About MF: I still have a pair of the old Philips MF-3-ways somewhere, and before one channel gave up the gost, I must say they have good bass. The reason I feel MF has never really picked up is because the correction you want it to apply is very predictable; it is the usual 12 dB/octave downward slope you see below the -3dB point. The approach taken by Linkwitz to correct for this is much more sensible in my opinion.
You can of course correct for more than just frequency response deviations with "real" motional feedback: You can also iron out nonlinear distortion.
The Sony sub under discussion has a current feedback loop added. With this trick, depending on whether one uses positive or negative feedback, it is possible to lower or rise Qts. If you make the current feedback frequency dependant you can even alter other TSPs as well. This was commercially used in the so called "ACE bass" by the Swedish company Sonab.
What Sony did we can't tell from the little piece of schematic that was posted. It is also difficult to say whether the design is well thought-out. It could in fact be anything form bad but loud to good and maybe not so loud.
Regards
Charles
The Sony sub under discussion has a current feedback loop added. With this trick, depending on whether one uses positive or negative feedback, it is possible to lower or rise Qts. If you make the current feedback frequency dependant you can even alter other TSPs as well. This was commercially used in the so called "ACE bass" by the Swedish company Sonab.
What Sony did we can't tell from the little piece of schematic that was posted. It is also difficult to say whether the design is well thought-out. It could in fact be anything form bad but loud to good and maybe not so loud.
Regards
Charles
Great opportunity.
If you can "listen" to the system with a mic and analysis software, so much the better. It is amazing what you can do with a laptop and an old Radio Shack sound-level meter these days. I'm sure curious about how the output across the tuned range sounds in a small BR box with MF!!!
Short of opening the box and shorting the .22-ohm resistor, I don't offhand think you can do much of an A-B test of the effect of the MF. Wish you could.
Try pushing the cone. MF does this weird thing: the cone feels stiff when you give it a sharp push (because the MF fights back against your hand... depending on the frequency capabilities of the feedback loop, of course). A BR cone (without MF) is floppy.
Ben
If you can "listen" to the system with a mic and analysis software, so much the better. It is amazing what you can do with a laptop and an old Radio Shack sound-level meter these days. I'm sure curious about how the output across the tuned range sounds in a small BR box with MF!!!
Short of opening the box and shorting the .22-ohm resistor, I don't offhand think you can do much of an A-B test of the effect of the MF. Wish you could.
Try pushing the cone. MF does this weird thing: the cone feels stiff when you give it a sharp push (because the MF fights back against your hand... depending on the frequency capabilities of the feedback loop, of course). A BR cone (without MF) is floppy.
Ben
Hi Y'all,
Yes, I think ACE Bass or K.E. Stahl (Pat No. 4118600) would apply here, and bjorno has done a lot of work on this subject.
As to the schematic, the service manual is very detailed, and is available here:
Sony SA-W2500 | Owners Manual, Service Manual, Schematics, Free Download | HiFi Engine
(or, PM me).
Regards,
Yes, I think ACE Bass or K.E. Stahl (Pat No. 4118600) would apply here, and bjorno has done a lot of work on this subject.
As to the schematic, the service manual is very detailed, and is available here:
Sony SA-W2500 | Owners Manual, Service Manual, Schematics, Free Download | HiFi Engine
(or, PM me).
Regards,
Since then Velodyne still use accelerometers in their servo subwoofers. It's still the most accurate way of implementing a motional feedback system
Hey Ben,
Thought you might be interested in this:
http://www.powersoft-audio.com/en/c...c-ipalmod-system-introduction.html?Itemid=111
The differential controlled IPAL system can make even a ported box track the input signal quite closely.
Ain't going to be available at Wall Mart though
...Expensive toys!
Art
Attachments
Last edited:
Thanks for link. Interesting pressure tranducer... something like the technology for air mass measurement in automobile fuel-injection throttle bodies.
Anybody who looks at those fabulous tone-burst traces (pages 12-23 from the link) has to shout, "I gotta have a motional feedback woofer." There is no other way to control a woofer cone like that.
Ben
Anybody who looks at those fabulous tone-burst traces (pages 12-23 from the link) has to shout, "I gotta have a motional feedback woofer." There is no other way to control a woofer cone like that.
Ben
Hey Ben,
Thought you might be interested in this:
http://www.powersoft-audio.com/en/c...c-ipalmod-system-introduction.html?Itemid=111
The differential controlled IPAL system can make even a ported box track the input signal quite closely.
Ain't going to be available at Wall Mart though
Expensive toys!
Art
Great stuff! I want one.
OK... let me comment on the article link from Weltersys.
There are two components to the article (not written in English clear enough for my full grasp). The first has to do DSP modeling of something or other - looks interesting but over my head. This relates, I think, to shaping the input signal to the driver.
The second has to do with MF using a transducer that compares pressure (1) inside the cabinet with that (2) near the front of the driver, as I understand it. They claim their system can be used with any driver; they use an 18 inch driver in a BR cabinet for testing in the article.
Mic results with a four-cycle tone burst at 5 frequencies associated with the BR tuning are shown. The astonishing results are similar to my own tests years ago. As I said before, one look at the improvement due to MF and a serious fan of woofing will want to rush to DIY. I'd like to see anybody seriously deny those results look immensely impressive.
I was wrong to say MF in a BR environment is an oxymoron - it really is only half-peculiar. As the traces show, there are monumental improvements to a tone burst response as far as listening-by-eyeball can reveal. No doubt this improvement would be a great leap forward in any acoustic system including OB (freq corrected in the usual ways) and BR (which is ordinarily beyond easy correction).
The problem for me in attempting BR using voice-coil based MF is that the sound output as you pass through the broad resonant frequency region is that there is an orderly but convoluted relation between speaker sound output and cone motion. That is because at some frequencies, the port adds to sound (upper bump) or subtracts from sound (lower bump) or just holds the cone steady so it doesn't go booooom too much (system resonance). Go figure.
What is especially peculiar to me, pending learning more from this forum, is how you can take the dramatic changes in pressure inside a BR as you pass through the hills-and-valleys of resonance of a BR (as described above) and make the speaker output uniform? I think it is fair to say they authors are "shy" about the freq curve.
While we all know about the output of a BR box (as described above), I have little knowledge of what the pressures inside the box are doing.... let alone how the comparison of front versus interior pressures can lead to a flat response.
On the other hand, it is easy enough to understand how any kind of MF gizmo can result in the fabulous tone-burst results shown in the article, even in a BR cabinet.
Footnote: the MF literature has a hidden agenda that makes the article a little weird. That "agenda" has to do with claims to support patenting. Each author wants to show his/her system is novel, different, solves new problems, etc. and otherwise distinct from earlier efforts. Including this article.
Ben
PS - I am curious about that Walmart $99 speaker... but I am not thinking about replacing my Klipschorn and giant OB with the little Sony box, even if it has MF.
There are two components to the article (not written in English clear enough for my full grasp). The first has to do DSP modeling of something or other - looks interesting but over my head. This relates, I think, to shaping the input signal to the driver.
The second has to do with MF using a transducer that compares pressure (1) inside the cabinet with that (2) near the front of the driver, as I understand it. They claim their system can be used with any driver; they use an 18 inch driver in a BR cabinet for testing in the article.
Mic results with a four-cycle tone burst at 5 frequencies associated with the BR tuning are shown. The astonishing results are similar to my own tests years ago. As I said before, one look at the improvement due to MF and a serious fan of woofing will want to rush to DIY. I'd like to see anybody seriously deny those results look immensely impressive.
I was wrong to say MF in a BR environment is an oxymoron - it really is only half-peculiar. As the traces show, there are monumental improvements to a tone burst response as far as listening-by-eyeball can reveal. No doubt this improvement would be a great leap forward in any acoustic system including OB (freq corrected in the usual ways) and BR (which is ordinarily beyond easy correction).
The problem for me in attempting BR using voice-coil based MF is that the sound output as you pass through the broad resonant frequency region is that there is an orderly but convoluted relation between speaker sound output and cone motion. That is because at some frequencies, the port adds to sound (upper bump) or subtracts from sound (lower bump) or just holds the cone steady so it doesn't go booooom too much (system resonance). Go figure.
What is especially peculiar to me, pending learning more from this forum, is how you can take the dramatic changes in pressure inside a BR as you pass through the hills-and-valleys of resonance of a BR (as described above) and make the speaker output uniform? I think it is fair to say they authors are "shy" about the freq curve.
While we all know about the output of a BR box (as described above), I have little knowledge of what the pressures inside the box are doing.... let alone how the comparison of front versus interior pressures can lead to a flat response.
On the other hand, it is easy enough to understand how any kind of MF gizmo can result in the fabulous tone-burst results shown in the article, even in a BR cabinet.
Footnote: the MF literature has a hidden agenda that makes the article a little weird. That "agenda" has to do with claims to support patenting. Each author wants to show his/her system is novel, different, solves new problems, etc. and otherwise distinct from earlier efforts. Including this article.
Ben
PS - I am curious about that Walmart $99 speaker... but I am not thinking about replacing my Klipschorn and giant OB with the little Sony box, even if it has MF.
Last edited:
Ben,
Some excerpts from the IPAL literature:
"At first it could to be noted, in fact, that one of the most important advantages of using a switching amplifier stage, usually referred as “Class D” amplifier, is the ability of this kind of power amplifier to manage any kind of Voltage to Current relation at its output, up to the full quadrature between them, practically managing a fully reactive load without any problem. Moreover, where the load is very reactive, the power that is bounced back from the loudspeaker to the amplifier output, in Class D operation, it could be easily recovered and recycled into the power supply rails. This kind of situation would be not possible to be managed by a traditional analog output stage instead, because the power dissipation would be extremely high in that case.
“ZERO LATENCY” DSP
An innovative architecture that ensures an astonishing 10us (microseconds) latency on the critical feedback paths allows “analog type” feedback approach with the flexibility of a DSP core."
I suspect an important part of what is only given a sideways mention in the white paper is that the entire signal can be delayed, then re-arranged as needed to comply with the original waveform. The "zero latency" is only a virtual latency, a pre-delay allows the system to "look forward" in relative time and perform the required adjustments.
Neither Class D amplifiers or the DSP horsepower to manage at the level of the IPAL were available until relatively recently.
Although all this makes for sine wave traces that look quite impressive, it should be noted that the transducers used in the IPAL are quite impressive on their own.
All the processing improves on a very good platform that in a home environment would be operating in it's linear range to start with.
Using high quality high excursion transducers in sealed enclosures with proper amplification and EQ one can see impressive duplication of sine wave traces at amplitudes up to the linear operating range, plenty of SPL to cause structural damage.
Using bass reflex or horn or tuned pipe enclosures, excursion and power demands can be lessened, providing more SPL in the linear operating range of a transducer.
When subs are time and phase aligned at the crossover point, the upper impulse response of any properly designed system will provide "punch", or "Pace, Rhythm And Timing" for the prats.
The fact that alternative systems may not track the original waveform exactly below the crossover point is largely academic in any room small relative to the wavelength produced, that is most domestic listening rooms, as the room response will mess up any "perfect" reproduction.
Current pricing of class D amps and DSP could easily allow low cost, non (or minimally) adjustable low cost versions of the IPAL technology.
Unfortunately, good high excursion transducers will never be as cheap as electronics, the tendency will be to "fix" cheap transducers with electronic "solutions".
While impressed with the IPAL technology, I would prefer to invest my money in good transducers and reasonably priced DSP rather than it, more bang for the buck.
Art
Yes, with DSP you can shape the input to a great degree. But that is not the same accomplishment as avoiding boxes with major resonances, with or without re-shaping.
Those lousy looking tone-bursts for non-MF drive seemed pretty realistic. I will patiently await seeing you post pictures of equally excellent tone-bursts without MF.
Odd to hear you say good tone-bursts are immaterial since a normal small listening room will add a lot of crap anywa. No my way to look at speaker performance.
Ben
Those lousy looking tone-bursts for non-MF drive seemed pretty realistic. I will patiently await seeing you post pictures of equally excellent tone-bursts without MF.
Odd to hear you say good tone-bursts are immaterial since a normal small listening room will add a lot of crap anywa. No my way to look at speaker performance.
Ben
You will be waiting a long time for me to post any tone burst pictures, I have survived for 56 years without an o-scope and it is not on my shopping list any more than a motional feedback system seems to be on yours
.Interestingly, the Ipal seems to have the least positive effect in the 120 Hz range, the place where the speaker response is not subject to the resonant behavior that you seem to abhor in principle.
That frequency range is typically the most critical to integration between the sub and top cabinet, but as the excursion is least there, motional feedback has the least potential to fix problems.
Oh well, the lower, more non-critical frequencies look great
.Art
Attachments
Hello Oliver, Rythmik are in your locale!
There's a good deal of info, mostly reviews, on the web. Mostly positive, sufficient for me to buy the 15" driver + amp kit & have it shipped to Australia. I'm in the process of building a 4ft^3 sealed front firing enclosure (plan from RA's site) using salvaged timber (after paying $260 freight, I can't justfy any additional expenses!)
Roj. Perth, Western Australia
There's a good deal of info, mostly reviews, on the web. Mostly positive, sufficient for me to buy the 15" driver + amp kit & have it shipped to Australia. I'm in the process of building a 4ft^3 sealed front firing enclosure (plan from RA's site) using salvaged timber (after paying $260 freight, I can't justfy any additional expenses!)
Roj. Perth, Western Australia
Pickup coils wound around the main drive voice coil are ineffective for motional feedback which is why this thing is ineffective at reducing distortion and why well designed servo speakers use accelerometers.
You may also note that there is virtually no information on the website regarding distortion performance which is the information you would expect to see from a well designed servo system.
regards
Trev
Last edited:
You will be waiting a long time for me to post any tone burst pictures, I have survived for 56 years without an o-scope and it is not on my shopping list any more than a motional feedback system seems to be on yours
Interestingly, the Ipal seems to have the least positive effect in the 120 Hz range, the place where the speaker response is not subject to the resonant behavior that you seem to abhor in principle.
That frequency range is typically the most critical to integration between the sub and top cabinet, but as the excursion is least there, motional feedback has the least potential to fix problems.
Oh well, the lower, more non-critical frequencies look great
Art
If the lower frequencies weren't that critical then you should be able to get away with using use a $3 woofer with a one ounce magnet and a 3 mm X-max to do the job of a properly designed woofer. Somehow this doesn't work in practice because maybe those frequencies are a lot more critical than what you think they are
regards
Trev
MFB principles and commercial available products
@bentoronto,
Have done some hobby projects with MFB long long time ago. I restored a pair of old Philips speakers and renewed the complete electronics, with proper current sources in the main amps (in stead of these dirty bootstrapping capacitors in the power amplifiers) and I have redone the filter circuitry.
Must say that I was stunned by the bass and the 'near apparent' sound of those beasts. Very good definition. New capacitors and new transistors certainly helped (remember, these are from the '70's ! )
I indeed experienced what you wrote in this forum: The speaker is 'working against' the movement you try to make with it, like a real servo. I still believe in this principle today: Philips did a great job back then. But the real eye-opener for me was to replace the tweeters with Focal tweeters. Long time ago, I do not remember the exact types, but I played these speakers for 15 years after my restoration. I still have them somewhere, cannot let them go...
Oh, to stay on topic: A technical aspect not mentioned here is the dynamic range of the MFB (or any other active FB principle). Always be critical of the sensor system you apply, otherwise it can do more harm than any good. The dynamic range is enormous... Cheap implementations often is current measurement. You might wonder what good can that do? The speaker in resonance does not like current, it even does not need much current. Makes me think...
@bentoronto,
Have done some hobby projects with MFB long long time ago. I restored a pair of old Philips speakers and renewed the complete electronics, with proper current sources in the main amps (in stead of these dirty bootstrapping capacitors in the power amplifiers) and I have redone the filter circuitry.
Must say that I was stunned by the bass and the 'near apparent' sound of those beasts. Very good definition. New capacitors and new transistors certainly helped (remember, these are from the '70's ! )
I indeed experienced what you wrote in this forum: The speaker is 'working against' the movement you try to make with it, like a real servo. I still believe in this principle today: Philips did a great job back then. But the real eye-opener for me was to replace the tweeters with Focal tweeters. Long time ago, I do not remember the exact types, but I played these speakers for 15 years after my restoration. I still have them somewhere, cannot let them go...
Oh, to stay on topic: A technical aspect not mentioned here is the dynamic range of the MFB (or any other active FB principle). Always be critical of the sensor system you apply, otherwise it can do more harm than any good. The dynamic range is enormous...
Cheap implementations often is current measurement. You might wonder what good can that do? The speaker in resonance does not like current, it even does not need much current. Makes me think...The harm you refer to is the terrible threat faced by those who try to produce MF systems commercially: feedback system go nuts when naive customers push them outside their working range... even just for an instant.
But when you think of power range or Xmax-like parameters, what first comes to mind is the monumental excursions you can see on BR boxes and infinite baffles when driven by amps riskily lacking low frequency roll-off. For the normal music frequency range - even for low freq passages - the excursion signal for 12- or 15-inch drivers is nothing like that.
Ben