It has always been taken for granted that motional feedback can only work with sealed boxes where the cone motion and the bass sound output are related in a simple way (and that might also include horns which often have sealed boxes behind the driver). That obviously isn't true of other enclosure types.
But if you have great powers of EQ by means of DSP, couldn't you adjust the frequency response of a motional feedback tuned-box or dipole sub accordingly*?
Ironically, wouldn't motional feedback be especially beneficial for enclosures that incompletely or unlinearlly constrain the motion of the cone (except by factors of driver construction)?
Ben
*"accordingly" might mean by doing room equalization
But if you have great powers of EQ by means of DSP, couldn't you adjust the frequency response of a motional feedback tuned-box or dipole sub accordingly*?
Ironically, wouldn't motional feedback be especially beneficial for enclosures that incompletely or unlinearlly constrain the motion of the cone (except by factors of driver construction)?
Ben
*"accordingly" might mean by doing room equalization
Motional feedback is not really necessary IMO because for the most part speakers act the way they are supposed to (within acceptable parameters) so I've never really looked into how it works.
BUT I am assuming you need a predictable frequency response curve and a predictable impedance curve so that you can have predictable excursion vs frequency to have information to base the amount of feedback necessary.
Most of the alignment types aside from sealed have wildly varying excusion at different frequencies even if they have a flat frequency response, this includes front loaded horns. Front loaded horns are not similar to sealed boxes just because there is a sealed chamber. I think you would be amazed how similar tapped horns and front loaded horns are in terms of impedance characteristics, excursion characteristics (above tuning) if you would actually LOOK at the information you are regularly presented with.
You need to know how speakers work with regard to frequency response, impedance and excursion before you can even begin to contemplate motional feedback for resonant.
Like I said, I don't know all that much about motional feedback and don't think it's necessary at all but I do know you need to have a firm grasp on how speakers actually work to use it. I'm sure it's not just as simple as using dsp to flatten the frequency response of resonant speakers and then it's good to go.
BUT I am assuming you need a predictable frequency response curve and a predictable impedance curve so that you can have predictable excursion vs frequency to have information to base the amount of feedback necessary.
Most of the alignment types aside from sealed have wildly varying excusion at different frequencies even if they have a flat frequency response, this includes front loaded horns. Front loaded horns are not similar to sealed boxes just because there is a sealed chamber. I think you would be amazed how similar tapped horns and front loaded horns are in terms of impedance characteristics, excursion characteristics (above tuning) if you would actually LOOK at the information you are regularly presented with.
You need to know how speakers work with regard to frequency response, impedance and excursion before you can even begin to contemplate motional feedback for resonant.
Like I said, I don't know all that much about motional feedback and don't think it's necessary at all but I do know you need to have a firm grasp on how speakers actually work to use it. I'm sure it's not just as simple as using dsp to flatten the frequency response of resonant speakers and then it's good to go.
I do think motional feedback is a good approach, not so much for distortion (for which the benefit is minimal because you can only add so much feedback without going unstable), but for maintaining frequency response and minimizing level compression. The feedback makes the acoustic output properties less affected by changes in voicecoil resistance (temperature) and variations in magnetic and suspension characteristics. With a sealed box, the relationship between cone acceleration and acoustic output is simple and direct, so the design is easier, and feedback ties the acoustic output more firmly to the electrical input.
With a ported box (or other more complex types), the relationship also depends on the port and box damping properties, so the acoustic output response relative to cone acceleration has a non-flat shape. It can still work, but you have to feed the input of the system with a response shape that has been pre-warped to give the needed flat acoustic output response.
That's assuming ideal conditions of course, in reality you won't get a flat acoustic response in either case since speaker and listener placement and the room get into things so strongly and have major effect on actual low frequency response. But the feedback will still take some of the variables out (including variations in sensitivity) and simplifies the response the speakers starts with, so that is a good thing.
With a ported box (or other more complex types), the relationship also depends on the port and box damping properties, so the acoustic output response relative to cone acceleration has a non-flat shape. It can still work, but you have to feed the input of the system with a response shape that has been pre-warped to give the needed flat acoustic output response.
That's assuming ideal conditions of course, in reality you won't get a flat acoustic response in either case since speaker and listener placement and the room get into things so strongly and have major effect on actual low frequency response. But the feedback will still take some of the variables out (including variations in sensitivity) and simplifies the response the speakers starts with, so that is a good thing.
Today's better drivers are already pretty linear with very symmetrical Bl and suspension curves as shown by Klippel testing. So for the most part, (as far as I know anyway) driver linearity is not really a big factor, if you want a very well behaved driver you just pay the premium for drivers that perform well.
As far as compression goes, is it not cheaper and easier to double up the speakers to relieve the thermal load for each coil? I don't see feedback as the best solution to keep drivers operating linearly when they are suffering thermal issues, if anything this would make the thermal issues worse by forcing the speaker to work even harder.
Undersized ports (anything over about 10 m/s) exhibit port compression, how would feedback deal with that? Ports are very rarely designed to have low velocity and most will have at least some compression, sometimes starting at moderate spl levels. The impedance curve changes when the coil heats up, not much but a little bit, motional feedback cannot address that. And resonances can vary in amplitude with differing power levels as shown at data-bass.com with progressive level power testing.
None of those things can be addressed with motional feedback but as previously suggested all of those things can be addressed by proper design with a high quality driver and using proper design and as many speakers as are required to keep response in the linear region.
And then there's this.
I do see slight advantages being possible from motional feedback but most of these issues are not really issues in the first place if you use high quality drivers, design for performance (large ports for example) and use as many speakers as are required to keep them in their comfortable linear range.
Last edited:
Not when the speakers require box volume and you want to have a number of them in the room to best work the room modes. Speakers are relatively cheap compared to adding cubic feet in a house.
That it can't, you have to have large enough ports (or shelf vents). Feedback can only control cone motion, what happens beyond that has to be stable on its own.
It absolutely can. That's what feedback does, affects the electrical drive into whatever impedance exists to force the cone motion towards the electrical system input. Ditto with the resonances (at low frequency, not cone or suspension breakups which can't be controlled by feedback), the cone motion is the controlled output. If the resonances are within the feedback bandwidth (and within the physical loop, not outside like the ports), effects of their variations will be reduced by feedback.
And, sure, you can use larger, high excursion, quality drivers in lots of boxes instead and approach the same performance. You can almost always spend money to improve things if you know what you are doing. You can also spend money on plane fare to travel and venue tickets to hear music live, for even better accuracy, for that matter!
I'm going to stick with the assertion that it's a very bad idea to control thermal issues by making the speaker work even harder with feedback. Throwing more power at a thermal problem to try to keep output linear is not the way to go.
I'm not sure what specific scenario you are talking about but doubling the amount of speakers in ANY given situation is usually going to only result in taking up a single digit fraction percentage of room volume. So I'm not sure anyone ever really needs to add an addition to their house to double up their subwoofers. Even four large subwoofers fit comfortably in a small room like the one I'm in now, 11 x 16 feet. Two as end tables, one in an ottoman and one in a coffee table and the 4 large subs take up 0 percent of the room's sq footage or volume that wouldn't already be used as furniture anyway. (That's theoretical, I don't actually have 4 large subs in my room but that's a cost issue, not a space issue.)
The point here is that if you want more subs it's usually pretty easy to fit them.
The point here being that cone motion is only one small part of the picture, I forgot to explicitly state that.
Ok, like I said I wasn't sure how it worked.
Assuming identical drivers and identical boxes, are you saying that doubling up a non motional feedback sub would only approach the performance of a single motional feedback sub? I would expect the dual subs to far exceed the performance of the single sub in every performance metric despite the single sub having motional feedback.
Good to have such evolved thinking posted. Thanks.
You are right to say MF never has a 1000000:1 feedback fraction like an IC has, due to terrible issues of stability in cone speakers (and I have an expensive ruined driver to demonstrate that). Commercial implementations must be esp. conservative.
But staying within practical feedback factors, you can materially reduce the awful distortion of Rice-Kellogg drivers (AKA cone drivers). Even more important, as most MF advocates will tell you, the transient response on a scope is dramatically better - and heard by ear too.
Combatting hot copper compression is an interesting benefit. But I have to say it was never as much an issue in my house as it seems to be in the homes and SUVs of many on this forum! I've been dreaming about double-voice-coil drivers: I wonder if MF with a DVC would fail to address hot copper compression due to heat transfer between coils? Maybe someone with a degree in physics could tell us.
Your term "pre-warped" - a clever description for room EQ! I think the way to think about MF is to think "locally" not globally. It corrects the driver but leaves the room to be fixed by other means. And actually, that's my question when I started this thread to ponder if MF really can be applied to tuned boxes.
Anybody who says drivers are already perfect and couldn't possibly be helped by feedback (in contrast to every other link in the audio chain), probably drives a 1925 Ford.... around the era that Rice-Kellogg developed the cone driver.
Ben
Not sure what awful distortion you are talking about - if you keep the drivers well within xmax as you should there is no awful distortion. I'm sure motional feedback provides some small benefit you can see on a scope but nowhere near as much benefit as using well designed drivers and doubling up drivers. And the very small issues that motional feedback can help with are orders of magnitude smaller than the much more important issues presented by the room and a dozen other factors.
You don't understand the issue if you think motional feedback can combat thermal issues. Making the speaker stay linear by adding more power is going to make the copper hotter, not combat the issue. This is not a long term strategy, it will lead to thermal failure.
Nobody said anything was perfect but motional feedback "fixes" issues that are not really issues if you choose drivers wisely and make a system that is powerful enough to stay well within it's comfortable and linear region at all times.
And if you think drivers haven't improved since 1925 you wouldn't notice the difference between a 1925 Ford or a modern car anyway so this is all lost on you.
Last edited:
I'd suggest perusing the patent papers on the Ipal and Mforce systems. Specifically the ones pertaining to the differential pressure control system. It does not detail exactly how it works with higher order resonant systems and I've not figured it all out myself but there are some nuggets of information in there.
There are some comparative measurements of a vented system with and without the feedback circuit and it effectively damps the ringing near the vent resonance. Not completely but notably.
Also just a quick aside about ports...A blanket statement of ports compressing at a certain or past a certain velocity cannot be made other than in a very general sense. A great deal of that depends on the area, shape, number and geometry of the vents.
There are some comparative measurements of a vented system with and without the feedback circuit and it effectively damps the ringing near the vent resonance. Not completely but notably.
Also just a quick aside about ports...A blanket statement of ports compressing at a certain or past a certain velocity cannot be made other than in a very general sense. A great deal of that depends on the area, shape, number and geometry of the vents.
Last edited:
Hi Josh,
Do you have any specific material you'd recommend reading about port compression? - I'm designing a few PA cabs at the moment and with the drivers I'm using its tough to get enough port into the cabinets.
Cheers
Chris
Do you have any specific material you'd recommend reading about port compression? - I'm designing a few PA cabs at the moment and with the drivers I'm using its tough to get enough port into the cabinets.
Cheers
Chris
JBL has an old AES paper that did an exhaustive study of various port shapes, surfaces and flares. I'm on my phone so I don't have it handy. A search may bring it up.
Also Collo's testing and results are worth a read. There are a couple of others most are older technical documents or studies but they still pertain.
In general, flaring = good and the larger the better. Minimizing the internal surface area of the port walls as much as possible for a given port area is also good for minimizing turbulence and allowing maximum air velocity prior to severe port overload. A single, circular port with a large flare at the ends is best in most cases. Of course this is not always practical for a number of reasons. A 3" port may severely compress at 25m/s but a 10" diameter pipe will still be relatively linear and noise free.
Also Collo's testing and results are worth a read. There are a couple of others most are older technical documents or studies but they still pertain.
In general, flaring = good and the larger the better. Minimizing the internal surface area of the port walls as much as possible for a given port area is also good for minimizing turbulence and allowing maximum air velocity prior to severe port overload. A single, circular port with a large flare at the ends is best in most cases. Of course this is not always practical for a number of reasons. A 3" port may severely compress at 25m/s but a 10" diameter pipe will still be relatively linear and noise free.
Many thanks for introducing the IPAL feedback gear. But I haven't been able to find any in-depth descriptions. Can you help please.
Ben
Just on a hunch, I'm guessing that JustAGuy has never worked with a motional feedback woofer. And to say that subwoofers take no significant floor space, I wonder if he's worked with those either?
The heating issue on voice coils isn't about voice coils burning up, it's about them changing the sensitivity from moment to moment while being used. The generated heat still changes the resistance for a bit even after the big pulse that heats the coil has gone and lower level is still trying to play.
The heating issue on voice coils isn't about voice coils burning up, it's about them changing the sensitivity from moment to moment while being used. The generated heat still changes the resistance for a bit even after the big pulse that heats the coil has gone and lower level is still trying to play.
Last edited:
If the second of the double-voice-coil which is used as a feedback sensor does not heat up*, then it would provide veridical feedback on the motions of the cone. Therefore, it would correct for the heating effect (as monumental source of anxiety day and night among the power-hungry posters to this forum), even if the heating effect was as anti-regular as you explain.
If the second coil heats up just as much, then, as I asked earlier, we need a wise person to model the consequences.
On the other hand, all the other MF sensors (bridge, accelerometer, ghostly apparitions, etc.) would also correctly correct, in proportion, for the heating effect.
In so far as I am made nauseous even thinking about bass reflex enclosures, my views may not matter much. But I think MF might help linearize cone motion in such boxes, but it could not sense (and hence, correct for) weirdo port compression phenomena.
Ben
*I believe that some dual voice coil drivers have the two coils wound on opposite sides of the former. If that former isolates the heat of the two then.... Seems to me that it would be simple to research heat transfer.
As I said there are very linear drivers and a doubling up on the subwoofers will easily outweigh any advantages a single sub with motional feedback, so no, I don't go looking for solutions to a problem that has never impacted me.
I had a 13.5 cubic foot subwoofer that took up 1.8 sq feet of floor space. And here's a 24 inch driver in a box that's 28 x 28 x 27. It's meant to be used an an endtable (see endtable option). It's not hard to fit subs even if small size is a goal. http://deepseasound.com/product/mariana-24sc-24-subwoofer/
So double up your subs and you have twice the capacity for thermal changes. This is going to work better than motional feedback to address this issue.
Last edited:
- Status
- Not open for further replies.