Servo woofers are rare these days and this questioner wants to know how they work. And check out our newest RU-vid channel / @octaverecordsanddsdst... Octave Records.
I recently came up with a design for a piezoelectric controlled mid woofer, and after some research figured it wouldn’t work well because nobody does it. This video put a big smile on my face, thank you
Thanks for that explanation and history lesson. For years I was under the misapprehension that servo=motor and imagined some kind of pistonic engine that moved the cone rather than a magnetic voice coil!
I was just trying to explain to my cousin why servo woofers are just better than your typical options. This video did a much better job of explaining it than I was doing. My next purchase is likely going to be a pair of GR-RESEARCH open baffle servo subs. Should go well with the magnepans when I find a pair for the right price.
Agreed. I've found the YAST II system to produce some nice clean bass. Is it the most powerful or the absolute lowest, nope. It does however just sound really good. Running two Yamaha subs in my HT system. Got rid of my boomy Polk's.
Never heard if servo woofer. Paul is a well of audio knowledge and experience. Watched this like 4x already..and I know I'm going to watch it again. This is fascinating!
Wonderful explaination! I could listen for hours. I am currently part of a team, designing a similar servo loop system at work, but for automotive suspension - i thought i knew control theory well... well it turns out i still have a lot to learn ;) i wish i could afford having one of those speakers, but i have to buy a house first :)
In the Netherlands there is a group of (Philips) MFB enthousiasts that are reinventing the principle to modern times. for instance using Hypex modules with with DSP and using this processing power to do the servo. Also the use for small active monitors with deep bass is being developed.
Not sure if Mr. McGowan was involved but I have a Genesis Servo 12B subwoofer that was purchased in 1995! It still does the job well. I like the fact that the control box/amp is separate so adjustments are easy.
back in the 80's i drooled over a set of Linear Power servo controlled car subs. 4 8's as i recall. As i recall they used dual voice coil woofers one was used for power and one for sensing.
Makes sense, presumably that'd also help cancel out the affect of the cars movement and vibration on the speaker and perhaps even decrease the noise inside the car
The the fastest most accurate subwoofer I have ever auditioned was an pair of light weight paper cone eminence 12 " 's in 2 8 foot boxes tuned to 16 hz. No servo control. Zero equalization. They used a very basic passive crossover. They also needed resistors to cut the efficiency back to the level of the Martin Logans they were running with. Effortless , efficient , quick , accurate, Cristal clear bass -3 Db at 16 hz. Very few people have room for 2 coffins in their living rooms.
smart dutch people, they invented a lot of cool thinks, coffe-pad machine, one blade razor this is the best electric razor, but i use a double edge razor, and airfloss
I studied electronic and electrical engineering, have owned, designed, built, and modified quite a few passive/active systems myself; and have always had an avid interest in audio and acoustics. Years ago I found myself working in a UK studio (this day we were repairing/tuning a well know brand of large 3 way far field active monitors) where the system we took out of the wall actually had a small laser attached to cone baskets of its woofers. The laser reflected off a strip attached to underside of the cone, and from this calculations were performed within a microprocessor to implement the correct feedback/control signal. Prior to this I had often thought about servo control of both woofers and other transducers beforehand, but the required processing power to achieve such things (at the time) was theoretically only (just) achievable within the low frequencies; and - whenever contemplating such a system - I was always aware that a closed loop feedback control system would still most likely be unable to deal with surface mode distortion across large diaphragms (such as with 12 inch and larger woofers) anyway. Anyway, this was the first time I had seen it implemented with a laser, and the surprise was that the monitor manufacturer was testing the technology and the studio themselves were unaware it existed. I recall this distinctly as I was the one that had to contact the manufacturer and ask about a servo electronics that was not in the engineering handbook. After several weeks of argy-bargy with the manufacturer they eventually/reluctantly mailed me a circuit diagram for this unannounced servo system (previously all they supplied was a circuit diagram for the amplification, Xover, and other circuitry) that just happened to “show up” in their product line, as complete documentation of all circuits and other acoustic systems so as to facilitate maintenance/repairs was an explicit requirement within the original purchase and support contract. Despite being reasonably crude, the servo system worked very well and (within certain SPL’s) it also measured better than other identical systems (which were implemented within the adjacent studio) from the same manufacturer that didn’t have low frequency servo control.
I built a motional feedback system in 1980 using a Philips AD8067MFB driver that had the accelerometer under the dust cap. It was the same driver that Paul Barton used in the PSB Beta II. The reduction in the non-musical 3rd harmonic distortion in this older (and not particularly linear) driver was huge. The human ear is a lot more sensitive to the frequency of the 3rd harmonic (say 60 Hz) than the fundamental (say 20 Hz) and since even today's good LF drivers generate quite a bit of distortion the reduction in distortion is significant and apparent.
Years ago, my audiophile friend described to me, if I remember correct, that "Amplifier Damping Ratio" works on the woofer to put it where it is supposed to be. The servo idea, as you describe, seems to do something similar, but more accurately. Currently I can not define damping ratio but maybe you could enlighten us on it and show how it effects the speakers.
BiddieTube, negative feedback does essentially the same thing but more simply and less expensively than the unnessarily elaborate servo design, often with better results. Decades of testing back me up on this.
This servo approach is used in CNC machines. In older CNC machines they used a tachometer feedback on a DC motor. The tach tells the amplifier how fast the motor is spinning. These are typically velocity mode drives, where the input voltage corresponds to a given RPM. The tuning parameters are "how fast do I react to input changes" and "how much do I try to control overshoot". This is just like a speaker. Normally the output of the speaker is proportional to the volume of the music. A loud volume will move the speaker faster than a quiet volume. DC motors work the same, a higher voltage will accelerate the motor faster than a lower voltage. So the idea is you put as high a voltage as you can, with as much current as you can, instantaneously, to get the armature (voice coil) to move the cone to the desired position as fast as possible. If you listen to speakers, you'll note there is a sweet spot where they start to sound much better, at low volumes they tend to sound muddy. Servo control essentially allows a speaker to sound good at low volume and high volume. Imagine a very dynamic musical piece that has quiet passages and loud passages. You crank it up when it's quiet, so you can enjoy the passage, but then it's too loud for the loud passage. A servo woofer would eliminate this need and it would just sound "good" all around. The intended frequency would be the same at low volumes as well as high volumes, just the intensity would change.
@@eformance your an idiot....volume doesn't change the speed in which a speaker move but the velocity the frequency of the music tone being played determins the speed
Those are some awful strong words. You have velocity, travel, and frequency. Power input (volume) influences the velocity of the cone and the travel of the cone, otherwise there would be no purpose in the servo technology Paul is talking about.
This is something I've never even heard of. It's a great idea, and would work very well for frequency response. I had a harebrained idea about making a voice coil activated port to tune the frequency of the box to the output of the speaker. That way, the speaker is always hitting its peak resonance within the box and port. It might help to make the system more efficient. Combine that with this technology, and you'll have an energy efficient clean audio response.
That wouldn't work because the speaker isn't only producing one frequency at a time. It's producing a whole bunch of sounds, not just one sine wave at a time. To make a speaker more efficient you make the cone lighter, the magnet stronger and the magnet gap smaller. Amplifiers are cheap these days in terms of dollars-per-watt. There isn't much need for ultra-efficient speakers anymore.
@@devilsoffspring5519 I was thinking more on the low end of speakers, the subwoofer. Have a low end filter that would interpolate the frequency like a sort of artificial auto tune. Especially with ported boxes that are tuned to a specific frequency. If you can make the box smaller, that's always a good thing. And a flatter frequency response is also good.
@@squishy312 There's no need for it, because ported speaker boxes don't only operate at only one frequency. They work over a range. The frequency that the bass port is tuned to is the resonant frequency, but their sound output is useful over a range of frequencies. So, no need for a "self-tuning" bass port that adjusts itself depending on input frequency.
Wow, this idea is amazing. Such a simple idea too. Making a speaker produce perfect square waves in the real world is impossible, but getting close to them is impressive! The only thing - if it's reading the speaker and applying a correction signal, it's surely lagging behind the audio a bit? Maybe only a few cycles, but how much time does this correction take to process?
No not a few cycles, but you are right there is some lag. However it can't be more than a few degrees. I understand that the reason we don't see servo woofers go very high in frequency is because the lag between the electrical signal and the woofer cone gets too great. If the signal and the motion lag close to 180 degrees you now have oscillation instead of control.
Brilliant! I'm thinking now, if you had a really good woofer with perfect linearity, powerful drive assembly, lightweight moving mass etc, and the accelerometer was absolutely accurate 100%, how do you compensate for the amplifier characteristics - slew rate and so forth? This technology is fabulous, I'm just wondering if it will expose weaknesses to the other components in the chain. Even measuring the actual output of a perfect servo controlled speaker with a microphone for example would add its own response characteristics to any measurement. Exciting times to be exploring things like this with modern technology, the engineers of old would tip their hats to this stuff!
Still have a pair of French 3a Master Control speakers with servo woffers i bought in 1980. Wish they still worked. Ribbon tweeters. Amazing sound. They alsi made a four driver sub of servo drivets that was also a coffee table.
Servo's , great decription of how servo's work . One of my labs in Electronics Eng was on servo feedback loop math . Now there's and interesting world of Calculus !!! Interestinng that you can get that much feedback loop control using a piezo , didn't know that !!!
I have Philips MFB speakers from the 70's, they work the same way. The accelero-meter is just in the feedback loop of the amplifier. 3 way speaker with 3 amps, 1 per chanel (active speakers). They had a whole line of those. MFB was in the line name.
What do you think of the Event Opal woofer technology that has another coil but it's out of phase, I think it works in the same way where you tense up your muscles while opening a bag of chips so when the bag finally opens the inertia of your arms doesn't fling chips around the room.
To summarise this video: In normal speakers, the actual physical position of the speaker cone is not being measured at all while it's being accelerated back and forth. "Servo" (not related to motor servos at all, apparently) woofers simply measure the cone's physical position with accelerometers or piezoelectric pressure sensors, and then use a PID system (look it up; very important and common concept) to adjust for where the speaker SHOULD be according to the input signal at any point in time, and then send that "difference" as a correction signal.
Very much looking forward to the AN series. Do the man justice.... I know you will. I let a used set of Genesis slide in anticipation of your upcoming effort.
Almost all modern amplifiers incorporate negative feedback circuitry, which increases damping factor (the ability of the amplifier to correct for aggregate motion error in speakers), which generally eliminates the necessity for complicated servo controls on speaker design. Tests have shown that servo speakers preform almost the same as properly-implemented high-quality speaker and amplifier pairings which cost only a fraction of what a servo-controlled system costs.
Are you referring to the Grotto when referring to the Martin Logan servo subwoofers? Any opinions on them? Would love your take if you're familiar with them.
Did you ever try using a hall effect sensor for measuring the travel of the woofer? A magnet on the woofer, with a hall effect sensor to measure the field strength of the magnet as it travels to and fro. This would give you a proportional feedback relative to the position of the cone. The distance would be proportional to the power input, but the rate of change would mirror the input signal.
@NDFOOTBALL it's not the imaging it is the control of the speaker cone itself and if you have heard a servo controlled sub they seem in a class of their own.
@NDFOOTBALL Can you please share what model number of the QSC you use? I am having a hard time finding powerful, well designed 8 inch subwoofers at a reasonable price.
Ideally all speakers should have feedback. Probably some day they will be more available at the market. And, Paul, the correct answer would be - people just don't care and prefer to put more in marketing than in development of more perfect speakers, that is reducing the production expenses and increasing non-essential features to sell things at higher price. From my experience of messing with them, they just avoid the topic, while showing the indication that they 100% realize that it can be done at the cost of some more expenses.
Given this auto EQ that is applied to the woofer signal, if you're running an EQ for the room, will this Auto EQ allow for these alterations or will it attempt to compensate? Is there a maximum given boost in the attempt to give a flat response from the woofer? Does it run passively? How does it alter the signal if it is running passively? Your explanation intrigues me to know more.
yep, and naming it Servo they found a way around the patent Philips had on MFB. When the patent expired Grimm audio started using DFB (Digital motional feedback) on some of their woofers. If you ever have a chance listening to their ls1be.. grimmaudio.com/hifi-products/loudspeakers/ls1be/ (just like Philips, Grimm audio is situated in Eindhoven .. )
I'm not Paul but I'll respond. I haven't had good results from chained speaker connections. Currently using separate pre outs to my powered subs. Also send the whole frequency to the subs. I tune the crossover level on the subs. Listening to something like the Birdman album is great.
Interesting! I didn't even know these existed. Couldn't the servo feedback thing be accomplished just using the voicecoil as a means of determining what the cone is doing, I mean, it's just a linear motor, so after initial calibration, measuring the voltage across the speaker being the lowest when it's either trying to change direction, or move off from a static position (i.e. drawing more current), or the highest when it's about to overshoot (i.e. voltage from amp is zero and the speaker coil has become a generator) Of course I'm no electrical engineer, so this may just be me taking nonsense. edit: Actually, brain fart time, one could have a second voice coil for feedback, and just measure the voltage across that, would make things easier.
What about using the main coil as source for the feedback signal? The emf coming from the loudspeaker coil should be proportional to the speed of the membrane. A "simple" differentiator [i.e. transfer function F(s)=s] following that emf should provide the needed acceleration feedback voltage.
Or even a sepparate coil next to the voice coil (on the same cilinder). Or on top of tve voice coil (with thin wires). That coil should produce a voltage only when you have a distortion and if you add his voltage to the input signal (matching levels) you have all the servo you want :)
Great discussion on servo controlled low frequency drivers. Can this circuit be passive or must it have its own dedicated amplifier requiring an AC power supply and source?
Technically it could feed data to digital processing unit before the amplifier, so you would need just 1 amplifier, but your digital processing unit should have data and software for that particular speaker.
Hi Paul, this is the first time i've heard of a servo woofer. Sounds like a neat idea to get really taught bass. You mentioned the piezo device that is relatively cheap but what is the lifecycle of that unit due to it's constant flexing before it would break or need replacing? thanks Kym from UK.
Yes Velodyne made them way back in time. The ULD (ultra low distortion) series had them and yes, they were expensive. They did claim very low distortion. I wonder how they would stack up against the high end subs of today, that do not use this technology but use incredibly powerful amplifiers and massive magnet structures (SVS subs for example).
I wonder if this applies to the Kenwood KAC-818 'sigma servo feed back'; had the amp (still do) as a teen and hooked up the extra lines a second set of inputs on the box to the extra tabs that were on the Radioshack 18" I was given. Seemed to control that well but had no idea if it was doing anything. Maybe I was lucky and just happened to buy a frugal amp that matched the free 18" my dad built a small sealed box for to fit in the back of my 88 RX-7. I always thought later on it might have been a sales gimmick so looking it up now :p
Instead of using a piezo element, could the same thing be done with a dual voice coil speaker. One coil for driving the woofer, the other for sensing movement?
Hoi Paul, my name is Hein van lokven and I am from the Netherlands. interesting item! But this system has been developed by Philips in the early seventies. It was know as motional feedback (MFB). I have several sets and I love them. So is your servo system different then MFB? Greetings from Holland.
Very good question, philips was way ahead of its time. And they performed the MFB using only simple transistors. No fancy op amps or DSP. Pure analog. And i also have a set of them the 544 and i find them pretty good. I love the thight bass.
Can those things be implemented on any speaker or a special speaker has to be built with the piezo element? I'm from The pro audio industry and i agree that subs and bass reproduction systems on the pro market aren't actually in the race with such drastic improvement Of low frequency distortion reducing and time impulse response improvement. Can we bring those things on boat for pro audio?
, I have the upmost respect for you being a professional and a pioneer in the sound industry. I have only one question, why in the hell do manufacturers use such small woofers now a days? I have a 35 year old pair of Sansui 12 inch woofer, 2 four inch midranges, one soft done tweeter, and one horn loaded super tweeter like JBL and Klipsch uses today, they will and do outperform any speaker that I have purchased in the last 30 years. I have JBL, POLK, ELAC, and a shit load of other very expensive speakers and none of them can hold a candle to the sound of my old 35 year old speakers !!!!!!!! I found that only a hand full of companies make speakers with larger woofers like Cerwin Vega, BIC, and some very expensive JBLs
There is a reason. Directivity depends on the size of a diaphragm. A driver that is relatively large to a wave it's producing will have a polar response in a form of a comb filter. 30 years ago they realised that directivity is important. A good sounding speaker is such that it's polarn pattern is smooth and close to constant. To achieve that a woofer needs not to be much larger than a tweeter. So nowdays both woofer and a tweeter on a speaker are smaller and closer in diameter. They make up for it with increased excursion. So in the end they maintain the same displacement, but fare much better in directivity. In adition large cones have problems with modes, breakup, nonpistonic motion. They're used in PA, subs, IB, but they're not good for high fidelity. As to why you like your Sansui more I have no answer ;)
Marcin, you are full of shit!!! You cannot have a woofer the size of a tweeter and expect deep bass, especially without needing gobs of power to achieve a respectable SPL due to the fact that lots of "mini-woofers" tend to have shit sensitivity ratings like 84dB at 1w/m. Deep bass requires the sweeping of large volumes of air. If you consider a 2-1/2" cone with an x-max of 5mm to be a woofer then you are deluded since the best kind of response you will get from a driver like that will possibly reproduce 100Hz at moderate volumes before the voice coil is bottoming out in the gap. That would make it a low-mid driver at best. If you really believe that large cone bass drivers cannot be hi-fi then you are truly lacking experience and exposure. I have a first gen DVC 12" JBL GT sub in a homemade custom box being driven by an Emotiva amp and it is still the most musical sub I've ever heard until you start bringing in reference subs costing many times what I paid. Only decent headphones can reproduce low bass with tiny drivers, but those aren't cones per say. They are super light weight diaphragms that easily respond to the full spectrum of audio signal seeing as they are easily manipulated by even the lowest amounts of power, yet their sensitivity is still so low that no one would ever consider them for for anything beyond "on ear" monitors.
Tim, I think older speakers like you describe have a certain quality to them that lots of modern speakers don't (though there were lots of junk speakers back then too) because the people who designed them were using a two approach method. First they had to design the physical and electrical parts (drivers, boxes and crossovers) and then they had to fine tune them by ear since there was no computer feedback. They had to sound like real life or close to it. Modern, mass produced speakers are simply made to try to get a ruler flat response across their spectrum. They may do what the amplifier tells them to do, but they don't "sing" like the oldies did. That said, you can still get some mighty nice sounding speakers for a fair price if you do your homework, but there are tons of them to wade through now so it could take a while to find some you like the same as your old ones if you are looking for new.
I'd suggest using non touch measurong method such as laser proximity measuring the cone's position. Besides, it is what is done inside the accelerometer. By implementing the laser proximity sensor of the cone relative to the frame we get much more accurate data amd faster on realtime, without adding any weight to the cone
I wonder at what size woofer you don't need a servo motor. If you have a newer designed modern day motor with a vented ported system do you even need a servo or is it extra cost with little benefit with smaller better designed drivers. At what size do you need this 10inch, 12inch, 18 inch etc. And how does it affect speaker designs in a two way with a need to go into mid range frequencies. Is this really a very high end option for clean bass thats really not cost effective for the vast majority of applications.
An accellerometer cannot measure the periods of standstill at maximum excursion during the two half-cycles of a square wave, so that rules out square waves.
I have owned a lot of mfb speakers from Philips who invented this system. Best bass ever but the mid and high were also very 70’s. So I ditched them. A theoretic problem is the slew rate because of the electronic circuit. In reality not a problem bevause the slew rate without the electronics is already longer.
HI, WHAT ABOUT THE CHANGE OF IMPEDANCE FROM THE MOVEMENT OF THE VOICE COIL OVER THE MAGNET. LOSS OF POWER AS THE OHMS CHANGE. HOW DO YOU COMPENSATE FOR THIS
This SERVO Idea is nice technology-and-all.. However, if the studio that created the MIX that is finally impressed on the CD/Vinyl OR Whatever storage medium did not have this listening/monitor equipment at the time of creation, it might be hard to recreate what the MixerMan/Producer heard when the Medium was created.. In my humble opinion, the best way to recreate what was originally created on the storage medium would be to have the exact reproduction equipment that the medium was created with.. Most everybody has a 4" speaker OR a set of Earbuds.. However, if the audio wasn't created-or-heard with the same listening devices, the audio will not be reproduced the way it was intended.. I hope I passed that Thought/idea along, correctly..
By your logic, it is useless to aim for lower distortion than the supposed level of distortion heard by the recording engineer. Sure, everyone should listen to Yamaha NS10s...
Even cheap Yamaha mass produced subwoofers have Servo Mechanism they brand it as YST II. But the cost of Servo Logic is relatively low today. Its the fine detailed engineering R&D work to execute it well that i'm guessing is majority of the cost. Servomechanism is just a basic concept, but has to be done right.
LOL, they were still selling the SM series to service men in the late 80s. Sounds like one could convert older speaker system to a servo system - maybe there's a DIY market you folks could fill. As you described it, my very first thought was "why not use a laser to measure the movement directly?". But your solution is MUCH cheaper, even if a lot more complex. How significant does that correction signal have to be compared to the original signal? Let's say I have a 2.2V square wave at 40Hz. (or specify your own numbers, if you like), what would the voltage of the correction signal be? (I specify the square wave knowing this would require the largest correction signal). So if I'm hearing you right, all your servo woofer systems are powered and have their own amp. I noticed you didn't mention the mass/stiffness conundrum. I've seen woofers made out of all sorts of material, even titanium. And even with your servo system, the woofer cone would still flex, causing distortion. Now you need to find a more rigid cone material to further reduce distortion.
Disculpe si usted pudiera ayudarme tengo unos subwofers acoustic sbt 2.6 servo y no se cómo hacer para se mueva la bocina como se tiene que mover si usted pudiera decirme que aplificador usar
We're talking about powered woofers or subwoofers, so this works because the amp design can include the additional difference amplifier there in the same circuit.
Woofers and subwoofers can be controlled very well with servo because you are working with low frequencies and the servo has time to react and compensate. As the frequency goes up it becomes impossible to fix the problem because the problem has already occurred.
Great info. I want a MUSICAL SUB. I play along with my guitar and AV Receiver. Is there a sub and or driver you might recommend? Size, brand, type, etc. I don't need it loud- 80dB max. I want it clean.
So I'm wondering about how this works with music not mastered with a servo woofer. I'm a student of music production (even though I'm no young'un) and know that mastering engineers use very carefully chosen components an speakers to put the final touches on music, with a certain expectation of what the listener will use (hence, checking mixes on various players, in the car, ear buds etc.) I don't know of anyone that mixes or masters with servo woofers in mind. Can they be too revealing for music not masters for these? Or is the reality that difference is generally a net positive?
@@NVMDSTEvilHi! I was talking about TL, Transmissionline. But with servo, you probably could doing smaller enclosures and don't need to know a lot of stuff about damping...
I remember a long time ago when Velodyne servo subs first came out and I was able to audition them. Scary bass. If you didn't know what low bass was they would take you to school. So I just looked them up right now and the are still in business but their servo subs are $4k-$6k! Continuing to search, I then found brand-x for $750 (F12-300 ). Why such a price range?
