Hello All,
Just for fun Today I took some blue tack stuff and stuck a stuck a SB26ADC-C000-4 tweeter on a VISATON WG220 x 150 wave guide.
Using a calibrated microphone, AP analyzer, and APx1701 test amplifier I gave the driver/waveguide a spin on the bench. The output of the amplifier was 2.83 volts. The microphone was 9 inches or so from the driver.
See the attached plots. I liked the low distortion levels.
Thanks DT
Just for fun Today I took some blue tack stuff and stuck a stuck a SB26ADC-C000-4 tweeter on a VISATON WG220 x 150 wave guide.
Using a calibrated microphone, AP analyzer, and APx1701 test amplifier I gave the driver/waveguide a spin on the bench. The output of the amplifier was 2.83 volts. The microphone was 9 inches or so from the driver.
See the attached plots. I liked the low distortion levels.
Thanks DT
Great. Thanks for the effort.
It shows nice wave guide 'gain' below 10KHz down to around 3KHz, this is what I am looking for. This 'gain' can be traded for lower distortion.
The very slight peak just under 3Khz should not be a problem as that is the crossover frequency I will be -6dB down. Look at 15° off axis, reference it to 10KHz and being down -6dB will mean response be pulled down by close to 10-12dB. I will also have a null LC filter tuned near the Fs and these two 'tricks' and based on what I see above, should mean very low distortion indeed.
Using more specific speaker hardware/sofware, I should be able to supply a fair bit more detail again. I use a Clio 11 setup, but I now also have a QuantAsylumQA403 (near unobtanium with chip shortages) that can produce 32bit stimuli and tried a signal into the Clio and that seemed to work well. So looking forward to seeing how that can be used.
I hope to be able to post the results with the Elsinore waveguide soon, all it requires is an O'ring and very slightly increase diameter of mounting holes; it looks like a good fit, but we shall see how that works out in practice.
It shows nice wave guide 'gain' below 10KHz down to around 3KHz, this is what I am looking for. This 'gain' can be traded for lower distortion.
The very slight peak just under 3Khz should not be a problem as that is the crossover frequency I will be -6dB down. Look at 15° off axis, reference it to 10KHz and being down -6dB will mean response be pulled down by close to 10-12dB. I will also have a null LC filter tuned near the Fs and these two 'tricks' and based on what I see above, should mean very low distortion indeed.
Using more specific speaker hardware/sofware, I should be able to supply a fair bit more detail again. I use a Clio 11 setup, but I now also have a QuantAsylumQA403 (near unobtanium with chip shortages) that can produce 32bit stimuli and tried a signal into the Clio and that seemed to work well. So looking forward to seeing how that can be used.
I hope to be able to post the results with the Elsinore waveguide soon, all it requires is an O'ring and very slightly increase diameter of mounting holes; it looks like a good fit, but we shall see how that works out in practice.
Can somebody explain this (from “Loudspeakers for Music Recording and Reproduction” by Newell and Holland):
In order to behave as a voltage source, and remain independent of load, the output impedance of the amplifier needs to be very low indeed – typically hundredths of an ohm.
I can’t get my head around the idea of a voltage source being independent of load, and why low output impedance would accomplish that.
I hope this isn’t off topic - seems like it’s related to Joe’s concern with current vs voltage.
In order to behave as a voltage source, and remain independent of load, the output impedance of the amplifier needs to be very low indeed – typically hundredths of an ohm.
I can’t get my head around the idea of a voltage source being independent of load, and why low output impedance would accomplish that.
I hope this isn’t off topic - seems like it’s related to Joe’s concern with current vs voltage.
The Voltage signal that the amplifier produces at it's output appears across the series combination of the amplifier output impedance and the speaker. By way of ordinary Voltage division, some of the Voltage will appear across each of the two elements. When the amplifier output impedance is non-zero, it will drop some of the Voltage and the Voltage of the speaker will no longer be the same as the amplifier output Voltage. It will also vary with frequency if both the amplifier output impedance is non-zero and the speaker impedance varies with frequency. In theory a source will be significantly a Voltage source when the source resistance is 0.8 ohms or less.
Hi Guys
Is this off topic? Well, it is something that is at the heart of the Elsinore design approach. I get reluctant, based on past experience, to talk about what some may deem 'disputed' facts, even when there really is no need to label them such. So I don't want an extensive discussion here on the subject and if necessary I will just get of the bus (this bus) and go to another bus, so-to-speak.
But here goes, briefly as it is a BIG subject:
All amplifiers can be categorised by their output impedance as a voltage source of zero Ohm in series with an impedance, which can be modeled as a resistor, but it really is not a resistor since not all impedances are resistors (anything that impedes current is an impedance and often that is not a resistor). But we can use a resistor to model it.
Note the speaker. This represents the load. Let us say if that load is a typical 8 Ohm, then whether we have voltage-drive or current-drive is basically down to that output impedance. It is the ratio that matters. What is that ratio? Is it 100:1 or does it need to be infinity?
About ten years ago I brought this ratio matter up with Esa Merilainen who we all know wrote this book:
Indeed I was among the first to buy the book because I instantly recognised it as important, and still do, even if there are a number of conclusions where I disagree with him.
So back to this ratio...
1. If our output impedance is zero, then we have a pure voltage source. The ratio with the load of 8 Ohm would be infinite. Infinitely low.
2. If our output impedance is infinite, then we have a pure current source. The ratio with the load would also be infinite, infinitely high.
Neither of these are attainable in real life, but is is clear that the ratio should either be infinitely low or infinitely high, that is not possible. This is about what is achievable in the real world. What would be the minimum ratio we can get away with, and this was the question I raised with Esa Merilainen, after all, he wrote the book and it seemed such an obvious question.
From here I draw from convention: A voltage source interface is when the insertion loss is no more than 1dB, that the voltage should not sag more that 1dB max under load. A quick mental calculation will tell you that is around a ratio of 10:1 - and if that is good for a current source too, then 10:1 should be applied current sources as well. The ratio 10:1 should apply to both, at least in my mind.
So we go back...
1. Voltage: With an 8 Ohm load and the insertion load to be no more than 1dB, the output impedance must not be higher than 8R/10 = 0.8 Ohm.
2. Current: With an 8 Ohm load and using the same 10:1 convention, the output impedance must at least 8R*10 = 80 Ohm.
That means with voltage-drive our output impedance should be nore more than 1/10th of 8 Ohm = 0.8 Ohm.
That also means with current-drive our output impedance should be at least ten times 8 Ohm = 80 Ohm.
But...
I asked Esa on the Vendor's thread, what would he consider the minimum ratio to achieve acceptable current-drive? I had a figure in mind when asking the question and it was 5:1, and I was delighted when he came back with 5:1. But I still maintain 10:1 is highly desirable, if you can get it.
So 5:1 is the working ratio and they work with both current sources and voltage sources.
Hence now we have 1.6 Ohm output impedance for voltage source drive. Some may say that is a damping factor of 5, but quite frankly, in the overall scheme of things, damping is a misnomer and best to largely ignore it. But this should keep tube amplifier users happy since they generally have higher output impedances.
We also now have a 40 Ohm output impedance required for current source drive with 8 Ohm speakers. That of course would be 10:1 if the speaker was 4 Ohm. But you get the general picture, right?
But...
And it is a BIG but...
What if we design the load so that the amplifier has to produce the same current with all frequencies? This would make the 5:1 ratio constant with frequency. That would cancel out the output impedance of the amplifier. As long as the ratio is constant with varying frequency, the ratio can be anything from zero to infinity.
We can cancel out the output impedance of whatever amplifier we are using.
This is what the Elsinore does, from around 30 Hertz and up. But it goes a lot deeper than that. Indeed the above is just a starting point.
Practicality: From the standpoint of Elsinore users, it is this, that they don't need to worry about the amplifier at all, that they get the same performance. At least in terms of the frequency staying stable.
Does it give you a sonic improvement advantage. I say absolutely yes, because the driver constantly wants to modulate the current of the amplifier (an often ignored fact) and this is suppressed. Current drive naturally suppress the modulation of current because it already supplies current that is constant with frequency and hence the non-linear part of the impedance of the speaker cannot modulate the current of the amplifier. This reduces distortion and whatever the sound that Esa says he is getting with current drive (because current drive suppresses the current modulation of the amplifier), we can get current source like improvements with all kinds of amplifier, including voltage sources. You do not need current drive! This is where Esa and I differ.
All of this I am in the process of documenting right now. And it won't be published here, it 's the wrong place.
But for the end users of the Elsinores, this is your thread and don't worry too much about all these niceties, just enjoy the sound know, that at least there is some serious science behind it.
Cheers, Joe
PS: An interesting fact hardly ever mentioned, but if a speaker already had a flat frequency response when under current drive, then we see that at low frequencies we see the cone visibly move, but not so at higher frequencies. But, the response is still flat when both the current is flat and the acceleration is flat. So the current is flat and the driver is able to track that current with flat acceleration, then you would have the perfect speaker that in theory would produce no distortion at the interface and a flat response... just a thought.
Is this off topic? Well, it is something that is at the heart of the Elsinore design approach. I get reluctant, based on past experience, to talk about what some may deem 'disputed' facts, even when there really is no need to label them such. So I don't want an extensive discussion here on the subject and if necessary I will just get of the bus (this bus) and go to another bus, so-to-speak.
But here goes, briefly as it is a BIG subject:
All amplifiers can be categorised by their output impedance as a voltage source of zero Ohm in series with an impedance, which can be modeled as a resistor, but it really is not a resistor since not all impedances are resistors (anything that impedes current is an impedance and often that is not a resistor). But we can use a resistor to model it.
Note the speaker. This represents the load. Let us say if that load is a typical 8 Ohm, then whether we have voltage-drive or current-drive is basically down to that output impedance. It is the ratio that matters. What is that ratio? Is it 100:1 or does it need to be infinity?
About ten years ago I brought this ratio matter up with Esa Merilainen who we all know wrote this book:
Indeed I was among the first to buy the book because I instantly recognised it as important, and still do, even if there are a number of conclusions where I disagree with him.
So back to this ratio...
1. If our output impedance is zero, then we have a pure voltage source. The ratio with the load of 8 Ohm would be infinite. Infinitely low.
2. If our output impedance is infinite, then we have a pure current source. The ratio with the load would also be infinite, infinitely high.
Neither of these are attainable in real life, but is is clear that the ratio should either be infinitely low or infinitely high, that is not possible. This is about what is achievable in the real world. What would be the minimum ratio we can get away with, and this was the question I raised with Esa Merilainen, after all, he wrote the book and it seemed such an obvious question.
From here I draw from convention: A voltage source interface is when the insertion loss is no more than 1dB, that the voltage should not sag more that 1dB max under load. A quick mental calculation will tell you that is around a ratio of 10:1 - and if that is good for a current source too, then 10:1 should be applied current sources as well. The ratio 10:1 should apply to both, at least in my mind.
So we go back...
1. Voltage: With an 8 Ohm load and the insertion load to be no more than 1dB, the output impedance must not be higher than 8R/10 = 0.8 Ohm.
2. Current: With an 8 Ohm load and using the same 10:1 convention, the output impedance must at least 8R*10 = 80 Ohm.
That means with voltage-drive our output impedance should be nore more than 1/10th of 8 Ohm = 0.8 Ohm.
That also means with current-drive our output impedance should be at least ten times 8 Ohm = 80 Ohm.
But...
I asked Esa on the Vendor's thread, what would he consider the minimum ratio to achieve acceptable current-drive? I had a figure in mind when asking the question and it was 5:1, and I was delighted when he came back with 5:1. But I still maintain 10:1 is highly desirable, if you can get it.
So 5:1 is the working ratio and they work with both current sources and voltage sources.
Hence now we have 1.6 Ohm output impedance for voltage source drive. Some may say that is a damping factor of 5, but quite frankly, in the overall scheme of things, damping is a misnomer and best to largely ignore it. But this should keep tube amplifier users happy since they generally have higher output impedances.
We also now have a 40 Ohm output impedance required for current source drive with 8 Ohm speakers. That of course would be 10:1 if the speaker was 4 Ohm. But you get the general picture, right?
But...
And it is a BIG but...
What if we design the load so that the amplifier has to produce the same current with all frequencies? This would make the 5:1 ratio constant with frequency. That would cancel out the output impedance of the amplifier. As long as the ratio is constant with varying frequency, the ratio can be anything from zero to infinity.
We can cancel out the output impedance of whatever amplifier we are using.
This is what the Elsinore does, from around 30 Hertz and up. But it goes a lot deeper than that. Indeed the above is just a starting point.
Practicality: From the standpoint of Elsinore users, it is this, that they don't need to worry about the amplifier at all, that they get the same performance. At least in terms of the frequency staying stable.
Does it give you a sonic improvement advantage. I say absolutely yes, because the driver constantly wants to modulate the current of the amplifier (an often ignored fact) and this is suppressed. Current drive naturally suppress the modulation of current because it already supplies current that is constant with frequency and hence the non-linear part of the impedance of the speaker cannot modulate the current of the amplifier. This reduces distortion and whatever the sound that Esa says he is getting with current drive (because current drive suppresses the current modulation of the amplifier), we can get current source like improvements with all kinds of amplifier, including voltage sources. You do not need current drive! This is where Esa and I differ.
All of this I am in the process of documenting right now. And it won't be published here, it 's the wrong place.
But for the end users of the Elsinores, this is your thread and don't worry too much about all these niceties, just enjoy the sound know, that at least there is some serious science behind it.
Cheers, Joe
PS: An interesting fact hardly ever mentioned, but if a speaker already had a flat frequency response when under current drive, then we see that at low frequencies we see the cone visibly move, but not so at higher frequencies. But, the response is still flat when both the current is flat and the acceleration is flat. So the current is flat and the driver is able to track that current with flat acceleration, then you would have the perfect speaker that in theory would produce no distortion at the interface and a flat response... just a thought.
We are still waiting for your back office science team to present the report that you announced a couple of years ago. Is it emerging?
//
Hi Alan
Don't worry about them. They don't even have the courage to identify themselves with actual names, when in fact I use a real name that anybody can Google. If they can't even give me their name, then why should I even care what they say.
Ah, what is a current source?
Yes, this is a commonly enough brought up and it does sound counter-intuitive, how can any current flow when you have an infinite value resistor on the output. The thing to understand is that this is not a resistor, it is an impedance. That is why I said at the very beginning that you can use a resistor to model, but it isn't exactly a resistor.
When you start looking at current sources, then there are a lot of things that are not intuitive. Voltage is very intuitive, but discussion re current is much less so.
Think of it this way, if a current source is set up to deliver 1 Amp of current, then for the current not to change with impedance and frequency, the output impedance should theoretically infinitely high. But of course, not in real life.
Let me explain it in a different way that might help.
Below is how I understand how Neville Thiele (of Thiele-Small Parameters fame) did the pioneering way of measuring the impedance of a driver using a current source.
Set up a voltage source to 10V and use a 1000R resistor to give you a current source effect. Since he used 1000 Ohm and that meant shorting the output, then 10mA would flow. Now insert the driver. So this is pure Ohm's Law. We have set up 10mA current source.
Since the impedance of the driver is much lower than 1000 Ohm, then the 10mA current hardly varies with changes in impedance of the driver. So far so good.
Using this setup, Thiele was able to measure the impedance of the driver by simply measuring the voltage across the driver terminals. Again pure Ohm'Law.
By measuring the voltage across the driver, he would get 60mV if the driver is 6 Ohm impedance. If the impedance of the driver went up 12 Ohm, he would then measure 120mV with reasonable accuracy, because the current has hardly changed. But of course it has changed slightly, we can do the maths.
With 6 Ohm the amplifier is delivering 10/1006 = 9,940mA
When 12 Ohm, the amplifier is delivering 10/1012 = 9.988mA
So the current hardly changed and that means the accuracy is within 1% or so.
But if the source impedance had been a theoretical infinite, then the 10mA would not change or reduce not matter what and we would have had 100% accuracy. That is why the theoretical impedance of a current source would be infinitely high.
Hope that helped. It must have been fourty years ago when I learnt this from Neville Thiele' example.
The above should hopefully not find any objection with the buffins, but you never know.
So here is an added tidbit that they might holler about:
If the DC resistor was 6R and the impedance was 11 Ohm at 2KHz (an actual driver, Vifa P17WJ-08), then if the current source was set to 1 Amp, then the 11 Ohm impedance would gives us 11V across the driver terminals. That means we have got an extra 5V over the 6V, hence 11V. Now for the punchline, that extra 5V = 5 Ohm. That extra 5V is a voltage source. When we are using a 1 A current source, 1V equals 1 Ohm.
Hence we have a voltage source that is both 5V and 5 Ohm at the same time. I have mentioned this before and there are some who don't like it. Well, it is easy enough to set up and test, no reason to be lazy about it. And it conforms 100% with Ohm's Law.
But what about if we are connecting the same Vifa driver to a voltage source amplifier?
If 1 Amp is flowing (adjust the voltage of the amplifier to achieve 1 Amp), that internal voltage source is still there and acting as 5 Ohm and hence reducing the current when it is a voltage source driving the speaker. The internal voltage source acts as an impedance under voltage-drive.
Now they will probably call that snake oil too. That just means they don't understand it, or just not trying.
It takes a long time go wrap one's mind around this stuff. If you are mentally lazy, then of course it means nothing. But Alan, trust me on this, I talk to some very knowledgeable people for hours at a time. They know that this ain't snake oil, but rather I have gotten only encouragement. Keep going. Also, it is important not to isolate oneself too much. I have a friend called Morris, a straight shooter and a crusty character, and if I go of the rails he will quickly toe me back in line. The only difference is that he is not a troll. And Morris is not the only one. I won't mention names, or they will accuse me of name-dropping. So I can't win... except time is on my side.
So there you are.
Cheers, Joe
Don't worry about them. They don't even have the courage to identify themselves with actual names, when in fact I use a real name that anybody can Google. If they can't even give me their name, then why should I even care what they say.
Ah, what is a current source?
Yes, this is a commonly enough brought up and it does sound counter-intuitive, how can any current flow when you have an infinite value resistor on the output. The thing to understand is that this is not a resistor, it is an impedance. That is why I said at the very beginning that you can use a resistor to model, but it isn't exactly a resistor.
When you start looking at current sources, then there are a lot of things that are not intuitive. Voltage is very intuitive, but discussion re current is much less so.
Think of it this way, if a current source is set up to deliver 1 Amp of current, then for the current not to change with impedance and frequency, the output impedance should theoretically infinitely high. But of course, not in real life.
Let me explain it in a different way that might help.
Below is how I understand how Neville Thiele (of Thiele-Small Parameters fame) did the pioneering way of measuring the impedance of a driver using a current source.
Set up a voltage source to 10V and use a 1000R resistor to give you a current source effect. Since he used 1000 Ohm and that meant shorting the output, then 10mA would flow. Now insert the driver. So this is pure Ohm's Law. We have set up 10mA current source.
Since the impedance of the driver is much lower than 1000 Ohm, then the 10mA current hardly varies with changes in impedance of the driver. So far so good.
Using this setup, Thiele was able to measure the impedance of the driver by simply measuring the voltage across the driver terminals. Again pure Ohm'Law.
By measuring the voltage across the driver, he would get 60mV if the driver is 6 Ohm impedance. If the impedance of the driver went up 12 Ohm, he would then measure 120mV with reasonable accuracy, because the current has hardly changed. But of course it has changed slightly, we can do the maths.
With 6 Ohm the amplifier is delivering 10/1006 = 9,940mA
When 12 Ohm, the amplifier is delivering 10/1012 = 9.988mA
So the current hardly changed and that means the accuracy is within 1% or so.
But if the source impedance had been a theoretical infinite, then the 10mA would not change or reduce not matter what and we would have had 100% accuracy. That is why the theoretical impedance of a current source would be infinitely high.
Hope that helped. It must have been fourty years ago when I learnt this from Neville Thiele' example.
The above should hopefully not find any objection with the buffins, but you never know.
So here is an added tidbit that they might holler about:
If the DC resistor was 6R and the impedance was 11 Ohm at 2KHz (an actual driver, Vifa P17WJ-08), then if the current source was set to 1 Amp, then the 11 Ohm impedance would gives us 11V across the driver terminals. That means we have got an extra 5V over the 6V, hence 11V. Now for the punchline, that extra 5V = 5 Ohm. That extra 5V is a voltage source. When we are using a 1 A current source, 1V equals 1 Ohm.
Hence we have a voltage source that is both 5V and 5 Ohm at the same time. I have mentioned this before and there are some who don't like it. Well, it is easy enough to set up and test, no reason to be lazy about it. And it conforms 100% with Ohm's Law.
But what about if we are connecting the same Vifa driver to a voltage source amplifier?
If 1 Amp is flowing (adjust the voltage of the amplifier to achieve 1 Amp), that internal voltage source is still there and acting as 5 Ohm and hence reducing the current when it is a voltage source driving the speaker. The internal voltage source acts as an impedance under voltage-drive.
Now they will probably call that snake oil too. That just means they don't understand it, or just not trying.
It takes a long time go wrap one's mind around this stuff. If you are mentally lazy, then of course it means nothing. But Alan, trust me on this, I talk to some very knowledgeable people for hours at a time. They know that this ain't snake oil, but rather I have gotten only encouragement. Keep going. Also, it is important not to isolate oneself too much. I have a friend called Morris, a straight shooter and a crusty character, and if I go of the rails he will quickly toe me back in line. The only difference is that he is not a troll. And Morris is not the only one. I won't mention names, or they will accuse me of name-dropping. So I can't win... except time is on my side.
So there you are.
Cheers, Joe
There's a thread here where these impolitic ideas can be discussed:
https://www.diyaudio.com/community/threads/current-drive-for-loudspeakers.250272/
https://www.diyaudio.com/community/threads/current-drive-for-loudspeakers.250272/
Hi Pete, thanks. They say "discuss in a friendly way" and I am with the Moderator to make it so. But that was back in 2014 when it started. Last post was back in 2021. I have noted that these discussions generally petter out and don't seem to get anywhere. The same old links are pointed at and then not much more. This seems to have ended up the same.
I am often thought of being into current drive, Actually I am not.
The world is not going to be converted to current drive. This is where Esa Merilainen and I differ. I think some things about current drive has been misunderstood. I am more interesting in this from a much wider perspective: All amplifiers are current-delivery-systems, whether they are voltage sources or current sources. They just deliver current in different ways. And once we understand that, and follow what the current does under both circumstances, only then will be making progress.
The question is not about current being constant a la current drive. This is about controlling the current, and that this is needed for both voltage drive and current drive. I repeat, current control needs to be exerted whether you use voltage drive or current drive. And it's about finding ways of doing that.
I have had to get my hands on equipment, but I now have it, one that is a bit rare, a 6.5" driver with very specific characteristics, also a test microphone of the highest quality and it costs money. Tests for distortion using current drive has not used really high quality test microphone, Esa used an expensive but rugged dynamic microphone in his later tests, the Sennheiser E815S >$100. Another used a cheap $50 Behringer ECM 8000 microphone. I will be using an Earthworks M30 mic. Measurements will be both electrical and acoustical, multi-tones as IMD shows up issues the best.
What's the saying? "I love it when a plan comes together."
Eight out of ten loves it when a plan come together:
I am often thought of being into current drive, Actually I am not.
The world is not going to be converted to current drive. This is where Esa Merilainen and I differ. I think some things about current drive has been misunderstood. I am more interesting in this from a much wider perspective: All amplifiers are current-delivery-systems, whether they are voltage sources or current sources. They just deliver current in different ways. And once we understand that, and follow what the current does under both circumstances, only then will be making progress.
The question is not about current being constant a la current drive. This is about controlling the current, and that this is needed for both voltage drive and current drive. I repeat, current control needs to be exerted whether you use voltage drive or current drive. And it's about finding ways of doing that.
I have had to get my hands on equipment, but I now have it, one that is a bit rare, a 6.5" driver with very specific characteristics, also a test microphone of the highest quality and it costs money. Tests for distortion using current drive has not used really high quality test microphone, Esa used an expensive but rugged dynamic microphone in his later tests, the Sennheiser E815S >$100. Another used a cheap $50 Behringer ECM 8000 microphone. I will be using an Earthworks M30 mic. Measurements will be both electrical and acoustical, multi-tones as IMD shows up issues the best.
What's the saying? "I love it when a plan comes together."
Eight out of ten loves it when a plan come together:
It was a sincere question - Joe, you said at one point that you where to publish a technical report on the subject and that you had summed a team to write it as it seems you felt you could now convey the concept intricate in full. Is this work canceled? I think many actually where seriously interested in that paper.
//
//
Yes, no,, yes, no and yes. Sorry. It's complicated.
But thank you, the moderation is tone is very much appreciated.
This is a much bigger undertaking that most people think. Trying to break through an obstinate thinking that has been set in cement for decades, is very difficult. So an approach had to be figured out that would make for a convincing story in the sense that the narrative is so focused that it can't be taken the wrong way round. For example, Pavel (PMA) describes this as speaker distortion:
He refers several times to "speaker current" and hence he has drawn the wrong conclusion. I tried to approach him and he only reacted harshly. Yet the speaker cannot produce current, only the amplifier can. Esa to an extent seems to have made a similar conclusion, but I hope he will be amiable to see it differently in the future. Only the amplifier can produce current. When the amplifier is a voltage source, the current produced it on demand by the load. But when the amplifier is a current source, the current is not on demand by the speaker load. Once I realised that, I also realised that the rules of the ball game was quite different to what most people think. That got me thinking in a very different way. But not wanting to trick myself me up, I am in the fortunate position to be able to consult with some very capable persons. I only well know the pitfalls of bamboozling oneself and lose touch with reality. But what I got was very encouraging instead, and told to keep going. But anything to be documented by me, this would then be thoroughly vetted by very the right kind of people. And then I promise, you will get to know about and I would be very happy for you to read it. But can you understand why I am very guarded and feel that this has to come out in a very calculated way. And it has proven to be a marathon and not a sprint.
BTW, please excuse me, I am a curious person by nature. I was born in Frederiksberg (near the original Carlsberg brewery), Copenhagen, and I see that you are from 'Sverige' and wondering from what part? I do of course have a number of friends there. My own brother Jannik moved to Norge in the late 70's and is now a Norge full citizen along with his two grownup children, both engineers, and he lives in Kongsvinger. His background is computers.
Cheers, Joe
But thank you, the moderation is tone is very much appreciated.
This is a much bigger undertaking that most people think. Trying to break through an obstinate thinking that has been set in cement for decades, is very difficult. So an approach had to be figured out that would make for a convincing story in the sense that the narrative is so focused that it can't be taken the wrong way round. For example, Pavel (PMA) describes this as speaker distortion:
He refers several times to "speaker current" and hence he has drawn the wrong conclusion. I tried to approach him and he only reacted harshly. Yet the speaker cannot produce current, only the amplifier can. Esa to an extent seems to have made a similar conclusion, but I hope he will be amiable to see it differently in the future. Only the amplifier can produce current. When the amplifier is a voltage source, the current produced it on demand by the load. But when the amplifier is a current source, the current is not on demand by the speaker load. Once I realised that, I also realised that the rules of the ball game was quite different to what most people think. That got me thinking in a very different way. But not wanting to trick myself me up, I am in the fortunate position to be able to consult with some very capable persons. I only well know the pitfalls of bamboozling oneself and lose touch with reality. But what I got was very encouraging instead, and told to keep going. But anything to be documented by me, this would then be thoroughly vetted by very the right kind of people. And then I promise, you will get to know about and I would be very happy for you to read it. But can you understand why I am very guarded and feel that this has to come out in a very calculated way. And it has proven to be a marathon and not a sprint.
BTW, please excuse me, I am a curious person by nature. I was born in Frederiksberg (near the original Carlsberg brewery), Copenhagen, and I see that you are from 'Sverige' and wondering from what part? I do of course have a number of friends there. My own brother Jannik moved to Norge in the late 70's and is now a Norge full citizen along with his two grownup children, both engineers, and he lives in Kongsvinger. His background is computers.
Cheers, Joe
Thanks, Joe. I had to read it a few times, but I think I got it.
The concept of "damping factor" seems to be that, by "shorting" the speaker terminals, the current created by driver motion is resisted by a back-EMF (not sure if I got that right) and resonance is "controlled", at least at some frequencies near the driver's natural resonance frequency.
If I got that right, then how does current drive perform the same function?
Last edited:
That's the common view. But this idea that there is some kind of brake going on is something I greatly disagree upon and the physics don't support it. Again I might get into trouble saying this, but motion of the cone is directly a function of current, yet here is a startling reality, that if the amplifier is a voltage source, it 'regulates' or fixes the voltage value at the expense of the current. In other words, if you have a voltage source you relinquish control over the current, and conversely, when you have a current source, you now 'regulate' or fixes the current (value) at the expense of controlling the voltage. Damping or braking can only be done when controlling the current. That is not possible with a voltage source.
So somehow we have managed (all of us) to be looking this the wrong way. A better explanation that fits the physics needs to be found. Damping is not controlled by voltage (only current makes sound, so the current has to be controlled), but it can be seen to be controlled as current versus frequency. Examining that seems to lead to a much more satisfying explanation. It also leads to the solution used in the Elsinores, which can be driven by any amplifier and yet has proper damping even if the amplifier is a theoretical infinite output impedance.
There are in fact two areas of concern, the box alignment which is a mechanical high-pass filter. The crossovers is the other area. They also have a high-pass filter on the tweeter. So we have one mechanical high-pass and we have another electrical high-pass. Both of these are very sensitive to the source impedance of the amplifier. In fact, for them to work as designed, we have to assume the amplifier is a voltage source and very low output impedance.
To make the speaker immune to the source impedance of the amplifier, we need to concern us with both the bass alignment and the crossover.
There are in the Elsinores three filters, one mechanical and two electrical.
Now let's look at the Elsinores and how sensitive they are to two source impedances 0.1 Ohm and 270 Ohm, the first is a voltage source and the other is a current source.
Is that not pretty cool? Both the box alignment is controlled and the deviation around 20 Herts is almost 20dB down (1/100th power). This is because this is a ported alignment. If this was a sealed or aperiodic alignment (high-pass), then that deviation would not be there, only with ported designs.
But look at that crossover!!!
We have achieved very nice control there as well. BTW, I have the amplifier with 270 Ohm output impedance right here and it is documented elsewhere on diyaudio.com, click here:
"Trans-Amp" - 40 Watt Transconductance "Current Amplifier"
On the Elsinores, with my current drive amplifier and 270 Ohm output impedance, I can play incredibly loud drum solos and deep organ tracks, and there is no loss of control.
KEY: Control the current and you can have all the damping to want with any amplifier.
It is the current that is the key to damping. We must find a way to control the current versus frequency and get away with being fixated with the voltage (not ignoring it of course) and we can start to find real world solutions. The Elsinores are compatible with both voltage drive and current drive. And the 'damping' does not change.
But when I start talking to people first time about current, they automatically think of me as somebody who is in the current-drive camp. The truth is that I am not in any camp. If anything I have doubts that current-drive is a practical solution (but it works great in active speakers).
Cheers, Joe
So somehow we have managed (all of us) to be looking this the wrong way. A better explanation that fits the physics needs to be found. Damping is not controlled by voltage (only current makes sound, so the current has to be controlled), but it can be seen to be controlled as current versus frequency. Examining that seems to lead to a much more satisfying explanation. It also leads to the solution used in the Elsinores, which can be driven by any amplifier and yet has proper damping even if the amplifier is a theoretical infinite output impedance.
There are in fact two areas of concern, the box alignment which is a mechanical high-pass filter. The crossovers is the other area. They also have a high-pass filter on the tweeter. So we have one mechanical high-pass and we have another electrical high-pass. Both of these are very sensitive to the source impedance of the amplifier. In fact, for them to work as designed, we have to assume the amplifier is a voltage source and very low output impedance.
To make the speaker immune to the source impedance of the amplifier, we need to concern us with both the bass alignment and the crossover.
There are in the Elsinores three filters, one mechanical and two electrical.
Now let's look at the Elsinores and how sensitive they are to two source impedances 0.1 Ohm and 270 Ohm, the first is a voltage source and the other is a current source.
Is that not pretty cool? Both the box alignment is controlled and the deviation around 20 Herts is almost 20dB down (1/100th power). This is because this is a ported alignment. If this was a sealed or aperiodic alignment (high-pass), then that deviation would not be there, only with ported designs.
But look at that crossover!!!
We have achieved very nice control there as well. BTW, I have the amplifier with 270 Ohm output impedance right here and it is documented elsewhere on diyaudio.com, click here:
"Trans-Amp" - 40 Watt Transconductance "Current Amplifier"
On the Elsinores, with my current drive amplifier and 270 Ohm output impedance, I can play incredibly loud drum solos and deep organ tracks, and there is no loss of control.
KEY: Control the current and you can have all the damping to want with any amplifier.
It is the current that is the key to damping. We must find a way to control the current versus frequency and get away with being fixated with the voltage (not ignoring it of course) and we can start to find real world solutions. The Elsinores are compatible with both voltage drive and current drive. And the 'damping' does not change.
But when I start talking to people first time about current, they automatically think of me as somebody who is in the current-drive camp. The truth is that I am not in any camp. If anything I have doubts that current-drive is a practical solution (but it works great in active speakers).
Cheers, Joe
Last edited: