Are you that full of hubris that you believe that you have more insight into how drivers are made than manufacturers that provide products for professionals? Take a look at stuff from the top pro sound manufacturers and tell me what's wrong with modern drivers.
There's plenty of information to look at including measurements of every type - klippels, CEA testing, compression sweeps, enough info to get a very good picture of what these drivers are doing at all power levels.
If you have a better solution then please present it. This is the subwoofer subforum of a forum with some members on the leading edge of the science. Suggesting that the primary tool used here (moving coil drivers) are a primitive mess demands some type of actual criticism based on evidence, and your argument would go a lot further if you actually had some kind of alternative to suggest. Constantly whining about moving coil drivers without any evidence that there's even a problem is not productive.
If you are talking about the Bagby quote, these are not errors that are counter balancing. These are shifting t/s parameters that shift in complimentary fashion as power is increased. i thought this was common knowledge, at least it should be. How else could you account for the fact that t/s parameters shift so dramatically at different power levels but yet compression sweeps show little to no change in the response curve until high power levels? The fact that you read the information and misinterpreted shifting t/s parameters as errors says a lot about how this discussion got to this place.
Is there a problem with presenting evidence that I didn't measure myself? If that's the criteria we must uphold I don't think you have presented anything of value that I can recall.
I'm sure it gets old really fast for you, as many of our recent conversations have not gone very well for you. Instead of posting my resume and years of experience over and over again I present actual technical information, quotes, links, measurements, simulations, whatever it takes to back up my position. This would get really old really fast for people that cannot defend their position in a technical manner.
Beats me why anybody in the world would think some stiff fabric and a roll of rubber makes a perfect spring
While I doubt if many good designers actually believe they are perfect springs, that assumption is cooked into the model and necessarily functions as an unqualified working basis in the absence of better models.
Ben
Are you sure you understand how cone speakers work (and are simulated)?
While I doubt if many good designers actually believe they are perfect springs, that assumption is cooked into the model and necessarily functions as an unqualified working basis in the absence of better models.
Ben
Last edited:
RonE has posted several formulas that I've seen that are the same information that simulators calculate so I'm pretty sure he knows how simulators work. I'm pretty sure he knows how drivers work too.
You've already been informed that nonlinearities like suspension nonlinearity are covered by other measurements like Klippel. The effects are well known and the sims still match the measurements so again, this is a known issue and it isn't that much of an issue.
Congrats, Ben, on another complete and total thread derailment. And you've once again turned it into a perfect trifecta of your favorite three themes:
1. The evils of resonance (you've even managed to include a backhand swipe at tapped horns)
2. The validity and inaccuracy of simulations (your weird idea that simulators COMMAND people to build certain alignments and provide inaccurate results)
3. The primitive nature of moving coil drivers (even though you use them yourself and they work as designed to provide very good results)
This latest attack on moving coil driver suspension nonlinearity is ridiculous. It's like saying that our knees only bend in one direction so they are useless, how could anyone think they are perfect, we should stop using them to walk. I don't have any alternative suggestion to accomplish the function of our legs at this point but knees are not perfect so they should be abandoned.
This weird world of misinformation piled on uninformed opinion that you create so often with your complete thread derailments to discuss these same issues over and over is amusing but it's all be covered before, several times, every time you've destroyed a thread in this fashion. The suspension nonlinearity is a new theme, I'll give you that, but you are grasping at straws. You need to point out a specific MEASURED flaw, not attack the moving coil concept piece by piece and suggest certain parts should be changed with no suggestion of how to improve them.
Planet 10, I kind of believe you hit the nail on the head without even knowing you were swinging the hammer. A speaker is a mechanical system with a mechanical resonance frequency which does not change much with amplitude until some sort of clipping begins to occur. You should look at your sealed box mechanical resonance frequencies at different levels easily found with a small series resistor measuring to find the peak voltage across the speaker or the minimum current. This frequency does not change much at all with amplitude in a sealed box with a not brand new driver. If the resonance is the same then it is likely the parameters are fairly stable which define that resonance.
Others, I do not "poo poo" the T/S alignments as those are simple straight filter theory and if implemented should show the predicted response. It is the Small method of determining T/S parameters which is usually useless. The electrical impedance is a reflection through the motor system of the mechanical reality of the speaker. Many factors affect this reflected impedance. This is why I prefer direct measurement of parameters rather than the Small method. The Small method can be quite erroneous which is why there is the Fudge Factory paper published in AES Journal.
Try reading some Micheal Lee and Susan Lampton to move toward using a mechanical version and use my measurement methods which in reality are not that difficult for a craftsman. Never mind I use a dial gauge ticked off in units of 0.0001 inch. A good site line will work too.
As I never saw it, the mechanical resonance is exactly at the frequency of highest impedance. The electrical resonance is at the frequency where Xl=Xc which means the current and the voltage are in exactly the same phase. These two frequencies are never the same and are one of the major sources of error using the Small electrical test method for woof parameter measurement. Electrical is always higher in frequency if I recall correctly.
Others, I do not "poo poo" the T/S alignments as those are simple straight filter theory and if implemented should show the predicted response. It is the Small method of determining T/S parameters which is usually useless. The electrical impedance is a reflection through the motor system of the mechanical reality of the speaker. Many factors affect this reflected impedance. This is why I prefer direct measurement of parameters rather than the Small method. The Small method can be quite erroneous which is why there is the Fudge Factory paper published in AES Journal.
Try reading some Micheal Lee and Susan Lampton to move toward using a mechanical version and use my measurement methods which in reality are not that difficult for a craftsman. Never mind I use a dial gauge ticked off in units of 0.0001 inch. A good site line will work too.
As I never saw it, the mechanical resonance is exactly at the frequency of highest impedance. The electrical resonance is at the frequency where Xl=Xc which means the current and the voltage are in exactly the same phase. These two frequencies are never the same and are one of the major sources of error using the Small electrical test method for woof parameter measurement. Electrical is always higher in frequency if I recall correctly.
Last edited:
The Lea/Lampton paper is nothing but a rehash of filter theory. google for "the theory of maximally flat loudspeaker systems" and you can download a nostalgic digital copy for free 😉 It does nothing that thiele didn't do 21 years earlier. Q=0.383 reveals that they are using an undamped vented box model, just like Thiele.
I have yet to see a speaker that has an impedance peak and zero phase point that are different to more than measurement error. Maybe this is some artifact of your measuring at 1/2 xmax (distortion, etc...)?
The vented box tuning frequency minimum and zero phase point are often different from each other. Inductance is the culprit.
I have yet to see a speaker that has an impedance peak and zero phase point that are different to more than measurement error. Maybe this is some artifact of your measuring at 1/2 xmax (distortion, etc...)?
The vented box tuning frequency minimum and zero phase point are often different from each other. Inductance is the culprit.
Yes, but inductance, eddy currents, motional feedback, mechanical non-linearities, and other factors aren't culprits, they are realities of real drivers.
The problem is the "small signal" model that may not take them into account or into account well. In turn, even if you could model all the factors nicely, you may never be able to get measurements for drivers so as to calculate these additional model elements. At least the more dynamic approach of sumaudioguy (as I poorly understand it) sounds better to my researchy ears as the way to measure even when we think we are talking about 'blocked" values.
Even with (marginally) better models, there still remains the question of goals. That is, what results do we want from a speaker and what kind of housing best struggles towards those results with shaking-cardboard drivers? That is not an engineering issue no matter how often I hear the deceptive phrase, "critical damping" or the euphemism "group delay".
Ben
Last edited:
What is the real world problem you want to address, Ben?
Sims match measurements if you accurately build what you simulated. If that is the case then what's the problem? There are examples everywhere of sims and measurements that overlay very well.
It isn't until high power levels that the sims start to deviate from the measurements and that's expected and we even know how to predict how the sims and measurements will be different at these power levels.
The only drivers that don't sim well with the traditional model are the ultra high excursion drivers and we even have a workaround for that now too.
So what is the actual problem? And what is the issue with the terms "critical damping" and "group delay"?
Sims match measurements if you accurately build what you simulated. If that is the case then what's the problem? There are examples everywhere of sims and measurements that overlay very well.
It isn't until high power levels that the sims start to deviate from the measurements and that's expected and we even know how to predict how the sims and measurements will be different at these power levels.
The only drivers that don't sim well with the traditional model are the ultra high excursion drivers and we even have a workaround for that now too.
So what is the actual problem? And what is the issue with the terms "critical damping" and "group delay"?
The effect of inductance on impedance phase at Fb is easily modeled. The effects of eddy currents on input impedance are easily modelled. Motional feedback has nothing to do with the topic at hand, but needs a model to tell the system what the feedback signal means. Nonlinearities can be modeled, but aren't as important as people might think for a box model specifically.
When making a model, you have to simplify it enough that it covers most, not all, of the expected behavior. If you can get to within less than a dB, for all intents and purposes, it is good enough for sizing a box/port whatever.
By "motional feedback" I mean "back EMF" which is motional feedback and which is, alas, one of the few kinds of corrective feedback available to shaking-cardboard drivers. (Have I previously mentioned my disdain for Rice-Kellogg drivers?*)
Perhaps you missed seeing the word "marginal" in my post. My point is that models are swell but there are more important issues to address than adding a decimal place to theoretical predictions... esp when those theoretical predictions are realized using a portable buzz saw in your back yard to produce carefully tuned Helmholtz resonator boxes.
Ben
*I also think internal combustion engines are a really, really stupid way to power a vehicle. But that hasn't stopped me from spending a lot of time with dynamometers tuning of my motorcycle engines over 55 yrs
Last edited:
back emf = motional feedback?
No.
Your disdain for direct radiators or ported boxes or everything since the AR-1 has nothing to do with interpreting impedance and just amounts to deflection and/or pensioner-style tangential thinking.
No.
Your disdain for direct radiators or ported boxes or everything since the AR-1 has nothing to do with interpreting impedance and just amounts to deflection and/or pensioner-style tangential thinking.
Watch your mouth sonny. Nobody has called me senile for hours.
By the way, that kind of personal insult isn't the norm here. There are lots of well-seasoned experts on this forum, not saying I'm one of them. How about an apology?
B.
You're doing it again: supposing that your personal, albeit anonymous authority is sufficient to post as a reply, turning it into he-said-she-said.
Degeneration is feedback, such as produced by low or negative output impedance in an amp acting on the back-EMF. Far better is active correction based on motional feedback.
Speakers* are the only element in modern HiFi systems without active error checking. Too bad.
Ben
*ESLs kind of excepted
Last edited:
Isn't that pretty much your entire modus operandi? A bunch of opinions, keywords, catchphrases and no technical information in sight. You keep complaining about "shaking cardboard" and resonances but you've shown no proof that there's an actual problem to address.
What type of active error checking do esl speakers employ?
Speakers don't really need any error checking. The distortion limitations are pretty well known. If you want lower distortion you need more or larger speakers. It's all pretty simple. There are moderately sized and priced speakers that could get you evicted easily while playing at negligible distortion levels. What more do you want?
Ron E- Am supposing you have some problem with your measurement technique because I have never seen the electrical resonance frequency equal the mechanical resonance frequency. I use this as one test for driver behavior. Indeed in certain tweeters this has been over 100Hz apart. Have you ever measured this actually or are you just blowing? Maybe you should use a scope and set up a Lissajous plot to verify your results as there is clearly some error in your method or equipment.
EMF occurs whenever a normal cone driver moves. Almost every amplifier output is connected to the inverting input which on better amps can react to this EMF. EMF occurs even when the speaker is being driven. It is fully possible to use a very simple feedback circuit to control driver motion via the back EMF. Some subwoofers in the past have used this "motional control" method in the design of the subwoofer.
The "who came first" card written about about Lee and Lampton is a bogus of a statement as seen here on DIY. All of the alignments are straight filter theory. The key point about the first paper by Lee and Lampton is the mechanical analysis instead of the electrical analog.
Believe it is time to give this thread over to the "experts" or Ron E and just a guy. They seem to know everything about everything woofer and build perfect systems while reading simulations. I often wonder if these guys even have an oscilloscope or any test equipment which cost more than lunch. No matter. My last post on this thread as the original questions have been pretty well answered I believe and hope the thread starter got his answer. The rest has turned in to a festival of simulation versus real world people with the simulators claiming their simulations are more valid than reality. A very common error based in the belief the ability to count outweighs actual data and observation. I get that kind of stupid throughout daily engineering work and have experienced far to much of that folly. Of course in real life power engineering the simulators people almost always end up firing themselves. Lack of reality or usefulness in the real business world. Computers do not make a good substitute for brains.
EMF occurs whenever a normal cone driver moves. Almost every amplifier output is connected to the inverting input which on better amps can react to this EMF. EMF occurs even when the speaker is being driven. It is fully possible to use a very simple feedback circuit to control driver motion via the back EMF. Some subwoofers in the past have used this "motional control" method in the design of the subwoofer.
The "who came first" card written about about Lee and Lampton is a bogus of a statement as seen here on DIY. All of the alignments are straight filter theory. The key point about the first paper by Lee and Lampton is the mechanical analysis instead of the electrical analog.
Believe it is time to give this thread over to the "experts" or Ron E and just a guy. They seem to know everything about everything woofer and build perfect systems while reading simulations. I often wonder if these guys even have an oscilloscope or any test equipment which cost more than lunch. No matter. My last post on this thread as the original questions have been pretty well answered I believe and hope the thread starter got his answer. The rest has turned in to a festival of simulation versus real world people with the simulators claiming their simulations are more valid than reality. A very common error based in the belief the ability to count outweighs actual data and observation. I get that kind of stupid throughout daily engineering work and have experienced far to much of that folly. Of course in real life power engineering the simulators people almost always end up firing themselves. Lack of reality or usefulness in the real business world. Computers do not make a good substitute for brains.
IIRC you've proposed only two arguments. The first was about ported box resonance being an issue. I countered this by asking what the modal ringing time in an average room was and stating that below 40 hz the resonant delay is mostly inconsequential and almost completely inconsequential when tuning below 20 hz. You had no answer to this at all.
The second argument concerns the traditional method of t/s measurement. Now you've expanded that to an assault on simulations. What is the point of measuring t/s parameters if you aren't going to use them? Whether you crunch the math and calculate an alignment or run it through a simulator the traditional method of measuring t/s is more than good enough.
No one has said that simulation is more valid than reality, this is just being over dramatic. What I actually said was the sims match measurements quite well IF you accurately simulate what you built. This holds true at low power levels and the deviations at high power levels are well known and easy to predict.
I've spent a lot of time learning how to simulate accurately. If you are not able to get sims with traditionally measured t/s to match measurements you are clearly not doing accurate simulations. There are literally hundreds (if not thousands) of examples of simulations matching measurements to a high degree of accuracy. This is basically what we do here. Even very complex high order designs are pretty simple to simulate accurately and match measurements.
"Simulators vs real world people", "that kind of stupid", "lack of reality or usefulness", "computers do not make a good substitute for brains"; this is all an extremely arrogant and lazy attack when YOU are the one that can't get sims to match measurements. Simulators are a great tool that enhance knowledge and theory, I'm not sure why you think people that use these tools are completely clueless about the theory.
People are using simulators every day to design enclosures and to add crossovers (active and/or passive) and getting simulated results that almost perfectly overlay measurements on and off axis. I'm not sure what simulators you have used (if any) and when (if ever) but modern sim tools are more than capable of providing results that very closely match measurements without any need to reinvent methods to measure t/s parameters. For the most part people that know what they are doing agree that simulators are a very powerful tool and it would be pretty silly not to use them.
I have measured using ARTA LIMP two unbranded 10 inch woofers and found that one has a higher impedance amplitude peak which is 30 Ohms at Fs 37 Hz , while the other one has a lower impedance amplitude 14.6 ohms at Fs 33 Hz . I would like to know what does that indicate. Tks for answer. ALDO
impedance amplitude explanation
Sorry I was perhaps not clear in my question. in fact I wanted to know in general whether a high free air amplitude at fs is an indication of a better driver acoustically and technically.
Sorry I was perhaps not clear in my question. in fact I wanted to know in general whether a high free air amplitude at fs is an indication of a better driver acoustically and technically.
No. You can imply little wrt that peak. Where it is tells you the Fs, and the height/width are related to the Q.
If there are ripples as you go up in frequency those are often related to resonances and could give some indication of problem areas.
dave
If there are ripples as you go up in frequency those are often related to resonances and could give some indication of problem areas.
dave
- Status
- Not open for further replies.