Ben,
With that type of accelerometer has been experimented too, as far as I know. But those experiments were not satisfactory.. also the price the distributor calculated sky-rocketed..
Sorry that the page intended is hard to reach...
perhaps this one works better : Motional Feedback Audio.
Thanks for links. But the ACH01 has been the "gold standard" for a while with very impressive linearity in published specs. That is, pending seeing those in the link replace it in the happy future.
Cost not biggest problem for a DIYer like me making home gear.
Any suggestions or warnings for mounting the ACH01?
Ben
Last edited:
Who started the OT about boom by saying "Isn't "group delay" a euphemism used for BOOM by people who build bass reflex enclosures? Others say, "boom".
The noise cancellation tech exists, it's in wide use and it works well. I'm not sure if people are using it for sound absorption in rooms (but it could be) and it has been used for quite awhile in applications like headphones. The device "listens" to the noise and counters it with an equal noise out of phase. I suppose that's vaguely similar to mic controlled MF but it's quite a stretch to say "there is no difference between the MF circuit and one you'd make if you were constructing an active room absorber".
Back in post 109 you were pretty sure MF doesn't eat external sound. Remember this?
"Yes, despite the obvious theoretical obstacles, nothing preventing you from using the IPAL item with vented enclosures as per his constraints. Providing that:
1. you are installing it where drunk dancers are the audience since it may do strange things to the freq response
At this point you've complimented MF for it's distortion reduction. Later your main goal with MF was extended frequency response and better impulse response. Now you want to explore MF's ability to eat external sound.
Instead of pondering the theoretical benefits of MF ad infinitum, why not buy or build and measure a MF speaker? It would be faster and easier to go that route than to continue this discussion.
Problem with the sensor was it's weight, I understood... and it didn't perform as well as the "reference" RH544 speakers. Weight appeared too much for rather small units (even the 8"). The picture on the site shows how the sensor was mounted on a small PCB. The PCB was glued upon the voicecoil-former.
Mounting of the alternative sensor PCB's is likewise, but they are much less in weight. If I were you I'd contact the guy....
Edwin
Last edited:
OK, since this thread has grown legs, both in size and subject matter, I'd like to say a few things. I'll probably get 'schooled' but that's perfectly fine, learning something new everyday is a good thing in my book.
Note: If I am not understanding what has been said correctly, just disregard this post.
I'll go back a bit to where bolserst and others think you can achieve what servo/MF can do with an EQ, I respectfully disagree. Even on hearsay and no proof (yet).
Why do servo companies use better transients (accuracy, starting and stopping) of, for instance, subwoofers as a main selling point if it couldn't deliver the goods?
Couple examples: Genesis Advanced Technologies › Technology › The Servo-Controlled Advantage
Rythmik Audio • FAQ - Frequently asked questions
I'll quote Rythmik talking about servo vs push-pull:
This is the main reason I'm even considering servo/MF subs, is this faster (ie; more accurate) starting and stopping.
It seems to me that just EQ, even in a small sealed box will not accomplish this, even an amp with great slew-rate and damping-factor thrown in for good measure.
Also seems that a modern high-powered driver will have it's way with the air inside a sealed box (overpower it's properties), and not provide what servo/MF could do...
Am I out to lunch on this?
AND, not to further complicate things, but if the servo/MF circuitry can over-ride the TS of a driver and make it behave the way it wants, why would you not try and use it in a ported box tuned very low, with a proper filter in the servo amp assuring that the system will remain above tuning frequency (12Hz in Rythmik's case [iirc]) therefore providing stability?
Just thinking out loud here, anyone want to comment?
Cheers
-Steve
Note: If I am not understanding what has been said correctly, just disregard this post.
I'll go back a bit to where bolserst and others think you can achieve what servo/MF can do with an EQ, I respectfully disagree. Even on hearsay and no proof (yet).
Why do servo companies use better transients (accuracy, starting and stopping) of, for instance, subwoofers as a main selling point if it couldn't deliver the goods?
Couple examples: Genesis Advanced Technologies › Technology › The Servo-Controlled Advantage
Rythmik Audio • FAQ - Frequently asked questions
I'll quote Rythmik talking about servo vs push-pull:
This is the main reason I'm even considering servo/MF subs, is this faster (ie; more accurate) starting and stopping.
It seems to me that just EQ, even in a small sealed box will not accomplish this, even an amp with great slew-rate and damping-factor thrown in for good measure.
Also seems that a modern high-powered driver will have it's way with the air inside a sealed box (overpower it's properties), and not provide what servo/MF could do...
Am I out to lunch on this?
AND, not to further complicate things, but if the servo/MF circuitry can over-ride the TS of a driver and make it behave the way it wants, why would you not try and use it in a ported box tuned very low, with a proper filter in the servo amp assuring that the system will remain above tuning frequency (12Hz in Rythmik's case [iirc]) therefore providing stability?
Just thinking out loud here, anyone want to comment?
Cheers
-Steve
Last edited:
No, but it's not a huge benefit until you push the drivers past their linear safe zone. Try MF and do some measurements with and without at the listening position in a real domestic living room with real music (not close mic in a carefully controlled anechoic chamber concentrating on zoomed in impulse signals) and see if you can see much (if any) difference.
Whether you are talking about driver fs or the tuning frequency of the box, usually if you limit your passband to the region outside resonance you are giving up a lot of efficiency. So there are consequences. What are the benefits? None.
Besides, you can design a very nice ported box with currently available drivers without using MF so what's the point?
You seem to have thought it through right.
What is really amazing is the term "negative output impedance amp". Just picture what that means when you have the amp and speaker sitting next to your test bench and you are checking it out on your oscilloscope.*
When you run a sine sweep across the resonance of the speaker, the voltage dips in the area where the speaker "lets go". That's how MF counter-acts "group delay". Likewise for any other erroneous motion, inertia, mass flop, or distortion, it counter-acts it.
And when you push on the cone, it holds back**. Like Frankenstein, "It's ALIVE". And so it is.
Ben
*on the test bench, you'd play a constant since wave into a voice coil MF amp of say 2 volts into 8 Ohms (substituted for the driver). The you'd substitute 7 Ohms, etc. When you calculate the amp's output impedance it turns out, amazingly enough, to be not just low but negative and not low. The same is true of accelerometer feedback, I guess, but the test set up is quite different. I bet this little bit of elementary shop practice is incomprehensible to some folks who think they are masters of plugging numbers into sims.
**depending on how good and how wide-band your MF is.
Last edited:
Fair enough, I will try this in my normal bass-trapped/treated listening environment.
But I was under the impression that servo/MF, applied this starting/stopping at all levels, not just when the driver is pushed past it's linearity. (but maybe you implied that won't make much difference listening-wise) We shall see, I'm ordering a Rythmik DS1500ci due to the fact that it's competitively priced to similar, non-servo, (all-on-sub) kits. I will measure using a simple bass-drum recording..and report findings...
BTW) If the Rythmik proves to be deficient for my needs, I plan on using the driver and amp as donors for an accelerometer-based project. Hopefully I can make that work, if not, I have a buyer..
What I will be building will take full advantage of resonances, and for exactly the reason you are talking about, efficiency. I'm just going to avoid going to where the system becomes unstable (ie; below tuning of the box).
Thanks for commenting.
Cheers
-Steve
All feedback systems have potential for going unstable, according to the well-known principles of phase and gain. Unlike the way BR boxes disappear below resonance, MF doesn't treat resonance any differently than any other point in the compass within its bandwidth. In the case of a sealed box, you know that the output is related in a predictable manner to the cone motion with the resonant point meaning nothing at all for a MF system. Well, almost nothing.
With BR box you don't know. In fact with most forms of MF, the output will disappear FASTER than 18dB/8ave below resonance. Think about it.
But you may find the challenge to stability is in the bandwidth of the amp and the sort of feedback signal you are sending back into the amp. Ordinarily you want something that goes pretty low - like DC if you're a bettin' man.
I used a DC-coupled amp with a single cap inserted along the way for a super-low AC-couple, for a long time with no crises.
None of this challenge should bother a good person who knows how to use their oscilloscope except for testing perturbations like wind and little kids (and visiting MF authorities) knocking on your cone.
Ben
Last edited:
These are very good points, duly noted.
I lol'ed at the frankenstein bit, but it's true, isn't it?
Finally ,yes, the reaction of the MF amp would confuse the sim if not accounted for... (will they ever be?), I think I finally know what you are getting at with your comments about them.(sims)
You do crack me up Ben, s'pose that's a good thing..
Oh and BTW, no further comments from me tonight ...I'm off drinking somewhere...lol...
Last edited:
Well, that isn’t very specific nor particularly helpful in improving understanding of your views. Perhaps I need to create a 10 question multiply choice survey for you?
Reading between the lines, I guess I would have to agree with zettairyouiki that you are saying you feel only sealed enclosures are suitable for hi-fi use. Is that correct? Nothing wrong with that, many designers and listeners share your view. Also nothing wrong with others finding enjoyment and technical challenge in pursuing other enclosure types. Concerning envelopes and sim…I don’t think there is anything wrong with people using engineering tools developed by people smarter than they are, although you usually have to have a certain level of understanding to take full advantage of them. But, I do share the feeling that many (including some engineers I work with) rely way to heavily on computers and fail to develop intuitive understanding of the physical principals involved in their field of work.MFB doesn’t eat group delay, but it does modify response in a manner that reduces it. Group delay is the rate of change of phase…basically a measure of how quickly the phase is wrapping up or changing as you go down in frequency. The sharper the LF roll-off corner and the faster the rate of roll-off below the corner the quicker phase changes, so higher group delay. MFB flattens and extends the response and reduces the rate of roll off at the corner, all of which reduce phase rotation and group delay. With VC feedback, a roll off rate of only -3dB/oct is not uncommon for an octave or so below the corner.
MFB does react to impinging pressure waves. But as you know there isn’t a very good impedance match between the air and a heavy woofer. As a result they make pretty poor microphones and don’t react strongly to sound waves from other sources. Active bass absorbers like the Bad End E-trap or AVAA from PSI Audio use carefully positioned microphones better suited for the purpose.
Active Bass Trap Archives - Bag End
AVAA C20 (Active Bass Trap) - PSI AudioPSI Audio
Is the 15” driver the Brutus BRZ15D4 you mention in your other thread? If so, I fear you may have to remove the outer layer of the cone to be able to attach the accelerometer in close enough proximity to the voice coil to avoid resonance in the structural path between VC and Accel < 1khz. I could be wrong, but I don’t think the center of the outer dished cone is bonded to the VC. As a test, get some 1/32” VHB tape and stick the Accel on to the center of the cone and see what you get. You will know pretty quickly if what you have is useable. Attached is an example magnitude and phase signals measured from a 15” DVC, ACH-01 mounted to the large, whimpy dust cap. Note resonance of dust cap with ACH-01 mounted to it(red curve) is around 300 – 500Hz…unacceptably low to be able to achieve useful amounts of feedback. Note also the velocity signal from the second VC starts to become contaminated by the voltage from mutual coupling with the driven voice coil around 60 – 70Hz. This is too low for building a MFB system to operate up to a 100Hz crossover frequency without significant compensation to null out the effects of the mutual coupling. The true velocity of the cone continues to fall at ~6dB/oct with rising frequency above the peak at resonance.
Attachments
Quote:
Originally Posted by weltersys
I have mixed in far too many "boom boom rooms" designed to make audience noise seem even louder than it already is.
As you originally asked the question about "boom", it is humorous for you to then single out and call one of four of my answers in post #156 "off topic".
In reply to your four points above:
1) MF and active room absorbers require completely different programming, though both can help provide a response that is closer to the recorded (or live) source.
2) No, since the MF sensor is typically placed in close proximity to the cone, it rejects the room sound at the inverse distance law, and does not "eat" group delay. To (virtually) eliminate group delay requires FIR filters.
3) Yes, I do.
4) Of interest (but somewhat off topic) I did extensive experimentation with the IBVA (Interactive Brainwave Visual Analyzer) during the decade+ from 1994 to 2005 (until Apple/Macintosh software/hardware "updates" rendered the computer based IBVA unusable…) which used three "sticky pad" electrode sensors placed on the subject's forehead to detect left and right brainwaves. The IBVA was my first actual user introduction to FFT (Fast Fourier Transform) analysis, the basis of audio comparator analysis programs like REW, SMAART etc.
As we (Dan Winter & company) discovered, a Callahan probe, a length of organic cord made of hemp or cotton soaked in salt (sea) water, wrapped around an oscilloscope probe with the ends attached to the + & - terminals, shaped like a turnip (or a stubby carrot), and wrapped tightly with a thin sheet of plastic wrap to retain the moisture, proved to be an effective transducer for the waves emanated from plant life, in specific, trees.
Of further interest, the output voltage of the Callahan probe placed on a two+ foot diameter tree trunk was much higher than any human brain wave using the standard IBVA electrode sensors, resulting in input clipping. Not understanding at the time (1994) that the results of clipping a FFT (or any analog to digital) interface results in a harmonic cascade, Dan concluded that the "tree was in a fractal harmonic interface with the (human) test subject", when the fact was that the harmonic cascade the tree's waves clipped input would "line up" with virtually any wave. To my knowledge, Dan never acknowledged the technical error, within the last five years or so I recall him still writing about the "fractal harmonic interface with the (human) test subject" as being fact.
One actual fact- the test subject could exhibit far greater control of his brain waves than any I have ever seen, most persons brain waves are fairly random (unless in subconscious, highly attentive or very meditative states) but his looked like a mountain range centering at around 15 Hz.
Long story short, different sensors are good for different things- a microphone (AKA differential pressure device) is probably better for an audio comparator than an accelerometer, voice coil, or locally grown organic carrot.
Have fun with your MFing "work"!
Cheers,
Art
Originally Posted by weltersys
I have mixed in far too many "boom boom rooms" designed to make audience noise seem even louder than it already is.
Ben,
As you originally asked the question about "boom", it is humorous for you to then single out and call one of four of my answers in post #156 "off topic".
In reply to your four points above:
1) MF and active room absorbers require completely different programming, though both can help provide a response that is closer to the recorded (or live) source.
2) No, since the MF sensor is typically placed in close proximity to the cone, it rejects the room sound at the inverse distance law, and does not "eat" group delay. To (virtually) eliminate group delay requires FIR filters.
3) Yes, I do.
4) Of interest (but somewhat off topic) I did extensive experimentation with the IBVA (Interactive Brainwave Visual Analyzer) during the decade+ from 1994 to 2005 (until Apple/Macintosh software/hardware "updates" rendered the computer based IBVA unusable…) which used three "sticky pad" electrode sensors placed on the subject's forehead to detect left and right brainwaves. The IBVA was my first actual user introduction to FFT (Fast Fourier Transform) analysis, the basis of audio comparator analysis programs like REW, SMAART etc.
As we (Dan Winter & company) discovered, a Callahan probe, a length of organic cord made of hemp or cotton soaked in salt (sea) water, wrapped around an oscilloscope probe with the ends attached to the + & - terminals, shaped like a turnip (or a stubby carrot), and wrapped tightly with a thin sheet of plastic wrap to retain the moisture, proved to be an effective transducer for the waves emanated from plant life, in specific, trees.
Of further interest, the output voltage of the Callahan probe placed on a two+ foot diameter tree trunk was much higher than any human brain wave using the standard IBVA electrode sensors, resulting in input clipping. Not understanding at the time (1994) that the results of clipping a FFT (or any analog to digital) interface results in a harmonic cascade, Dan concluded that the "tree was in a fractal harmonic interface with the (human) test subject", when the fact was that the harmonic cascade the tree's waves clipped input would "line up" with virtually any wave. To my knowledge, Dan never acknowledged the technical error, within the last five years or so I recall him still writing about the "fractal harmonic interface with the (human) test subject" as being fact.
One actual fact- the test subject could exhibit far greater control of his brain waves than any I have ever seen, most persons brain waves are fairly random (unless in subconscious, highly attentive or very meditative states) but his looked like a mountain range centering at around 15 Hz.
Long story short, different sensors are good for different things- a microphone (AKA differential pressure device) is probably better for an audio comparator than an accelerometer, voice coil, or locally grown organic carrot.
Have fun with your MFing "work"!
Cheers,
Art
Last edited:
To sell product? I don’t think there are any particularly false claims on the websites you linked to, you just have to read between the lines and understand what they are really saying. BTW, notice that Genesis uses dual woofer force cancelling arrangement which is particularly advantageous for accelerometer based MFB which does respond to motion/vibration of the enclosure, not just the woofer (as mentioned on the Rythmik website).
I started experimenting with MFB for the exact reasons you mention…the verbal picture presented in glossy brochures of MFB taking a firm hand with the woofer motion. My early tests showed just what I had expected, reduced distortion at extremes of stroke, and improvement to the look and sound of tone bursts at all signal levels. But, this was comparing the same woofer in the same box, just with MFB turned on or off. A mentor of mine suggested I compare results with response matched by EQ with what I was getting with MFB turned on. It was definitely a light-bulb moment for me.
If you have access to a mini-DSP, DCX2496, or similar it might be educational for you to observe some tone bursts passed thru them. Then experiment with different HF and LF roll-off filters to emulate the response of typical sealed of ported box alignments. The effect on the start and stop of tone bursts even for frequencies in the middle of the subwoofer passband might surprise you. If at some later date you see similar time history traces from a woofer measurement, you might consider that it is not necessarily a woofer in need of a firm hand causing the tone burst to look that way but rather a simple consequence of the overall response(magnitude and phase).
No reason not to
Last edited:
Bolserst,To sell product? I don’t think there are any particularly false claims on the websites you linked to, you just have to read between the lines and understand what they are really saying.
If you have access to a mini-DSP, DCX2496, or similar it might be educational for you to observe some tone bursts passed thru them. Then experiment with different HF and LF roll-off filters to emulate the response of typical sealed of ported box alignments. The effect on the start and stop of tone bursts even for frequencies in the middle of the subwoofer passband might surprise you. If at some later date you see similar time history traces from a woofer measurement, you might consider that it is not necessarily a woofer in need of a firm hand causing the tone burst to look that way but rather a simple consequence of the overall response(magnitude and phase).
No reason not to
All your points are exactly correct, though seem to make no difference to the OPster.
Art
Well, all the benefits except the ability to get higher output with low distortion from a given enclosure size.
I’d still encourage first hand experimentation, it’s fascinating really. Also, you never know when something you learn might come in handy.
Brilliant +1.forr rightly mentioned
Also one of the better written patents in terms of education rather than obfuscation.
Last edited:
Correct on all points. But I can honestly say I'm not interested in experimenting with MF, I think the time and money is much better spent elsewhere. I think I mentioned to you awhile ago that I was considering designing and building a 125 cubic foot front loaded horn subwoofer with four high excursion 18 inch drivers. That is contingent on two points -
1. Ability to obtain the driver's at dealer cost.
2. Ability to assuage the guilt associated with spending the $3000 it would require rather than saving it for more "adult" investments.
THAT is the path to low distortion as far as I'm concerned, not MF. Of course that's not an option for people with cranky spouses or people living in small apartments but I don't have problems like that.
That would constitute a real solution to both the "good design" and "properly sized system" aspects. The third remaining factor, as I previously mentioned, is getting the room well treated. Until all that is done I'm not even moderately interested in considering MF, as it's the wrong tool to fix these three real problems. AFTER all that is done I might consider it but with the real problems addressed adding MF wouldn't improve anything much.
It's true there is a trade-off between passband and loudness (the efficiency is not affected!) using MF.
Try to listen to a MF once before you dare to compare the bass reproduction with a vented box....
The way wreckingball described his intentions sounded like he wanted to eliminate the ported box resonance and/or driver fs from the passband - that would affect efficiency. But later he stated he didn't want to do that, he just wanted to eliminate the frequencies below tuning, what we usually use a high pass filter for.
Try to listen to a MF once before you dare to compare the bass reproduction with a vented box....
I've got Sony MF speakers sitting right beside me. While they might not be the perfect implementation of MF I find nothing about them to be impressive in any way.