Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 28th April 2010, 01:25 AM   #11
diyAudio Member
 
Join Date: Jan 2008
Location: Toronto and Delray Beach, FL
Explanation for beginners

Whenever the voice coil moves too much or too little (hey, that's called distortion) it raises (or lowers) its back EMF. If you can sense that (across a resistor or better, across a bridge), you've got an error signal which can be fed back to the amp.

Below resonance in a sealed box, the speaker gets held back by air pressure and hence doesn't move as far as the amp output thinks it should. Therefore again, you get an error signal and the amp increases its effort to make the recalcitrant driver move more. Which is exactly what you want.

But for a bass-reflex cabinet, things are altogether different and MF is inapplicable.
__________________
Dennesen ESL tweets, Dayton-Wright ESL (110-3200Hz), mixed-bass Klipschorn + giant OB using 1960's Stephens woofer; Behringer DSP
HiFi aspirations since 1956
  Reply With Quote
Old 28th April 2010, 07:03 AM   #12
diyAudio Member
 
steph_tsf's Avatar
 
Join Date: Mar 2008
Quote:
Originally Posted by bentoronto View Post
Yes, producing what is essentially a mirror-image (positive output impedance, if I am not confused too badly) of the speaker is the goal with bridge.
Hello, I'm afraid we are not on the same wavelength yet.

1. In this thread we are discussing about the control of a loudspeaker membrane without relying on a sensor. The aim is to have no accelerometer, no extra coil winding, no optical pickup, no RF system.
2. There is one post in this thread describing a very interesting RF system. This truly deserves a separate thread, as from what I'm understanding, there was a working prototype and maybe results or issues to be discussed on diyAudio.
3. The Servo-Sound ver.15B still needs to be properly reported. I can reverse-engineer the printed circuit. But I don't know what's inside the little black box next to the trimmer (see attached pictures).
4. The Servo-Sound ver.15B is sampling the current using a small resistor in series with the loudspeaker, and this signal seems to be used in a bridge substracting the I*R component from the overall voltage, hence coming close to the pure motional voltage.
5. If you do the maths, you'll see that point 4 is equivalent to feeding the loudspeaker using a negative resistance power source. Therefore, one can say that the Servo-Sound is a variant of the negative resistance power source. If the negative resistance is different in module from the actual coil copper resistance, the extraction of the pure motional voltage fails. It still contains a dumb (delta R * I) component. This is not a good start for organizing the feedback around it. But we can live with that.
6. More problematic, if the module of the negative resistance is greater than the actual coil copper resistance, you end up with a system that's fundamentally unstable.

Quote:
Originally Posted by bentoronto View Post
Hate to seem anti-intellectual, but a basic issue is that you are messing about trying to make a simulation work. For example, fussing over coil temperature is not a real-world issue. As I said, perhaps too cryptically, if you can grab some feedback voltage, run with it.
Fundamentally I disagree with bentoronto. Point 5 above is indeed "grabbing some feedback voltage" with a planned inacurracy for ensuring that the module of the negative resistance is always smaller than the actual copper resistance. Back in 1957, Werner-Carell were right and prudents in limiting their ambitions in designing a negative resistance power source exhibiting 80% of the actual coil copper resistance. However, I think that the first generation Yamaha commercial AST amplifiers dating back from the early nineties (and associated personality modules) never dared to use such high percentage. Okay, Yamaha engineers job was no easy as the first generation Yamaha AST amps were separated and independents from the loudspeakers. And due to this, those AST amplifiers needed to be compatible with closed enclosures 2nd order, bass reflex 4th order, dual chamber 4th order, triple chamber 8th order aka Bose, transmission line, and so on.

7. One may say that the situation gets better if the amp is integrated in the box, like a closed box. Then we may use a 80% cancellation maybe. And yes, the results are fine. But at the end of the day you will lose your markets. Because a closed box approach is not the most efficient way to generate deep bass. Bass reflex 4th order, dual chamber 4th order and triple chamber 8th order aka Bose are more efficient, with less excursion for a given output level below 60 Hz. You can't ignore that the production cost of a loudspeaker driver is highly dependent of the max excursion he can deliver. You can't ignore that money matters in this world.

8. You may then want to try using the Servo-Sound principle on those other enclosures, and doing so you will need to put a frequency response amplitude and phase shaper in the feedback path. That's tricky. Then your colleage, boss or mentor seing your equalizer in the feeback path, and seing the risk that the wole system gets unstable if something goes wrong over there, will ALWAYS tell you "Hey, we are losing our time. This won't pass the consumer market test requirements. Let's drop this Servo-Sound idea. However, let us recycle it into a system that gets marketed as making a cancellation of the connecting wire resistance - about 0.5 ohm. This way our shareholders won't see we wasted R&D money and time".

9. Then, if you want to deeply extend the bass range, you will rely on a conventional pre-equalizer possibly tuned to the loudspeaker you are using, possibly equipped with a safe operating area module, containing a continuously variable high-pass filter, for avoiding overdriving the loudspeaker in the deep bass extension, at high levels.

10. This is all what will happen (and actually, happened) to people trying to follow or extend the Servo-Sound approach. Those are the consequences of applying the coarse idea of "if you can grab some feedback voltage, run with it". Personal opinion only, of course. See no offence to bentoronto. We are in a constructive debate.

What's the construction indeed ?

11. Well, the merge of two contradictory ideas. Starting from now we are not going to use a negative resistance power source. Instead we will use Peplinsky-Miller AES 1995 collateral finding about the "signature" of a system consisting on a loudspeaker connected on an amplifier having a POSITIVE resistance (or preferably, wideband impedance) EQUAL to the load it sees. There, on page 7 of Peplinsky-Miller AES 1995, you see the main clue of the collateral concept of "signature". Basing on the unity damping factor idea. Quite interesting findings they made, like there is less stress everywhere. Very sexy for the audiophile markets. And they have not used this concept in full. But in year 2010, we will. Let me show you.

12. Multiply the instantaneous current by the intantaneous voltage. If you are in focus, you get a particularity when you correlate this in a specific way with the input signal. This particularity enables to tell if you are in focus. Being "in focus" means that the wideband impedance of the power source is precisely equal (say max 0.1% error) to the load represented by the loudspeaker. You thus know everything about your equivalent schematic. You thus can extract the pure motional voltage. With a precison of 0.3% if you allow another 0.2% error in the rest of the circuit.

13. Then, if you want to drive the loudspeaker using the motional voltage, all you need is to put everything inside an outer feedback loop. Hoping it is stable.

Now everybody should be de-confused about negative and positive impedance power sources, and their use in the context of loudspeakers drive.

Please note that points 11, 12, 13 never been disclosed publicly, so I consider owning the paternity of this drive method (points 11, 12, 13 above) basing on Peplinsky-Miller AES 1995 collateral finding about the "signature" of a system consisting on a loudspeaker connected on an amplifier having a positive resistance (or preferably, wideband impedance) equal to the load it sees. Use it at your will, as I'm too old and too poor for dealing with patents and surrounding sharks and thieves. Who cares about audio and loudspeakers those days ? Seems we are the last Mohicans, here in diyAudio.

Steph
Attached Images
File Type: jpg Servo-Sound 15B PCB comp vue small.JPG (113.1 KB, 464 views)
File Type: jpg Servo-Sound 15B PCB solder vue small.JPG (127.5 KB, 379 views)

Last edited by steph_tsf; 28th April 2010 at 07:14 AM.
  Reply With Quote
Old 28th April 2010, 08:11 AM   #13
diyAudio Member
 
Join Date: Jan 2008
Location: Toronto and Delray Beach, FL
A sealed box is a very good, safe, and clean way to make bass. Power is cheap and it is just a matter of feeding the speaker more power below system resonance esp. if you can have MF as your guide. I'm pretty hostile to tuned boxes even if you can configure the parameters to approximate an OK output curve and being "efficient" by surfing on two resonances with weird phases and amplitudes is not my idea of getting ahead of the game.

Rather presumptuous to talk about "losing markets" with a sealed box.

Even a little MF is a whole lot better than none.

It is inconceivable to make MF except when dedicated to one speaker system, not aiming for the last bit of perfection, and unless it was very very smart. I struggle to find words strong enough to convey my disparagement of the thought of a universal MF amp.

Yes, as I said yesterday, you want the output impedance of the amp to be mirror image of the speaker: 7 ohms when the speaker is 7 ohms and 2 ohms when the speaker is 2 ohms or someting like that. Is that roughly what Peplinsky says?

Yes, Werner talks about making the copper resistance disappear by using the bridge. Acutally just kind of arbitrary - in practice - what you are making disappear and hardly worth getting all too anxious about making the bridge too precisely to simulate the driver. Just grab some feedback, as I've been saying. If you were all-knowing about driver modeling in health and in distortion, you'd be able to build a perfect driver and not need MF.

A very interesting issue in MF is the ability to eat room modes, even if the woofer is not playing music at the time. I've never known enough tech stuff to look into it but a very intriguing "active absorber" idea, eh? Now there is a new idea.

Maybe you are saying something novel and different from what Werner didn't already say and try to patent for RCA. Remember, folks like RCA (or Bose, etc.) try to show that their "invention" is somehow curved differently than plain old obvious state of the art. Current feedback for loudspeakers goes back to the 40s.

Time to get a mic and do some real-world lab work and stop messing with models.
__________________
Dennesen ESL tweets, Dayton-Wright ESL (110-3200Hz), mixed-bass Klipschorn + giant OB using 1960's Stephens woofer; Behringer DSP
HiFi aspirations since 1956

Last edited by bentoronto; 28th April 2010 at 08:15 AM.
  Reply With Quote
Old 28th April 2010, 02:28 PM   #14
diyAudio Member
 
Join Date: Jan 2008
Location: Toronto and Delray Beach, FL
Quote:
Originally Posted by bentoronto View Post
Yes, Werner talks about making the copper resistance disappear by using the bridge. Acutally just kind of arbitrary - in practice - what you are making disappear and hardly worth getting all too anxious about making the bridge too precisely to simulate the driver. Just grab some feedback, as I've been saying. If you were all-knowing about driver modeling in health and in distortion, you'd be able to build a perfect driver and not need MF.
Sorry to quote myself.

Even Werner, if I recall, added a capacitor here and there on the bridge to further emulate the speaker. Problem is that some of those elements you want to model are inherent driver features and some are The Problem you are trying to correct (and hence, you DON'T want to bridge them out of the feedback loop)... and not clear who is who.

And that is why I favor simple, mostly resistance-balanced, bridge-derived feedback.

The greatest progress of hifi has been the enlargement of feedback. MF is the last frontier of that loop by enclosing the speakers in the loop.

Always seemed odd to me that some much energy is lavished on minor issues of hifi in this forum while MF, with enormous benefits, has gotten little attention. Perhaps because it is a messy, cross-discipline area, including electronic, speaker acoustic, and room acoustic.
__________________
Dennesen ESL tweets, Dayton-Wright ESL (110-3200Hz), mixed-bass Klipschorn + giant OB using 1960's Stephens woofer; Behringer DSP
HiFi aspirations since 1956
  Reply With Quote
Old 28th April 2010, 08:45 PM   #15
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
This might be of interest:

The first two pictures could give you some clues about the insides of "the black box": it is a thermal image shunt.

The following pixes illustrate a tentative electrostatic MFB speaker.
Despite its apparent crudeness, it probably deserves some more attention: preliminary tests showed an astonishingly good behavior and stability when enclosed in a relatively high-gain closed loop, without any precautions, calculations or compensation.

See you sometime on Futura
Attached Images
File Type: jpg ThermImg1.JPG (295.6 KB, 353 views)
File Type: jpg ThermImg2.JPG (260.7 KB, 358 views)
File Type: jpg ElecStatMFB.JPG (321.6 KB, 322 views)
File Type: jpg ElectStatMFB2.JPG (320.6 KB, 165 views)
File Type: jpg AmpMFB1.JPG (301.9 KB, 147 views)
File Type: jpg AmpMFB2.JPG (312.5 KB, 130 views)
  Reply With Quote
Old 28th April 2010, 09:57 PM   #16
diyAudio Member
 
Join Date: Jan 2008
Location: Toronto and Delray Beach, FL
Very interesting.

There are main routes to development (in my poor understanding).

One can aim for a piston-action driver (like the KEF B139 styrofoam eliptical woofer). Next, you must ensure the motion is linear with MF. Using the voice coil as a reference, you have 100 years of R&D are behind this approach.

Or you can measure the cone motion by using an accelerometer, position sensor of some kind (like capacitive/RF), photo, or otherwise, with or without smart processing. In as much as drivers have been carefully groomed for 100 years to have the smallest possible distortion, you need a sensor system that is better than the speaker or where the errors of the sensor are benign relative to the ear or the driver. The sensor has to have a range of about an inch. Where can I buy such a sensor cheap?

Otherwise, you might try acoustic feedback. But there are enormous obstacles to getting a mechanical, an acoustical, and an electronic system to work togther.

... and that's why I like the voice coil in a bridge concept.



2. You can
__________________
Dennesen ESL tweets, Dayton-Wright ESL (110-3200Hz), mixed-bass Klipschorn + giant OB using 1960's Stephens woofer; Behringer DSP
HiFi aspirations since 1956
  Reply With Quote
Old 29th April 2010, 08:38 PM   #17
diyAudio Member
 
Join Date: Jan 2008
Location: Toronto and Delray Beach, FL
I like the idea of MF (and Steph's efforts in that direction) that I have one more thing to add... now that my aged memory is piqued.

When you push on an MF woofer cone, it kind of pushes back (sometime just feels rather solid like a wall and sometimes pushes back too much. It's ALIVE. A funny experience. Even had that experience with other woofers?

That is why I long-ago thought MF woofers would "eat" room resonances. Recently I read there's a patent on just such a concept. I wonder it is also eats room resonances while simultaneously making music? Wouldn't that be a cut-above passive woofers!

Again, I am implicitly making a distinction between fun experience of bench-top experimenting versus the dry (to me, at least) engineering approach.
__________________
Dennesen ESL tweets, Dayton-Wright ESL (110-3200Hz), mixed-bass Klipschorn + giant OB using 1960's Stephens woofer; Behringer DSP
HiFi aspirations since 1956
  Reply With Quote
Old 30th April 2010, 08:21 AM   #18
diyAudio Member
 
steph_tsf's Avatar
 
Join Date: Mar 2008
Quote:
Originally Posted by bentoronto View Post
That is why I long-ago thought MF woofers would "eat" room resonances. Recently I read there's a patent on just such a concept. I wonder it is also eats room resonances while simultaneously making music ?
Agree ! It was at the end of the eighties, 1988 or 1989 maybe. I remember taking our Philips RH541 MFB loudspeakers dating back from 1978, in a loudspeaker shop for comparing and buying new speakers like the newest Bose, B&W, Cabasse, Celestion, Dynaudio, JM Lab, JBL, KEF, Mission, Philips, Pierre-Etienne Léon, Sonus Faber, Sony, Wharfedale, Yamaha.

After one hour comparison if was very clear that our Philips RH541 MFB loudspeakers were reproducing the human voice 10 times better than any other speakers. You could not find a speaker having a decent deep bass response, and still capable of accurate, realistic human voice reproduction. On all those conventional loudspeakers, bass was always in excess on the human voice, in a strange variable way. Hard to describe. The more we were listening, the more we had the conviction that the Philips RH541 MFB loudspeakers were taking control of the room, making it sound flat, less reverberant, and then, on such flattened and damped environment, delivering their audio signal. Even the owner of the shop was agreeing with us, about that.

We were fully aware that the Philips RH541 MFB were not perfect, with a high range being far from silk-smooth, provoking listening fatigue. Some speakers in this shop were vastly better than the Philips RH541 MFB, about smoothness in the high range, and we were agreeing with the shop owner about this. Globally, at the end of the process, we felt more comfortable with the Philips RH541 MFB listening experience, than with any other speaker that was in this shop.

We came back home without purchasing anything new. I remember the conversation with my dad, in the car upon coming back home. We developed the idea that the Philips RH541 MFB was eating the resonance modes of the listening room, as an undisclosed feature. We asked ourselves why Philips did not promoted that feature for marketing those loudspeakers. We concluded that it was maybe impossible to objectivate this feature.

Today in 2010 my parents are still listening to those same same Philips RH541 MFB loudspeakers. They are now connected on the TV set, where human voice reproduction accuracy matters.

In 1991, I bought Philips best available MFB loudspeakers, the F9638 models with flat cone woofer, flat cone medium, ribbon tweeter. It proved unreliable with problems with some connectors inside, and it was not featuring the impression that it was eating the resonance modes of the room. The sound delivered by the F9638 was dull. Less listening fatigue than the RH541, must say. Around year 2001, I couldn't cope with the reliability issues of the F9638 anymore, so I carefully dismantled them for re-using the transducers in a custom-made version. I dropped the project upon discovering that the flat woofer membrane was faked. Actually, it was the usual MFB woofer, now with a flat styrofoam plate glued on the front. Also, I got dissapointed when seeing how small was the excursion of the medium driver. I only kept the ribbon tweeters. I dumped the rest. I think that the F9638 was not eating the resonance mode of the room because of the faked flat membrane woofer. In such configuration, the accelerator sensor doesn't face the listening room anymore. There is that flat styrofoam membrane inbetween.

At the moment I own 4 or 5 MFB drivers type AD 7066/MFB
2422 257 47005
DD 00 946 126
packed as servicepart
4822 240 50099
Still waiting me to design something around.

I would appreciate if somebody points a link or posts information about the objectivation that some MFB speakers have the capability to eat the listening room resonance modes. Is there a patent about this, really ?

Cheers,
Steph

Last edited by steph_tsf; 30th April 2010 at 08:51 AM.
  Reply With Quote
Old 30th April 2010, 08:53 AM   #19
diyAudio Member
 
Join Date: Jan 2008
Location: Toronto and Delray Beach, FL
Quote:
Originally Posted by steph_tsf View Post
snip

I would appreciate if somebody points a link or posts information about the objectivation that some MFB speakers have the capability to eat the listening room resonance modes. Is there a patent about this, really ?

Cheers,
Steph
In MF (esp bridge type) the speaker is a mic too.

Is this relevant? It was kindly sent to me by a distinguished European poster.

Audio Engineering Society
Convention Paper 6520
Presented at the 118th Convention
2005 May 28–31 Barcelona, Spain

Active Acoustic Absorption and Reflection
Peter A. Swarte
P.A.S. Electro-acoustics,
Eindhoven, The Netherlands
e-mail: paswarte@cs.com
Abstract
The subject called "Active Absorption and Reflection" has already been theoretically discussed during one of the scientific sessions in 1986 of the Dutch Acoustical Society (NAG) in Delft, The Netherlands. The idea behind the subject was led down in a pending US patent. The patent has been granted to Philips in Dec.1987. The principle with the possible applications is described in publication nr 84 of the NAG d.d. April 1987.

Many others have investigated the phenomenon: a search on the Internet delivered in 2001 all publications of the Drittes Physikalisches Institut of the University of Göttingen, Germany. A lot of these publications had underwater
acoustics as application field. It was not an easy matter to give proof of the feasibility of the idea at that time. Nowadays, the technical possibilities
to do investigations and to gather convincing results have dramatically improved. With the help of adequate software implemented on a high quality personal computer and a special acoustical set up, the measurements can be performed in order to collect results that give a convincing support to the theoretical background. Two types of transducers i.c. a dynamic and an electrostatic (electret) transducer have been
investigated.
The transducers are an element of an electronic positive or a negative feed back circuit. The feed back circuit has to adapt the reflecting or absorbing nature of the diaphragms of the transducers to the desired degree of acoustic absorption or reflection. It is obvious that, at the same time, the transmitting properties of the diaphragms are also influenced. This phenomenon was in fact not a subject of this feasibility study. But a quick
investigation has shown that the transmission of sound is indeed influenced by the feed back control circuit.
In this feasibility study, the measurement set up and the results are explained and discussed. The results are encouraging. But this is only the beginning; a couple of investigations must be executed in order to
arrive at a manageable product. The products are expected to cover a wide range of acoustic and electro-acoustic applications.
__________________
Dennesen ESL tweets, Dayton-Wright ESL (110-3200Hz), mixed-bass Klipschorn + giant OB using 1960's Stephens woofer; Behringer DSP
HiFi aspirations since 1956
  Reply With Quote
Old 30th April 2010, 09:23 AM   #20
diyAudio Member
 
steph_tsf's Avatar
 
Join Date: Mar 2008
Hello,

in may 1985, I was at the RTBF (radio television broadcast of belgian french communauty) as student in applied electronics. Everywhere, there were Philips MBF loudspeakers used : RH541, RH544 and RH545. So why ?

1) they used parts made in Belgium
2) the RH541 delivered an unbeatable quality/volume ratio in mobile use
3) the RH545 delivered the most consistent result with headphones, in control room use

One of the RTBF sound engineer took some time explaining me that the Philips RH545 was not the "best", but in reality "the only" system providing a nice consistency in a control room, not forcing you to reassess everything when you switch from headphones to free field (loudspeakers) and vice-versa, especially when balancing human voices.

The concept of the MFB "eating" the resonance modes of the listening room was thus not there, back in 1985.

I would like to know if the industry made some progresses in 25 years. Are there loudspeakers on the market, providing a better consistency than the Philips RH545, when switching from headphones to free field (loudspeakers) in a control room ?

P.S.
And thus, it is now clear why the Philips RH541 were so good at home and in a hifi shop on human voices (radio program listening) between 1980 and 1990. The RTBF sound engineer was balancing the sound, on his side, in the control room, using Philips MFB RH545 units, quite the same bass rendition as the Philips MFB RH541.

Cheers,
Steph

Last edited by steph_tsf; 30th April 2010 at 09:35 AM.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Ultra sound transducer for servo feedback subwoofer? e96mlo Subwoofers 28 11th July 2013 04:59 PM
Philips made in Belgium, new production engels Multi-Way 40 22nd July 2010 02:20 AM
Hello from Belgium pigsnoot Introductions 3 13th June 2009 09:44 PM
..didn't know how they sound till i made me one.. tmyr Planars & Exotics 137 26th November 2006 09:01 PM
home made surround sound Hitman Digital Source 7 22nd December 2002 06:15 AM


New To Site? Need Help?

All times are GMT. The time now is 06:12 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2