OK,
I simulated three setups using Visaton FRS8.
Setup 1:
A standard OB with the driver in the middle of a 30 x 17.5 cm baffle.
Setup 2:
The same size baffle with a 17.5 cm cube (5 litre internal) behind its lower half, driver facing up just below and behind the OB driver. A 400 Hz second order LP filter was added.
Setup 3:
As for Setup 2, but the box driver moved to the front face of the box (on the same baffle as the OB driver, facing forward).
In the first attached graph:
The black curve is the response of the standard OB.
The red curve is the combined response of the OB with "cancel box".
The green curve is the response of the "cancel box".
In the second attached graph:
The black curve is the response of the standard OB.
The red curve is the combined response of the OB with "cancel box".
The green curve is the combined response of the OB with "bass assist". This could also be called a "1.5 way" design.
My conclusions:
The standard 1.5 way design with both drivers facing forwards will perform better than the "rear canceled" design. Optimisation of box volume, crossover frequency and driver spacing on the front baffle will produce even better results.
Attempting to cancel the OB rear wave is pretty much futile. At the frequencies of interest there's not enough coupling between the drivers to be useful. Any significant cancelation will also reduce the radiation resistance of the OB driver, and thus reduce its efficiency and frequency response. You can't afford to do that with a 3 inch driver.
I simulated three setups using Visaton FRS8.
Setup 1:
A standard OB with the driver in the middle of a 30 x 17.5 cm baffle.
Setup 2:
The same size baffle with a 17.5 cm cube (5 litre internal) behind its lower half, driver facing up just below and behind the OB driver. A 400 Hz second order LP filter was added.
Setup 3:
As for Setup 2, but the box driver moved to the front face of the box (on the same baffle as the OB driver, facing forward).
In the first attached graph:
The black curve is the response of the standard OB.
The red curve is the combined response of the OB with "cancel box".
The green curve is the response of the "cancel box".
In the second attached graph:
The black curve is the response of the standard OB.
The red curve is the combined response of the OB with "cancel box".
The green curve is the combined response of the OB with "bass assist". This could also be called a "1.5 way" design.
My conclusions:
The standard 1.5 way design with both drivers facing forwards will perform better than the "rear canceled" design. Optimisation of box volume, crossover frequency and driver spacing on the front baffle will produce even better results.
Attempting to cancel the OB rear wave is pretty much futile. At the frequencies of interest there's not enough coupling between the drivers to be useful. Any significant cancelation will also reduce the radiation resistance of the OB driver, and thus reduce its efficiency and frequency response. You can't afford to do that with a 3 inch driver.
Attachments
Last edited:
Don, those graphs looks pretty ragged, is that the problem of the Visaton or the problem with the OB and cancellation driver, or both?
Thanks very much for your info, I think it's just impossible to cancel the OB rear wave completely provided that the real life situation is the same as the theoretical graphs .
However, I may still wanna test a pair of OB/cancellation driver speaker. High fidelity is not all about science, after all. No one's even attempting to get ruler-flat response from full range speakers either.
Regards,
Boris
Thanks very much for your info, I think it's just impossible to cancel the OB rear wave completely provided that the real life situation is the same as the theoretical graphs .
However, I may still wanna test a pair of OB/cancellation driver speaker. High fidelity is not all about science, after all. No one's even attempting to get ruler-flat response from full range speakers either.
Regards,
Boris
Physical speakers will have a little less raggedness than the simulation, but all dipole speakers are like that. Linkwitz explains it very well at Linkwitz Lab - Loudspeaker Design.
Do feel free to try it, though. I suggest you build the box as a cube, and the OB as a section that sits on top of the box, so that one face of the box forms the lower half of the baffle. That way you can try it with the box driver facing up behind the OB driver, then rotate the box 90 degrees so the box driver faces forward with the OB driver.
Some random thoughts:
The aim is to cancel the back wave to improve the frequency response in a way that won't affect the "boxless" sound of an OB speaker.
One problem is that part of the characteristic sound of an OB speaker is due to its dipole function. If you magically cancel the back wave, you create a monopole which will sound quite different due to different reflection patterns from the room surfaces, as well as elimination of the dipole comb effect.
But if you're certain that cancelling the back wave is the way to go, the most effective method is to mount the driver in an enclosure.
You could use a long tube the same diameter as the driver, filled with damping material, extending straight back behind the driver. This would remove the requirement for a baffle completely, unless you want a baffle to generate a baffle step to tailor the frequency response. I have attached a graph showing the Visaton driver mounted in a long 100mm diameter tube.
Or you could use a box the same dimensions as the baffle, deep enough to make the volume at least 10 times that of the driver's Vas. This would put the driver in "infinite baffle" conditions, apart from the baffle step.
What we really need is a frequency sensitive baffle material. It should be opaque (solid) at bass frequencies and transparent at mid and high frequencies. It will prevent cancellation of the bass, and allow the mid and HF "ambience" reflections from the walls behind the speaker to reach the listening position.
Another barrier to flat response is the inductance of the voice coil. You can see that the relatively high inductance of the Visaton voice coil causes the HF response to droop. I originally used the Tang band W3 1364SA driver in the simulation. It has a much lower claimed inductance and the response was almost flat at HF.
Finally, you're going to be excursion limited to about 4 watts with the Visaton driver and 17.5 x 30 mm baffle.
.
Do feel free to try it, though. I suggest you build the box as a cube, and the OB as a section that sits on top of the box, so that one face of the box forms the lower half of the baffle. That way you can try it with the box driver facing up behind the OB driver, then rotate the box 90 degrees so the box driver faces forward with the OB driver.
Some random thoughts:
The aim is to cancel the back wave to improve the frequency response in a way that won't affect the "boxless" sound of an OB speaker.
One problem is that part of the characteristic sound of an OB speaker is due to its dipole function. If you magically cancel the back wave, you create a monopole which will sound quite different due to different reflection patterns from the room surfaces, as well as elimination of the dipole comb effect.
But if you're certain that cancelling the back wave is the way to go, the most effective method is to mount the driver in an enclosure.
You could use a long tube the same diameter as the driver, filled with damping material, extending straight back behind the driver. This would remove the requirement for a baffle completely, unless you want a baffle to generate a baffle step to tailor the frequency response. I have attached a graph showing the Visaton driver mounted in a long 100mm diameter tube.
Or you could use a box the same dimensions as the baffle, deep enough to make the volume at least 10 times that of the driver's Vas. This would put the driver in "infinite baffle" conditions, apart from the baffle step.
What we really need is a frequency sensitive baffle material. It should be opaque (solid) at bass frequencies and transparent at mid and high frequencies. It will prevent cancellation of the bass, and allow the mid and HF "ambience" reflections from the walls behind the speaker to reach the listening position.
Another barrier to flat response is the inductance of the voice coil. You can see that the relatively high inductance of the Visaton voice coil causes the HF response to droop. I originally used the Tang band W3 1364SA driver in the simulation. It has a much lower claimed inductance and the response was almost flat at HF.
Finally, you're going to be excursion limited to about 4 watts with the Visaton driver and 17.5 x 30 mm baffle.
.
Attachments
BOTH. the FRS8 isnt much of a brilliant driver, its budget, and if i remember correctly, the FR is very rough before a somewhat early hi end roll off. For a real cheap FR project it is a handy driver though, although i like a slightly bigger diameter driver...like the visaton flatcone FR driver, um i forget the part number....I know its not the not the b200 anyways.
It is my suspicion that the cancelling driver merely causes more transverse lobing, since there is a physical displacment between that and the OB driver. This time delay would worsen phase issues, and deepen baffle nulls.
I dont think this can ever REALLY work unless used in some form of infinite baffle/wall speaker application, OR, maybe you just need MORE cancelling drivers to 'overwhelm' the displacement of the main driver. THis would likely kill all bass, BUT you can reduce the effect by padding ther cancellers, until a good response is achieved....and thats a whole other can of worms.....i mean where would the canceller go? and does the OH really want to listen to my music in the next room?
Last edited:
Chris,
I use sims to get "in the ballpark", to decide if an idea is worth constructing or not. As a general rule, such sims are fairly accurate at low frequencies. In this particular case, it says to me that getting meaningful "bass" from 3 inch drivers whose response drops like a rock below 150 Hz anyway is a waste of time.
But like I said, try it anyway. Try to keep environmental effects to a minimum in terms of positioning in the room etc. You might have been exciting a favourable room resonance when you heard the increase in bass.
Increasing the box volume didn't help much - a little more bass from the sealed unit. Once you get the box volume over 5 to 10 times the driver's Vas value, you're pretty much into infinite baffle range and futher increases in volume (or stuffing) make no significant difference.
I use sims to get "in the ballpark", to decide if an idea is worth constructing or not. As a general rule, such sims are fairly accurate at low frequencies. In this particular case, it says to me that getting meaningful "bass" from 3 inch drivers whose response drops like a rock below 150 Hz anyway is a waste of time.
But like I said, try it anyway. Try to keep environmental effects to a minimum in terms of positioning in the room etc. You might have been exciting a favourable room resonance when you heard the increase in bass.
Increasing the box volume didn't help much - a little more bass from the sealed unit. Once you get the box volume over 5 to 10 times the driver's Vas value, you're pretty much into infinite baffle range and futher increases in volume (or stuffing) make no significant difference.
Chris, looking forward to hearing from your re-test with this concept soon.
Mr.Mondogenerator, why would you pad down the cancellation drivers after adding a lot of them? I think by adding a hell lot of cancellation drivers that can overwhelm the OB driver, you will end up with a lot more omnidirectional bass than you need. (say adding 3 cancellation drivers, 1 cancels the back radiating wave of OB driver, there are still 2 cancellation drivers pumping bass that is in phase with the front OB, so you'll have a bit more bass.)
Please point out my misconception
Mr.Mondogenerator, why would you pad down the cancellation drivers after adding a lot of them? I think by adding a hell lot of cancellation drivers that can overwhelm the OB driver, you will end up with a lot more omnidirectional bass than you need. (say adding 3 cancellation drivers, 1 cancels the back radiating wave of OB driver, there are still 2 cancellation drivers pumping bass that is in phase with the front OB, so you'll have a bit more bass.)
Please point out my misconception
@ boris:
i was merely speculating to be honest. i speculate that more cancelling drivers would perhaps overcome the shortfalls of such an arrangement with equal 'main' & 'cancelling' drivers. how many would be needed would be down to experiment. I mentioned padding these drivers, in case the cancellation goes too far and rearward radiation becomes the dominant radiation direction.
i was merely speculating to be honest. i speculate that more cancelling drivers would perhaps overcome the shortfalls of such an arrangement with equal 'main' & 'cancelling' drivers. how many would be needed would be down to experiment. I mentioned padding these drivers, in case the cancellation goes too far and rearward radiation becomes the dominant radiation direction.
How's this for a thought (and it is only a thought...)
Instead of a single, flat baffle behind the OB, make a U-shape, where the sides of the U go on each side of the OB driver. On each vertical, put another driver. Surely, this'd help cancellation.
If you put all 3 drivers in series, then parallel the OB one, you'd get 6ohm impedance, and much higher percentage cancellation...
Instead of a single, flat baffle behind the OB, make a U-shape, where the sides of the U go on each side of the OB driver. On each vertical, put another driver. Surely, this'd help cancellation.
If you put all 3 drivers in series, then parallel the OB one, you'd get 6ohm impedance, and much higher percentage cancellation...
Chris, I was thinking about that too, but I'm not sure if that's really the way to go. Now I'm going from 2 drivers per channel to 4 drivers per channel!
It should have better cancellation, but does that really worth the effort?
BTW I think Aurasound NS3-193-8A it a good candidate for my OB/cancellation setup and probably also for your 4 driver setup.
It should have better cancellation, but does that really worth the effort?
BTW I think Aurasound NS3-193-8A it a good candidate for my OB/cancellation setup and probably also for your 4 driver setup.
is this really the case? im not sure....since radiating area has an impact on efficiency, doesnt it?
As for Chris' idea...im not sure thatll achieve any more than a bipole TL would, except the pipe resonances wouldnt be managed in the deisgn, unless it was designed for the specific purpose.
hypothetical question:
DO you think a pipe of perforated material would behave as a pipe, or transparently, or as a large series of helmholtz resonators....like studio sound absorbers do.......
this leads to another question.
forget the canceler driver. instead make a perforated rear panel to the OB, U shaped or whatever you would like. use the thickness of the panel like a port tube length dimension. calculate hole areas/diameters to create resonance at the desired frequency range/s. damp the perforated panel with BAF to reduce Q of resonances, and broaden their damping bandwidths. With a lot of planning, thought and experiment; a very effective canceller could be made.......perhaps greatly more efficient than a canceller driver.
Im actually so pleased with the idea(although its nothing new) i may try it myself........im thinking 12mm stock, holes diameters from 2mm to 8mm, constituting >70% surface area........although thicker stock may be better from a structural POV
Last edited:
@mr.Mondogenerator
That is already done some time ago by Linkwitz (Mini open-baffle speaker - PMT1).
I dunno how it works but it seems it'll still require terrible equalization as only a relatively small proportion of the back radiating wave can be cancelled and have very low efficiency.
If you go even further, you'll end up with a Transmission Line.
That is already done some time ago by Linkwitz (Mini open-baffle speaker - PMT1).
I dunno how it works but it seems it'll still require terrible equalization as only a relatively small proportion of the back radiating wave can be cancelled and have very low efficiency.
If you go even further, you'll end up with a Transmission Line.
@mr.Mondogenerator
That is already done some time ago by Linkwitz (Mini open-baffle speaker - PMT1).
I dunno how it works but it seems it'll still require terrible equalization as only a relatively small proportion of the back radiating wave can be cancelled and have very low efficiency.
If you go even further, you'll end up with a Transmission Line.
yup. You would end up with a 'true' TL.
However a quick glance at the PMT1 on lintwitz's site will show that it is just a simple dipole....no perforated sound absorbers used......
Last edited:
I'm still looking for a material that's acoustically transparent at HF, but opaque at LF...
Looks like I'm looking for some more drivers. I used to have 4 of the FRS8 drivers. 2 of them, I've (in a moment of insanity) painted a crude enabl pattern on. One of those was then cooked with a few hundred watts. The other 2 have had their dustcap removed, and a coat of PVA applied to the cone.
On the up side, I've been lent 4 CHR70 drivers.
Chris
Looks like I'm looking for some more drivers. I used to have 4 of the FRS8 drivers. 2 of them, I've (in a moment of insanity) painted a crude enabl pattern on. One of those was then cooked with a few hundred watts. The other 2 have had their dustcap removed, and a coat of PVA applied to the cone.
On the up side, I've been lent 4 CHR70 drivers.
Chris
The U-Frame approach will anyhow move the radiation pattern
towards a cardioid shape. There is no cancellation driver
needed. The effect comes from baffle plane being folded
towards the rear side.
If cardioid radiation pattern is the goal with OB like structures
it will be difficult to beat a U-Frame in compactness and efficiency.
Of course it is limited from its upper usable frequency because of
lambda/4 resonances according to the depth of the U-Frame.
I had sympathy with the "cancellation" approach, because
i thought it was possible to have less ragged response
obove the usual "dipole hump". However Don Hills simulation
shows only a marginal advantage of the cancellation approach
in that range. In the low bass there is a disadvantage compared
to "OB and closed box on same baffle".
For experimenting it could be interesting to search for configurations
and baffle dimensions which have smooth response from upper
bass to midrange, mabe there is still a chance for the cancellation
approach. Combining "Cancellation" approach with U-Frame will
introduce resonant behavior into the midrange and destroy
the last chance for that approach in having any advantage.
As can be seen "Tekton" designers are consequent in using BR
for low bass augmentation ...
They are also consequent in offset mounting the driver on the
baffle, which gives consideraby less ripple from bass to midrange
when the position is optimized.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.