| griffon |
My theory to be ripped to pieces today is;
taking say a Lowther full range driver like the DX55 or PM4A which have possibly the straightest responses out of the Lowther drivers and placing them in a sealed enclosure which gives them a relative Qtc of 0.6.
Then putting a subwoofer with a matched air displacement to the Lowther (so possibly another lowther, dread the thought of the cost) in the chamber isobarically, and in close proximity to the first cone. the internal cone would have base above say 70hz reduced by possibly even first order crossover to stop internal midrange and hopefully achieving a fairly linear base response if figured out.
benefits
As far as i can tell the base in a horn or transmission design can't be in perfect phase as compared to a sealed or infinite baffle because of the extra travel needed before contact with the ears. infinite baffle bass would be in perfect phase with the mids and treble when using a single driver. Essentially using an isobaric driver set up turns the sealed enclosure into a simulated infinite baffle enclosure in respect to Qtc values in the bass spectrum.
Isobaric is supposed to produce a resonant frequency that is lower than that of the single driver by a factor of .707 in a similar sized enclosure space.
I also wonder whether the decibal rating of the sub base would be increased into perhaps a far more linear frequency response with correct tweaking.
Ideas? |
|
|
| mightym |
Griffon,
I don't even pretend to Know enough to comment on your theory.
I have looked a a similar layout at this link:www.artec-france.com/etudes.php?lang=en
I don't know if this will help or not, but it seems like an interesting idea.
John |
|
|
| mightym |
OOPS, It's the compact system at the aforementioned link
John |
|
|
| DaveCan |
I think in the pic you attached the drivers are to far away from each other and may not work like that??, I'm no expert though...
Dave:) |
|
|
| Henkjan |
| quote: | Originally posted by griffon
Isobaric is supposed to produce a resonant frequency that is lower than that of the single driver by a factor of .707 in a similar sized enclosure space. |
yes, but with a much lower Qb and an f3 that is probably higher or at best the same
for lower bass, isobaric is not the way... (for smaller cabinets it is)
but do I understand you correctly that you want to put another type of driver inside? |
|
|
| griffon |
I am guessing that to put a different driver in would not really work because they would not match in the speed and x-max perfectly, especially given the very fast nature of the fostex.
now if it works better for a smaller cabinet why wouldn't putting an isobaric configuration in a large enclosure drop the base even further than without of the same size.
the aim for this setup I am suggesting would be to attempt to drop the resonant frequency of the full range cone making it even more full range. like closer to twenty something hertz.
the drawing was just to give an idea and is not to scale but you would probably want the internal cone directly behind the external one, cone to magnet. there have been isobaric setups similar to the one I drew though.
In all honesty the theory definitely needs testing because a single cone that could cover a frequency from <30 to 20,000hz would be incredible to hear. |
|
|
| GM |
| quote: | Originally posted by griffon
In all honesty the theory definitely needs testing because a single cone that could cover a frequency from <30 to 20,000hz would be incredible to hear. |
Indeed, the late Babb Lorelei with its 1" (1.2"? I can't remember now) linear stroke could actually do 20 - 20 kHz in only a 1 ft^3 sealed cab at loud enough levels that in a side by side comparison in a fairly large room at a ~16 ft listening distance, a pair was still sounding composed when Jim Griffin's original Linus Arrays were in audible distress.
Supposedly it's being redesigned by the group that bought out Babb, but it's been several years now and no announcement that I'm aware of............
GM |
|
|
| just a guy |
Isobaric loading does nothing to change the resonant frequency of the driver afaik. What it does allow you to do is make the box half the size of a box using the same two drivers in non-iso configuration but it won't go any louder than a single driver since the displacement is the same as a single driver. AFAIK, economy of space and possibly distortion reduction are the only benefits of an iso load.
Now you CAN use a box 2x larger than max flat alignment would be, and you can do that with a single driver OR an iso loaded pair and the results will be the same. The resonant frequency of the driver doesn't change so much as the tuning frequency of the box. You can make the box as large as you like and you can tune it as low as you want, but it won't be flat response, you will end up with an extended bass shelf type response. The lower you tune it the less power it will handle due to increasing excursion demands at lower frequencies.
All of this can be modelled very quickly with winisd and a number of other free programs.
OTOH, once you start playing with drivers that have 1 inch stroke, like GM mentioned (and maybe a touch of eq for good measure) you can force them to do all kinds of tricks.
Based on the questions you've asked in this thread and the other one, and the amount of industry buzzwords and you've been tossing about it sounds like you've been reading the forums I(or reviews) a lot. Some people here are like walking textbooks with built in calculators and some (like me sometimes) are just plain wrong (or purely subjective). It takes a lot of research to figure which is which and forums and reviews (with so many subjective opinions) can be a tricky tool to handle properly. I'd suggest a few good hours with some good software simulation programs and technical papers to gain a firm grasp of the fundamentals (which I have to admit I haven't even sorted out myself - so I should probably follow my own advice).
Anyway, good luck with this very ambitious project. |
|
|
| LineArray |
If you make up a compound driver like you propose
out of two identical ones and wire them in parallel the
following happens:
Moved mass x2
Compliance /2 -> Vas /2
Power Efficiency /2
Impedance /2
Voltage Sensitivity stays the same.
Advantage: You can build the compound driver
into a cabinet of half size, for the price of double
current with same voltage.
Qb does not change, if cabinet is half of the size for
the single driver. No change in bass performance at all,
just a reduced size cabinet.
----------
DX 55 e.g. is a high fs, low Qts, low excursion driver.
The only way to get some bass out of it (if you don't want to
build a horn) is a properly designed bass reflex cabinet or
(better) crossing it with a dedicated subwoofer.
If you want to use 2 ones, a 1.5 way design might be
possible using a lowpass filter for one of them.
Both drivers mounted in a BR enclosure ontop
of each other, both facing the outside air.
This can increase sensitivity in Bass by 6dB without
changing excursion, because radiation impedance
of the drivers add up and you drive twice the power
through them in the low bass range.
The enclosure would have twice the volume as an
enclosure with same alignment for one driver.
Since the XO frequency has to be set fairly low, it is
difficult to manage this with a passive XO.
You need very low resistance inductors not to
increase the Q of the lowpassed driver excessively.
Maybe it is useful to tolerate some resistance in the
inductor and compensate that by adding a series
resistance to the unfiltered driver to tune both to
the same Q ...
An active XO would be recomended to circumvent these
problems ...
Cheers
|
|
|
| badman |
| This won't work with a full range driver. You'll have reflections through the cone and and cancellations and all sorts of oddball effects. Basically, it'll sound like trash. |
|
|
| LineArray |
@ badman
You mean mids radiated into the cabinet and
going through the (lightweight) cone of the
lowpass filtered driver by reflection ?
This will happen, but this happens with a single
driver too, if the enclosure is damped usufficiently.
Build separated volumes then to avoid this.
Then port and volume can be tuned to each drivers
effective Qts individually, while maintaining same fb .
Only disadvantage is ports getting longer, but
should not be problem with alignments possible for
DX 55 e.g.
This way even the problem with the resistance in
a passive XOs inductors can be solved ...
If you fear mids radiated from the port ... well place the
port at the back, build a port with an absorbent lining
or use a passive radiator with a solid diaphragm ...
But this would be a problem of every ported design
driven in the midrange and independent from the kind
of driver used.
Cheers |
|
|
| badman |
There will necessarily be a frequency dependent lag between the drivers no matter what they do. The question was regarding isobaric loading, which means you absolutely cannot separate the volumes. |
|
|
| LineArray |
What i proposed, while taking your issues into account,
is a way to improve bass response using 2 drivers.
Saparated volumes solve the issues you
mentioned before.
As i and others pointed out before, isobaric loading
(compound driver) does not improve bass performance and
is useless for that purpose.
My preferred way to improve bass performance when using
that kind of lowther drivers would be a subwoofer.
Cheers |
|
|
| LineArray |
Consequently a subdivison opens the possibility to
wire a special driver in parallel, which is able
to deliver the desired bass augmentation ...
From this 1.5 way construction it is not far to a 2 way ...
But omitting a low pass for the fullranger may be an
interesting design goal ...
|
|
|
| LineArray |
The term "compound driver" /"compound box" is
established for two drivers coupled by a small volume,
while only one driver is facing the outside air.
What i said concerning bass response using two
lowthers as a compound driver in the established
definition of the term stays valid.
@sreten:
The experiment on the website you posted a link to,
is IMO based on an oversimplified equivalent circuit,
because the authors weren't able to simulate the
"true" circuit of the system and relied on using software
only capable to calculate boxes using a single driver or
a multiple of equal drivers.
Just averaging the parameters will IMO not be sufficient
to predict the actual bahaviour of this type of
"blended driver" in a given box.
At least it would be necessary be to predict excursion vs.
frequency for both of the drivers separately to decide
whether a given design has sane mechanical limits ...
This is only possible by using a proper equivalent circuit
in the simulation, which the authors havent done by now.
As long as this is missing, i feel there is a lot of
wishful thinking in that idea.
But nevertheless i do not want to say it is impossible,
maybe "driver blending" will be a sport of the future. :cool:
But as long as theoretical background is unsure it will not
be a save technique for the average DIY er.
btw. "blended driver" is the term i would suggest for this
approach, since the term "compound driver" is already used
for a different configuration.
Cheers |
|
|
| sreten |
Hi LA,
I agree a more robust engineering approach would have helped matters.
But the idea is "sound" IMO. A low Qts driver in a largish box is so
overdamped even a significantly lower sensitivity driver, but with
higher Qts and excursion would juice up the bass end.
Theorectically I'd expect there to be a particular cabinet division
that would be equivalent to it not being there, one point of the
article being this is the case without knowing the division.
I also agree the maths for the compound driver are a bit of joke,
it is far more complicated, and somewhat different IMO, (I have
not attempted a robust analysis), but it is an interesting idea.
:)/sreten. |
|
|
| LineArray |
All one has to do is to draw the proper equivalent
circuit, representing BR type (or closed) cabinet and
the two different drivers ...
Use a Program suitable for simulation of passive R-L-C
networks and that's it.
Unfortunately i deleted my own program i wrote 15
Years ago, it was running on Atari ST ... port to a
recent system did not seem worthwile to me,
it was written in "GFA Basic" ...
Are there "all purpose" simulation programs, which are
easy to handle besides SPICE ??? or maybe an easy
to use spice derivate ?
Cheers |
|
|
| LineArray |
@sreten:
I do not believe in equivalence possible between a subdivided
enclosure with a single volume enclosure. At least if there is
one driver with higher fs and smaller Vas its excursion below
fs will be different in both approaches ...
But i'd be curious to do a proper simulation for a given
example. Maybe the idea has potential.
But one thing should be made clear:
The design of such a system will in every case be more
complex than designing a conventional BR type.
The theoretical background being more save, then
the next step would be "driver blending" formulas or rules.
But there is one question i'd like to ask before we start
decades of research:
WHICH PARTICULAR ADVANTAGES CAN POSSIBLY OCCUR
USING A SINGLE VOLUME INSTEAD OF TWO SEPARATED ?
If no one has an answer to that,
i will not lift a finger concerning that idea ...
---
Just a thought:
Making a traditional "compound driver"
- cones coupled by small volume,
only one facing the outside-
out of two different drivers "a blended compound driver ?"
would be possible. The coupled drivers would behave like
one single driver at low frequencies, the new parameters
of the compound driver would be relatvely easy to calculate.
Not too far away to which was proposed on that website.
You could simulate an enclosure using existing programs.
This is were we were at the start:
But the "blended compound" drivers displacement volume
is equal to that of the driver having the lower displacement
volume of the two drivers, so its mechanical limits will
limit the whole system ...
This approach could only be interesting when using
a fullranger (more than one?) with a displacement
volume large enough for the purpose.
You can only "tweak" the TSPs of the fullranger,
without changing its max displaced volume.
In case of a Lowther DX 55 e.g. the best result could be
a well designed BR Box with good bass performance but
limited max SPL. Is that attractive ?? Maybe. But the
design has to take cone excursion vs. frequency into
account properly.
Cheers |
|
|
| LineArray |
@griffon:
What about using just a series resistor, to push up Qes ?
Would open the door to more interesting BR
alignments for e.g. DX 55.
Of cause the problem with limited displacement volume
will persist, as it does with your compound box design ...
Too cheap? :bigeyes:
Too simple? :xeye:
Fear resistors spoiling sound quality? :hot:
There are many resistor technologies we could talk about...
Cheers |
|
|
| sreten |
| quote: | Originally posted by LineArray
Too cheap? :bigeyes:
Too simple? :xeye:
Fear resistors spoiling sound quality? :hot:
Cheers |
Hi,
I guess not a popular choice when the point is to be able to use
flea power amplifiers, case of the baby with the bath water.
:)/sreten. |
|
|
| LineArray |
Depends on how much flea power your amp has got ...
Using <4 Watt SET this might be an issue, but even in that
case i would not overestimate that. Same with usual tube amps.
The limiting factor is the ridiculous low displacement volume
with an DX 55 e.g.
Want bass or not ? In the resonance region, the loss in
sensitivity will be compensated by using a more balanced
BR alignment, which is more efficient at low frequencies.
Even 3-6 dB loss in voltage would be compensated easily
in the resonance region and since the drivers impedance
rises at higher frequencies due to voice coil inductance the
effect of a series resistor having 20 .. 30% of the voice coil's
DC resistance is far lower than you might think ...
If you would calculate a
bandwith (octaves) X average sensitivity
the resistor version will probably look far better.
So please "don'panic" !
It is always the word RESISTOR which causes fear. :hot:
But it is just a word and for commonly used solid state amps
it is not an issue at all.
Cheers |
|
|
| LineArray |
Suggestion for the tube amp community:
If you have a high fs / low Qts driver and want to use BR,
push up the Qts using a series resistor Rs until driver is
appropriate for a useful BR alignment , assuming an amp
with 0 Ohm output resistance.
Build and tune the box using a Solid State amplifier with
low output impedance.
Make the test resistor Rs' adjustable, to take the dynamic
output impedance of your tube amp into account.
Adjust the speaker to a certain tube amp / transformer tap
in a way that output resistance Ramp of the amp plus
the reduced series resistor make up the Rs to which the box
is aligned.
Ramp + Rs' = Rs
Any problems ? |
|
|
| sreten |
Hi,
you could rearrange the feedback loop to give output resistance.
:)/sreten. |
|
|
| LineArray |
Hi,
for someone who is into amplifiers,
maybe an interesting way to go ...
Cheers |
|
|
|
