well...tell me if im wrong but the sound radiation is a vector? hence, to my mind at least, equal amplitude(force) AND opposite direction(or phase) are required for near perfect cancellation.
Attachments
In the bass range where the cones are very small compared to wavelength
only the displaced volume counts. There is no "direction", so the cancellation
works also with that 90 degree orientation, given that the distance is
small compared to wavelength too.
And that monopole radiation at the lower end is de facto the
monopole radiation from the box ... you can disconnect
the OB driver for low frequencies if the baffle is small and
cone velocity is identical (given the infinitely large box model for
the "cancellation driver").
If the (open) baffle is upsized the cardioid radiation will
be maintained towards lower frequencies. With the baffle going
towards infinite size you will have radiation into the front halfroom
and no radiation into the rear halfroom. Transition is continouusly
and the frequency where a certain intermediate radiation pattern
between monopole and cardioid appears is dependent from the
baffle size.
I know that the manufacturer has also a subwoofer with that
"OB in front of box technique". They argue with better "transient response".
They do not claim it to be cardioid themselves. If i was to improve
"transient response" i would save the cost for the OB, the additional driver
and the mounting. I would invest the money and the box space saved into
a conventional BR or closed subwoofer with improved performance.
But i feel the "Tekton" design discussed in here is interesting and
can be further modified.
I see the problem here.
To completely cancel the rear wave, you need complete coupling of the "canceller" driver to the rear of the OB driver. In other words, a tube connecting the two. This is, of course, a classic "isobaric" configuration.
Without it, you will only get, at best, partial cancellation at higher frequencies, and reducing cancellation as the frequency drops. A few moment's work with Falstad's "ripple tank" simulator will illustrate this.
Surely you mean lower frequencies, with increasing cancellation as frequency lowers...
The drivers, when any distance apart, will always be a bit out of phase. I expect this could be fixed with time delays etc. The lower the frequency, the longer the wavelength, the less out of phase the rearward driver is, because the distance between the drivers becomes increasingly titchy compared to the wavelength.
This design, from what I can see, is effectively an OB iso-braric design, but the coupling isn't as good.
As I see it, when the rear-wave is cancelled, it's only the front-wave of the OB driver giving any output, because when the rear speaker moves into the box, it creates low pressure. When the OB speaker moves towards the box, it creates high pressure. Those two cancel to give radiation from the front of the OB speaker only, and pressure inside the rear box.
Let's get back to the Tekton. I, too feel it could be modified, (firstly by sealing the port, but we know where that went), perhaps playing with the port tuning is in order, as there's a measured midbass bump...
Chris
OK, if we build a compound driver by connecting the OB driver and the box driver using
a tube, we have a conventional (closed or BR) box. A compound driver is from outside
the box not distinguished from a single driver. The new driver will have half Vas, same
Fs, double moving mass etc.
Before running in a circle, we should ask the question what the "Tekton" design may be
good for. Since i am not the designer, i do not have to be its advocate.
It does not make sense to ignore, that with relative dimensions of box and OB given from
the "Tekton" picture, the whole speaker is at the lower frequency end dominated from the
radiation of the box.
The discussion, whether the radiation comes from "the front wave" of the OB driver
- because "the back wave" is cancelled - or from the box is rather fruitless.
The keyword here is superposition IMO: We have a dipole source (OB) and a monopole
source (box), which are superimposed.
Which one of the two sources is dominant at low frequencies can easily made clear
by removing one of the two components.
If you remove the dipole source at LF the output will stay more or less the same in
power and also in radiation pattern, which is nearly a monopole.
If you remove the monopole source (box), power will decrease dramatically and the little
amount of radiation left will have dipole radiation pattern.
If a system consists of two components one if which may be omitted without significant
change in behaviour, common sense and science as well conclude the system to be
dominated - in the field of behaviour investigated - by the component which cannot be
ommitted to produce the observed behaviour.
Ergo: If the design is good for "something", this "something"
- reagrdless of the intention of the designer - will not be a significant change
of bass performance by adding a small OB in front of a monopole box.
A change in radiation pattern can only occur at higher frequencies, where the
dipole radiation becomes comparable in sound pressure to the monopole.
This will typically be from mid or upper bass towards the midrange. At that frequencies
the dipole radiation is strong enough, to change the overall radiation pattern
of the system by superposition of dipole and monopole radiation, resulting
in a cardioid.
The total system does not act like an infinite (open) baffle, it does not
even act like a large (open) baffle. From the radiation point of view it acts
pretty much like the box with the OB part dismounted at low frequencies.
I cannot even understand the argument, OB baffle size not being important
because "the rear wave is cancelled". If so, you could omitt the OB and
place a naked driver somewhere in front or above a box claiming
a change in bass performance. Again obviously this is not the case.
It is determined by the (open) baffle size, at which frequency, call it feq, the dipole
and the monopole component will be of same magnitude, making up a cardioid.
The baffle size in fact is the only parameter determining feq as long as you feed
both drivers with the same voltage.
The larger the (open) baffle, the lower feq will be. Since feq is well above the
resonance of the OB driver AND the boxed driver as well
(a closed boxed driver will have higher resonance than the OB driver),
there will be small difference in phase between both drivers at feq.
The distance of the drivers itself will not cause a significant phase difference
at feq, since the distance is small compared to wavelength at feq, when mounted
closely.
The synthesis of a cardioid by superposition of a monopole and a dipole is often shown
in acoustics textbooks. It is nothing new or mysterious.
Maybe someone in here knows a good english textbook to be cited ...
On Linkwitzlab you can even find a spreadsheet for calculating the radiation of a
dipole dependent from path length and compare it with a monopole on the same sheet.
Nearly everything has already been done, to describe a system like the
"Tekton". There is no place for mystery IMO.´
Regards
a tube, we have a conventional (closed or BR) box. A compound driver is from outside
the box not distinguished from a single driver. The new driver will have half Vas, same
Fs, double moving mass etc.
Before running in a circle, we should ask the question what the "Tekton" design may be
good for. Since i am not the designer, i do not have to be its advocate.
It does not make sense to ignore, that with relative dimensions of box and OB given from
the "Tekton" picture, the whole speaker is at the lower frequency end dominated from the
radiation of the box.
The discussion, whether the radiation comes from "the front wave" of the OB driver
- because "the back wave" is cancelled - or from the box is rather fruitless.
The keyword here is superposition IMO: We have a dipole source (OB) and a monopole
source (box), which are superimposed.
Which one of the two sources is dominant at low frequencies can easily made clear
by removing one of the two components.
If you remove the dipole source at LF the output will stay more or less the same in
power and also in radiation pattern, which is nearly a monopole.
If you remove the monopole source (box), power will decrease dramatically and the little
amount of radiation left will have dipole radiation pattern.
If a system consists of two components one if which may be omitted without significant
change in behaviour, common sense and science as well conclude the system to be
dominated - in the field of behaviour investigated - by the component which cannot be
ommitted to produce the observed behaviour.
Ergo: If the design is good for "something", this "something"
- reagrdless of the intention of the designer - will not be a significant change
of bass performance by adding a small OB in front of a monopole box.
A change in radiation pattern can only occur at higher frequencies, where the
dipole radiation becomes comparable in sound pressure to the monopole.
This will typically be from mid or upper bass towards the midrange. At that frequencies
the dipole radiation is strong enough, to change the overall radiation pattern
of the system by superposition of dipole and monopole radiation, resulting
in a cardioid.
The total system does not act like an infinite (open) baffle, it does not
even act like a large (open) baffle. From the radiation point of view it acts
pretty much like the box with the OB part dismounted at low frequencies.
I cannot even understand the argument, OB baffle size not being important
because "the rear wave is cancelled". If so, you could omitt the OB and
place a naked driver somewhere in front or above a box claiming
a change in bass performance. Again obviously this is not the case.
It is determined by the (open) baffle size, at which frequency, call it feq, the dipole
and the monopole component will be of same magnitude, making up a cardioid.
The baffle size in fact is the only parameter determining feq as long as you feed
both drivers with the same voltage.
The larger the (open) baffle, the lower feq will be. Since feq is well above the
resonance of the OB driver AND the boxed driver as well
(a closed boxed driver will have higher resonance than the OB driver),
there will be small difference in phase between both drivers at feq.
The distance of the drivers itself will not cause a significant phase difference
at feq, since the distance is small compared to wavelength at feq, when mounted
closely.
The synthesis of a cardioid by superposition of a monopole and a dipole is often shown
in acoustics textbooks. It is nothing new or mysterious.
Maybe someone in here knows a good english textbook to be cited ...
On Linkwitzlab you can even find a spreadsheet for calculating the radiation of a
dipole dependent from path length and compare it with a monopole on the same sheet.
Nearly everything has already been done, to describe a system like the
"Tekton". There is no place for mystery IMO.´
Regards
Last edited:
Oliver,
This make sense to me. The problem with internet is that there are so many opinions. Some things have to be tested, people can and must live with certain compromises.
What I'm aiming at though, is to have cardiod response in the bass with fairly small and unobtrousive boxes. Looking at the quote above, it seems impossible...
Peter
This make sense to me. The problem with internet is that there are so many opinions. Some things have to be tested, people can and must live with certain compromises.
What I'm aiming at though, is to have cardiod response in the bass with fairly small and unobtrousive boxes. Looking at the quote above, it seems impossible...
Peter
An OB based design a la Tekton will need a large baffle area
if the cardioid pattern should work for really low frequencies.
If you really want that directivity in the deep bass, you
maybe should consult
Music and Design
and look for U-Frames and damped U-Frames
Such a design is more compact IMO, since the path length
needed is built into one dimension only.
But again, i do not think that the Tekton idea is bad, it is
just mainly monopole at the low end, when built tiny.
Regards
if the cardioid pattern should work for really low frequencies.
If you really want that directivity in the deep bass, you
maybe should consult
Music and Design
and look for U-Frames and damped U-Frames
Such a design is more compact IMO, since the path length
needed is built into one dimension only.
But again, i do not think that the Tekton idea is bad, it is
just mainly monopole at the low end, when built tiny.
Regards
The front-wave pressure from the dipole would be equal to the pressure from the monopole, if it wasn't being cancelled by the rear-wave.
Having built a sealed-box version of the design, I can say that, when the drivers are seperated, there is less bass output than when they are mounted together. I tested this by making a sealed box, and a baffle to sit in front of it. I wasn't the only one to observe this effect.
I don't think this discussion is getting us very far.
Chris
No, I mean what I said. Cancellation will reduce at lower frequencies.
As the frequency is reduced, the OB driver becomes less effective as a dipole source because the front and back waves bend around the "baffle" and cancel each other ("short circuit"). At the same time, the "baffle" becomes less effective as a barrier to the wave from the rear speaker, and its wave propagates out into space instead of cancelling the back wave from the OB driver.
You're also trying to help a driver that doesn't want that sort of help. Even if you do effectively cancel the back wave, you're also halving the radiation resistance of the driver. You then have to add filters to fix the frequency response. It will lead to higher excursion at low frequencies, which is always a bad thing. You're much better off going to a classic "OB with sealed sub on a common baffle" arrangement.
This arrangement of canceling of the back wave of an Open Baffle system with the wave of a closed-box system "sounds" like something that I would think to do.
However, I am skeptical. I think that instead what occurs is that the total acoustic output of the system is simply the summation of the acoustic radiation of the two drivers. As the OB system without any active compensation and a narrow flat baffle has little if any bass output, the bass output of the closed-box system very obviously supplements the bass range output.
Think about it. In the bass range, both drivers are radiating into full (4 pi) space. The wavelength of sound of the frequency 100 Hz is about 3 meters. So the flat baffle that one of the drivers is mounted is not an appreciable obstruction to radiation by the driver back-loaded by the closed- box.
Reproduction in the bass-range is almost completely that of the closed-box system.
Regards,
Pete
However, I am skeptical. I think that instead what occurs is that the total acoustic output of the system is simply the summation of the acoustic radiation of the two drivers. As the OB system without any active compensation and a narrow flat baffle has little if any bass output, the bass output of the closed-box system very obviously supplements the bass range output.
Think about it. In the bass range, both drivers are radiating into full (4 pi) space. The wavelength of sound of the frequency 100 Hz is about 3 meters. So the flat baffle that one of the drivers is mounted is not an appreciable obstruction to radiation by the driver back-loaded by the closed- box.
Reproduction in the bass-range is almost completely that of the closed-box system.
Regards,
Pete
Thank you all, gentlemen, for your idea and theories. I think I've learnt quite a lot from this discussion.
However, instead of arguing based on the conventional text book theories (I am sure there are quite a number of creditable theories out there so there's no definite rights or wrongs), why don't we try to have an experiment on where the bass is from the cancellation driver or the OB?
Line array had came up with a method, and I can suggest another easier one.
We can close-mike (near field measurement of SPL) the SPL of space between the cancellation driver and OB and the SPL of the front OB respectively when a low frequency signal (say 150Hz) is played.
In this experiment, we should use a relatively small OB (say 6" wide).
This way we can know where the bass is from.
BTW, Chris, can you please tell me how your Visaton OB/cancellation driver setup sound like? I am pretty interested in it.
However, instead of arguing based on the conventional text book theories (I am sure there are quite a number of creditable theories out there so there's no definite rights or wrongs), why don't we try to have an experiment on where the bass is from the cancellation driver or the OB?
Line array had came up with a method, and I can suggest another easier one.
We can close-mike (near field measurement of SPL) the SPL of space between the cancellation driver and OB and the SPL of the front OB respectively when a low frequency signal (say 150Hz) is played.
In this experiment, we should use a relatively small OB (say 6" wide).
This way we can know where the bass is from.
BTW, Chris, can you please tell me how your Visaton OB/cancellation driver setup sound like? I am pretty interested in it.
I think this could be summed up by saying that trying to get one driver to cancel the backwave of another in the suggested "loose coupled" configuration is like herding cats. Those pesky waves just don't go where you want them to.
Chris and Boris, if you provide some driver T-S parameters and baffle dimensions, I can model it for you in Akabak to show you what would happen. An initial try with Tang Band 3 inch drivers on a 20 x 30 cm baffle appeared to show that the disadvantages outweigh the advantages. The main disadvantage is phasing between the two drivers because they have different path lengths to the listening position. This causes severe dips and peaks in the frequency response in addition to the normal OB dipole peaks and dips. To get around this you would need to low-pass the drive to the "canceller" speaker, limiting its contribution to the bass region. And if you're going to do that, you may as well put the driver on the baffle in the same plane as the OB driver. That way you get the bass enhancement without the phasing problems.
Chris and Boris, if you provide some driver T-S parameters and baffle dimensions, I can model it for you in Akabak to show you what would happen. An initial try with Tang Band 3 inch drivers on a 20 x 30 cm baffle appeared to show that the disadvantages outweigh the advantages. The main disadvantage is phasing between the two drivers because they have different path lengths to the listening position. This causes severe dips and peaks in the frequency response in addition to the normal OB dipole peaks and dips. To get around this you would need to low-pass the drive to the "canceller" speaker, limiting its contribution to the bass region. And if you're going to do that, you may as well put the driver on the baffle in the same plane as the OB driver. That way you get the bass enhancement without the phasing problems.
Here's the driver I used
Visaton - Lautsprecher und Zubehör, Loudspeakers and Accessories
7" wide baffle, 12" tall. Box behind it was a 7" cube, sealed.
Crossover was 400Hz (second order, had some components lying around)
Chris
Edit - it sounded lovely on anything considered intimate music. Excelled at girl+guitar. They were capable of playing rock, but not loud enough to be good.
Visaton - Lautsprecher und Zubehör, Loudspeakers and Accessories
7" wide baffle, 12" tall. Box behind it was a 7" cube, sealed.
Crossover was 400Hz (second order, had some components lying around)
Chris
Edit - it sounded lovely on anything considered intimate music. Excelled at girl+guitar. They were capable of playing rock, but not loud enough to be good.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.