB contour optimising" if I ever succeed
Michael
mige0 said:
Hi Michael,
A little bit of an intro ; I'm from the SR side of things and a formula that I use to compare various drivers , is this ;
DIF = ( Bl / Re ) divided by ( Mms / Sd )
- Sd is stated in square inches
- This returns an index number ( @ 1 through 8 ) that I use to grade the dynamic behaviour of any particular example under study.
- You'll see that HiFi drivers will typically index as some number below 3 / while SR products will range 4 thru 8 .
- Granted this a pretty primitive formula / and if one closely scrutinizes it, one will quickly find some obvious weaknesses .
- One such weakness is that "Re" is an obvious oversimplification ( of what should be a much more complex factor ).
- This simplification skews results and suggests that lower impedance drivers will have a more dynamic character than comparable higher impedance drivers ( even of the same model type ) .
- FWIW, I'm comfortable with this "skew" as it actually mirrors my SR experience.
- Perhaps one reason for this may be my exclusive use of modern solid state amps ( & their almost negligible output resistance ) .
>< cheers
ps StikEric ; Nice project ! & I apologize for the OT deviation .
CLS said:
CLS - left a answer for you at a better place:
http://www.diyaudio.com/forums/showthread.php?postid=1808235#post1808235
Michael
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.