Well I managed to get through most of this. Great thread. I have long been interested in current drive. My current system has 5 1/4" Radio Shack Foster OEM FR and 10" high qts woofer both on open baffle with a 12" sealed subwoofer all tri-amped with sand amps. I think that in particular the FR could benefit from current drive.
I intend to go hollow state (with possible exception of the sub). The design I had planned for the FR was UL pentode with local FB on the output stage. It seems that a pentode mode without feedback similar to the older German system mentioned earlier might be worth some investigation as well.
Since the OB will not raise the Qm it seems this speaker system is an ideal test bed. I don't expect to get to it very soon as I have other projects but it seems to me that there are a couple of issues that make the zero NFB pentode only a partial solution.
The first is the non-linear gm of the pentode that we normally use NFB to address. The second is that even an open loop pentode with lowish turns ratio on the OPT still is not a full fledged current drive.
To address these I suppose that a NFB solution which uses a current sensing as the source of the feedback signal with the correction applied in a way that increases rather than decreases the Zout would be in order. So I will need to do a little bit of study on NFB theory to see what such a system might look like. Since power requirement is modest class A operation (even SE) is fine so cathode bias without bypass cap seems to be the starting point.
I intend to go hollow state (with possible exception of the sub). The design I had planned for the FR was UL pentode with local FB on the output stage. It seems that a pentode mode without feedback similar to the older German system mentioned earlier might be worth some investigation as well.
Since the OB will not raise the Qm it seems this speaker system is an ideal test bed. I don't expect to get to it very soon as I have other projects but it seems to me that there are a couple of issues that make the zero NFB pentode only a partial solution.
The first is the non-linear gm of the pentode that we normally use NFB to address. The second is that even an open loop pentode with lowish turns ratio on the OPT still is not a full fledged current drive.
To address these I suppose that a NFB solution which uses a current sensing as the source of the feedback signal with the correction applied in a way that increases rather than decreases the Zout would be in order. So I will need to do a little bit of study on NFB theory to see what such a system might look like. Since power requirement is modest class A operation (even SE) is fine so cathode bias without bypass cap seems to be the starting point.
There has now appeared a couple of customer reviews of the book at the Amazon.com page. One is positive but the other merely prejudiced. My comment to the latter can be read here.
I encourage here those who have the book and have grasped the issue to some extent to also post reviews at the Amazons or elsewhere, especially if you consider the book to be worth more than the present rating. (Amazon allows posting if you have only purchased something from them.)
By the way, I wonder if there is found any measured data of the output impedance of existing tube amps as a function of frequency. I think there is potential for considerable variation with frequency; and this, along with the basic impedance level might well define the sonic characters of the myriad designs more than the distortion properties do.
I encourage here those who have the book and have grasped the issue to some extent to also post reviews at the Amazons or elsewhere, especially if you consider the book to be worth more than the present rating. (Amazon allows posting if you have only purchased something from them.)
I don't see any actual reasons to go 'hollow state' with current-drive. If you, however, manage to construct a working example piece, share it here, as there is definitely not too many of them published.mashaffer said:
By the way, I wonder if there is found any measured data of the output impedance of existing tube amps as a function of frequency. I think there is potential for considerable variation with frequency; and this, along with the basic impedance level might well define the sonic characters of the myriad designs more than the distortion properties do.
Hi,
I have measured a range of Tube Amp's. In general there is not a lot of variation with frequency, except near the bandwidth limits, where the output transformers parasitics (lack of primary inductance & capacitance, leakage inductance) become material.
So depending on precise design the output impedance usually falls below 50....100Hz due to the limited primary inductance and often rises above around 10-20KHz due to the leakage inductance and the fact that the negative feedback needs to be reduced in the interest of stability.
Most non-feedback SE Triode Amplifiers measure around 2-4 Ohm output impedance across most of the bandwidth on the 8 Ohm connection. Other, more exotic types (like the Wavac Amplifiers using Class B RF Triodes in Class A2 with feedback) can be less predictable and often have higher output impedance.
Push-Pull Amplifiers vary more, I have seen as much as 8 Ohm on the 8 Ohm Tap for some "maximum power" designs with low/no negative feedback though I already considered the around 3 Ohm from the big CAT JL-2 Amp as above average, low feedback "Ultralinear" designs average around 2 Ohm, designs with more feedback lower this.
Stereophile's (on line) archives contain a range of tube amplifiers measured with their so-called "speaker load" which allows good levels of inference of output impedance in general. Soundstage (on line) usually posts "damping factor" measurements with frequency if they post measurements, which again can be used to rather precisely infer output impedances.
BTW, I find it incredible how you could post a thread sensationalistic claiming "The Secret of Tube Amplifiers Revealed - and much more!", when clearly your understanding how they perform is rather sketchy, but that's just me.
Ciao T
I have measured a range of Tube Amp's. In general there is not a lot of variation with frequency, except near the bandwidth limits, where the output transformers parasitics (lack of primary inductance & capacitance, leakage inductance) become material.
So depending on precise design the output impedance usually falls below 50....100Hz due to the limited primary inductance and often rises above around 10-20KHz due to the leakage inductance and the fact that the negative feedback needs to be reduced in the interest of stability.
Most non-feedback SE Triode Amplifiers measure around 2-4 Ohm output impedance across most of the bandwidth on the 8 Ohm connection. Other, more exotic types (like the Wavac Amplifiers using Class B RF Triodes in Class A2 with feedback) can be less predictable and often have higher output impedance.
Push-Pull Amplifiers vary more, I have seen as much as 8 Ohm on the 8 Ohm Tap for some "maximum power" designs with low/no negative feedback though I already considered the around 3 Ohm from the big CAT JL-2 Amp as above average, low feedback "Ultralinear" designs average around 2 Ohm, designs with more feedback lower this.
Stereophile's (on line) archives contain a range of tube amplifiers measured with their so-called "speaker load" which allows good levels of inference of output impedance in general. Soundstage (on line) usually posts "damping factor" measurements with frequency if they post measurements, which again can be used to rather precisely infer output impedances.
BTW, I find it incredible how you could post a thread sensationalistic claiming "The Secret of Tube Amplifiers Revealed - and much more!", when clearly your understanding how they perform is rather sketchy, but that's just me.
Ciao T
I am going with tubes simply because that is what I like and understand best not because I am wanting to try current drive. What ever I do topology wise it will be vacuum tubes all around with the possible exception of the subwoofer. For me it is two separate issues. That being said the high plate resistance of pentode mode outputs leads to high Zout in open loop so it seems at least a reasonable place to start.
The part that concerns me is that in open loop mode the harmonic distortion (particularly HD3 and above) seems to run rather high in pentode mode. So even though the distortion generated in the speaker would be reduced we are increasing distortion in another area. I have no idea whether the overall result would be superior.
One possibility would be to use an un-bypassed cathode resistor for output stage bias (though I have heard rumors that the HD spectrum is not necessarily improved even though THD is) but I wonder if there might be a simple way to sample the current either in the OPT primary or secondary and use that as a correction signal to apply back to the driver stage. After all, in this application it is output current that we wish to linearize not necessarily output voltage. Maybe the voltage from a current sense resistor in the output could be coupled to the grid of the driver stage. The grid is very high Z so shouldn't load the output.
Of course just as voltage based gNFB does not always result in better sound than local varieties we may find the same thing in current FB.
Anyway, just some random musings.
The part that concerns me is that in open loop mode the harmonic distortion (particularly HD3 and above) seems to run rather high in pentode mode. So even though the distortion generated in the speaker would be reduced we are increasing distortion in another area. I have no idea whether the overall result would be superior.
One possibility would be to use an un-bypassed cathode resistor for output stage bias (though I have heard rumors that the HD spectrum is not necessarily improved even though THD is) but I wonder if there might be a simple way to sample the current either in the OPT primary or secondary and use that as a correction signal to apply back to the driver stage. After all, in this application it is output current that we wish to linearize not necessarily output voltage. Maybe the voltage from a current sense resistor in the output could be coupled to the grid of the driver stage. The grid is very high Z so shouldn't load the output.
Of course just as voltage based gNFB does not always result in better sound than local varieties we may find the same thing in current FB.
Anyway, just some random musings.
mashaffer,
I can't tell if you already suggested this, but taking feedback from the cathode (in an unbypassed cathode resistor arrangement) is one approach to raising the output impedance of an output triode. The cathode bias resistor also functions as the current sense resistor. The DC component of this feedback tap could be an issue, as might the rolloff caused by a current source driving the input capacitance of the OPT primary.
I can't tell if you already suggested this, but taking feedback from the cathode (in an unbypassed cathode resistor arrangement) is one approach to raising the output impedance of an output triode. The cathode bias resistor also functions as the current sense resistor. The DC component of this feedback tap could be an issue, as might the rolloff caused by a current source driving the input capacitance of the OPT primary.
Hi,ThorstenL said:
It is true that I haven't found the time end interest to delve much into tube amp design(s), for I only see them as a partial solution or partial improvement to the mainstream. The title, however, refers mostly to tube amps as opposed to solid state; the determining factor being here the remarkable output impedance of the former, that is able to effect real change by helping to purify the load current. To conclude that doesn't take so much knowledge on their detailed performance; and the statement, while seemingly bold, is factually justified and not intended to harm anyone's businesses. I also tried a less prominent heading before this one, but it didn't attract any interest.
In my previous posting, I was more thinking about the difference between various tube designs. In high-feedback solid state amps, the output impedance usually increases tenfold towards high frequencies. As I went through the Soundstage reviews, it seems, however, that tube amps perform more evenly in this, and only in some models the frequency variation was really remarkable. Thus, somewhat against my musings, in most cases a single number seems to define this property with reasonable confidence.
Hi,
Then, with all due respect, how can you, from such a position of ignorance, claim to reveal their secrets?
The "remarkable" output impedance is (as I pointed out) revealed to be fairly unremarkable in most cases (1-3 Ohm), enough to cause some modest changes in frequency response for the attached speaker, but not enough to give any material alteration to the kind of distortion to which voltage drive is causal.
Your view that that it is the output impedance that gives the tube amp their characteristic sound is simply wrong, I can tell you that ex-cathedra.
Correct, only someone rather substantially ignorant of the real state of things would draw such a conclusion.
It is plainly and patently wrong, factually speaking. Did you read my post giving the actual measured impedances of tube Amplifiers? Did you chack the other sources I gave?
Yes, now we are at the real point, you used a sensational statement which you believed to be true due a lack of knowledge of the subjsct and which you took no effort to remedy to attract attention to your book.
Yes. High Feedback Solid State Amplifiers have increasing output impedance with frequency, however the absolute values are still minuscule. Moreover, when driving multiway speakers with passive crossovers the additional impedances in the crossover readily swamp out any from common style solid state amplifiers.
Yes, competently designed tube amplifiers generally have an output impedance that is moderately low and does not change appreciably with frequency.
Yes. And a reasonably low number at that, in most cases.
Ciao T
Then, with all due respect, how can you, from such a position of ignorance, claim to reveal their secrets?
The "remarkable" output impedance is (as I pointed out) revealed to be fairly unremarkable in most cases (1-3 Ohm), enough to cause some modest changes in frequency response for the attached speaker, but not enough to give any material alteration to the kind of distortion to which voltage drive is causal.
Your view that that it is the output impedance that gives the tube amp their characteristic sound is simply wrong, I can tell you that ex-cathedra.
Correct, only someone rather substantially ignorant of the real state of things would draw such a conclusion.
It is plainly and patently wrong, factually speaking. Did you read my post giving the actual measured impedances of tube Amplifiers? Did you chack the other sources I gave?
Yes, now we are at the real point, you used a sensational statement which you believed to be true due a lack of knowledge of the subjsct and which you took no effort to remedy to attract attention to your book.
Yes. High Feedback Solid State Amplifiers have increasing output impedance with frequency, however the absolute values are still minuscule. Moreover, when driving multiway speakers with passive crossovers the additional impedances in the crossover readily swamp out any from common style solid state amplifiers.
Yes, competently designed tube amplifiers generally have an output impedance that is moderately low and does not change appreciably with frequency.
Yes. And a reasonably low number at that, in most cases.
Ciao T
ThorstenL,
I find it incredible how your contribution here has so suddenly turned from reasonable and constructive to offensive and also garbling. The title has always been there. It has not disturbed you hitherto. Why does it right now launch such an arrogant burst at this stage of the discussion? Why such ado about one advertisement heading placed in a commercial forum, be it pleasing or not?
I deem it appropriate for advertisers to use the most interesting and appealing features of their wares to effectively address their target audience. Sensation is never a goal per se, but neither should messages be diluted for fear of it.
The book is not about tube amplifier design, and you don't need to be a tube amp specialist to know that they exhibit relatively high output impedance, which is the point. It would not change anything here if I had created a few tunings of my own instead of just listening to ready commercial implementations. This is not a knowledge contest on all the tricks available in tube electronics.
In many tested tube amps, the output impedance is in the range 5-10 ohm. As speaker impedances are also typically of this order, the driving mode can then technically (though not yet sonically) be regarded as a halfway between voltage and current drive, at least outside the crossover regions. As the total impedance of the current path is now double compared to pure voltage mode, the parasitic current components generated by the driver and hence all the impurities contained by them become reduced roughly by half. If e.g. a 6-dB reduction in these essential distortion sources throughout the mid and treble frequencies is nothing material to you, what on earth then might be?
When the output impedance is 3 ohms, these current impurities are reduced by about 3 dB. Even this is still an appreciable figure, though not anything dramatic, and is likely to be perceptible at least in prolonged experience.
If we exclude magic in these issues (like I prefer to do) and possible instability and fault conditions occurring, we have these principal categories of distortion in the amp/speaker system:
- distortion due to amplifier itself
- mechanical/acoustic distortion of the speaker
- electrical distortion of the speaker
The first and second ones are relatively well recognized and understood and as such not news. The third, however, which is largely dependent on the output impedance, has not been generally recognized or properly understood before. This, therefore, is the news that has been forgotten and ignored all too long.
I find it incredible how your contribution here has so suddenly turned from reasonable and constructive to offensive and also garbling. The title has always been there. It has not disturbed you hitherto. Why does it right now launch such an arrogant burst at this stage of the discussion? Why such ado about one advertisement heading placed in a commercial forum, be it pleasing or not?
I deem it appropriate for advertisers to use the most interesting and appealing features of their wares to effectively address their target audience. Sensation is never a goal per se, but neither should messages be diluted for fear of it.
The book is not about tube amplifier design, and you don't need to be a tube amp specialist to know that they exhibit relatively high output impedance, which is the point. It would not change anything here if I had created a few tunings of my own instead of just listening to ready commercial implementations. This is not a knowledge contest on all the tricks available in tube electronics.
In many tested tube amps, the output impedance is in the range 5-10 ohm. As speaker impedances are also typically of this order, the driving mode can then technically (though not yet sonically) be regarded as a halfway between voltage and current drive, at least outside the crossover regions. As the total impedance of the current path is now double compared to pure voltage mode, the parasitic current components generated by the driver and hence all the impurities contained by them become reduced roughly by half. If e.g. a 6-dB reduction in these essential distortion sources throughout the mid and treble frequencies is nothing material to you, what on earth then might be?
When the output impedance is 3 ohms, these current impurities are reduced by about 3 dB. Even this is still an appreciable figure, though not anything dramatic, and is likely to be perceptible at least in prolonged experience.
If we exclude magic in these issues (like I prefer to do) and possible instability and fault conditions occurring, we have these principal categories of distortion in the amp/speaker system:
- distortion due to amplifier itself
- mechanical/acoustic distortion of the speaker
- electrical distortion of the speaker
The first and second ones are relatively well recognized and understood and as such not news. The third, however, which is largely dependent on the output impedance, has not been generally recognized or properly understood before. This, therefore, is the news that has been forgotten and ignored all too long.
You shouldn't distort. The 'secret' was in singular, I'm not promising to analyze the inner workings of the different configurations or their performance in general.
Do not make up things. I have only remarked that the output impedance of certain solid state amplifiers behaves so and so. I've not said a word, explicitly or implicitly, about the effects of that impedance.
WHOA!
Hi Esa
I think calmer heads should prevail. The problem is that you both have a somewhat combative style, certainly Thorsten is well known for his. LOL
The way I see it, you are both making some valid points, but reading your book and on page 111-112 you take a dismissive attitude toward tube amplifiers as effectively as "distortion generators" (my phrase, not yours). This is likely to cause friction with some of us, admittedly me included, and especially here on diyaudio.com that is not the most palatable position to take. Diplomacy is not your strong suit.
It also ignores that not all tubes are equal, as indeed not all SS is. But we are much more open to discuss that devices (and how they are used) have different transfer characterists, regardless whether one type or another - even if one kind of driving you recommed near miraculously overcomes that too (but only at the interface and that leaves a lot more before it).
The sound of tubes is not just summed up by the supposed fact they are higher in distortion (ignoring the kind of distortion) but that sonic footprint of devices goes much deeper than that. For example, the statement by M. Omar Hawksford that the initial movement (one assumes away from a static state) of bi-polars is always nonlinear. He does not suggest NFB as a solution, indeed NFB may only makes it worse.
As it is, I have learned some good things from your book, but your IC solutions are not for me, so the challenge is to come up with non-IC solutions. Hey, challenges like that are good for the soul.
As for tubes being able to supply current-steering, take a look at the work by the likes of Gary Pimm - and indeed hybrid solutions where tubes can have a major role to play. There is the challenge to find zero NFB solutions and not just put the speaker in a voltage feedback loop to achieve current drive. Maybe one day the opportunity to listen to one of those "hollow state" solutions (or hybrid variation) may avail itself to you and start seeing things in a different light.
As for PP Triode amps (my forte') only benefitting 6dB, there are still a lot of speakers out there that have 1:2 ratio, lowest to highest impedance (and then also likely to have flatter electrical phase) - so these amps are still compatible with a large number of speakers. So they are still practical and also of commercial value, important for survival for those making them. Speakers compatible with full current drive are going to continue to be sparse in numbers.
In terms of compromise, it is not a bad solution and at least avoid voltage drive. But, then there is the challenge to loudspeaker designers to make their speakers compatible with both current and voltage drive. It as can be done. And there is potential there as they will sound better with people who are stuck with voltage amps.
I agree with Thorsten that good Triode amps have fairly flat Z across the board without relying on NFB and speaking to people like Menno Vanderveen (a proper physicist, audio transformer designer extraordinaire and tube lover) they also have stable dynamic impedance. With the right speaker I have yet to hear more lifelike dynamics than any SS power amp.
If Martin Colloms is reading this, may I mention the Vacuum State DPA-300B amp (which he knows and I did the earliest prototyping). Again PP of an unsual kind but with OPZ similar to the average speaker load (there is a better and more high power version coming). They don't sound like tubes or were designed to sound like tubes, just the best sound possible!
Cheers, Joe R.
Last edited:
Hi,
So we both find things incredible.
Yes. Your complete ignorance of how tube amplifiers work has only been revealed recently.
What I object to is that you use such a sensationalist title when in fact you know very little about tube amplifiers (and as shall be illustrated nothing about their secrets).
However, we shall just put the title and my objections to the side and instead clear up your fundamental misunderstandings about tube amplifiers, shall we?
Hmmm, define many.
Okay, I will do it for you. I have taken a sample of 28 Tube Amplifiers from 16 different manufacturers for which either Soundstage or Stereophile published output impedance measurements. I noted down the maker & Type, measured output impedance and the link to the source of the data.
I have attached this in Excel format.
My sample was SOMEWHAT managed.
I did exclude Single Ended Amplifiers, as their low output power tends to require matching them to high efficiency speakers. In HE Speakers the problem of current distortion is already drastically reduced while these kind of amplifiers usually also show very high distortion levels (with 3rd HD not dissimilar to that in voltage fed speakers), making them irrelevant to a discussion of mainstream audio; disregarding my personal preference for such amplifiers (with suitably efficient speakers of course).
I further excluded any brands with what I would call wonky designs (designs that highly unorthdox and measure highly unorthodox).
The reason for my sample was to illustrate the common output impedance performance for well designed tube amplifiers, as the term is generally understood nowadays, not to find a small number of extreme outliers.
Here are the results, normalised to driving 8 Ohm Speakers, meaning if the Amplifier has an 8 Ohm Tap the output impedance is for that output (most tube amplifiers tend to also have 4 Ohm Taps from which the output impedance is usually a little more than halve of that at the 8 Ohm Tap).
Maximum Output Impedance for a sample of 28 Tube Amplifiers 3.4 Ohm
Minimum Output Impedance for a sample of 28 Tube Amplifiers 0.1 Ohm
Average Output Impedance for a sample of 28 Tube Amplifiers 1.1 Ohm
Now as anyone knows, when analysing data with a large standard deviation and apparent non-gaussian data distribution it is usually better to use Median and/or geometric means instead of averages.
In this case we find:
Median Output Impedance for a sample of 28 Tube Amplifiers 0.9 Ohm
Mean Output Impedance for a sample of 28 Tube Amplifiers 0.8 Ohm
The above suggests that it is probably fair to suggest that most well designed tube amplifiers exhibit an output impedance of around 1 Ohm when matched to nominal 8 Ohm loads.
However the actual amplifiers exhibit a wide range of actual impedances, though cases with more than 2 Ohm are rare (4 out of a 28 sample or 14% incidence in the actual sample) while cases with an output impedance equal to or lower than average of the sample at 1.1 Ohm are very common (19 out of a 28 sample or 68% incidence in the actual sample)
So, in all amplifiers out of my substantial, but arguably not exhaustive sample, we find much lower impedances than you claim and most are lower by a factor of 5 to 10...
Note, I do not say that with a careful enough search you cannot find a tube amp with 5 or even 10 Ohm output impedance, just that chance of encountering one such is quite low.
So we can save ourselves all your theoretical waffle about what would happen if tube amplifiers had such output impedances, for the mainstream they do not have them and the whole line of enquiry is moot.
It is a complete mystery how you could possibly come up with the notion that
"In many tested tube amps, the output impedance is in the range 5-10 ohm."
It would take magic.
First, it news only in the sense in which what you read in the tabloids is "news" (as in made up on the spot in the whole cloth or if not blown way out of proportion by holding extreme examples as an illustration of what is common, when it is not), nor is it new, as current drive was investigated in detail at least as far back as the 80's (Quarter of a century ago) when it came to my attention.
Pot. Kettle. Black!
Apologies, my mistake, of course you where refering to "THE SECRET", not secrets in general, as in you where revealing the main and only worthwhile secret. I stand corrected.
I do not. Again. Pot. Kettle. Black!
Correct. I in turn remarked that while such a tendency is indeed observable, the actual absolute magnitude of the impedances was minuscule, compared for example to the output impedance of even a well designed tube amplifiers or even the ESR of a crossover capacitor and would be likely swamped by other factors in a complete music replay system.
I did not bring it up, you did.
I merely attempted to put the effects you attached such a great importance to in your post in an actual practical context.
Ciao T
So we both find things incredible.
Yes. Your complete ignorance of how tube amplifiers work has only been revealed recently.
What I object to is that you use such a sensationalist title when in fact you know very little about tube amplifiers (and as shall be illustrated nothing about their secrets).
However, we shall just put the title and my objections to the side and instead clear up your fundamental misunderstandings about tube amplifiers, shall we?
Hmmm, define many.
Okay, I will do it for you. I have taken a sample of 28 Tube Amplifiers from 16 different manufacturers for which either Soundstage or Stereophile published output impedance measurements. I noted down the maker & Type, measured output impedance and the link to the source of the data.
I have attached this in Excel format.
My sample was SOMEWHAT managed.
I did exclude Single Ended Amplifiers, as their low output power tends to require matching them to high efficiency speakers. In HE Speakers the problem of current distortion is already drastically reduced while these kind of amplifiers usually also show very high distortion levels (with 3rd HD not dissimilar to that in voltage fed speakers), making them irrelevant to a discussion of mainstream audio; disregarding my personal preference for such amplifiers (with suitably efficient speakers of course).
I further excluded any brands with what I would call wonky designs (designs that highly unorthdox and measure highly unorthodox).
The reason for my sample was to illustrate the common output impedance performance for well designed tube amplifiers, as the term is generally understood nowadays, not to find a small number of extreme outliers.
Here are the results, normalised to driving 8 Ohm Speakers, meaning if the Amplifier has an 8 Ohm Tap the output impedance is for that output (most tube amplifiers tend to also have 4 Ohm Taps from which the output impedance is usually a little more than halve of that at the 8 Ohm Tap).
Maximum Output Impedance for a sample of 28 Tube Amplifiers 3.4 Ohm
Minimum Output Impedance for a sample of 28 Tube Amplifiers 0.1 Ohm
Average Output Impedance for a sample of 28 Tube Amplifiers 1.1 Ohm
Now as anyone knows, when analysing data with a large standard deviation and apparent non-gaussian data distribution it is usually better to use Median and/or geometric means instead of averages.
In this case we find:
Median Output Impedance for a sample of 28 Tube Amplifiers 0.9 Ohm
Mean Output Impedance for a sample of 28 Tube Amplifiers 0.8 Ohm
The above suggests that it is probably fair to suggest that most well designed tube amplifiers exhibit an output impedance of around 1 Ohm when matched to nominal 8 Ohm loads.
However the actual amplifiers exhibit a wide range of actual impedances, though cases with more than 2 Ohm are rare (4 out of a 28 sample or 14% incidence in the actual sample) while cases with an output impedance equal to or lower than average of the sample at 1.1 Ohm are very common (19 out of a 28 sample or 68% incidence in the actual sample)
So, in all amplifiers out of my substantial, but arguably not exhaustive sample, we find much lower impedances than you claim and most are lower by a factor of 5 to 10...
Note, I do not say that with a careful enough search you cannot find a tube amp with 5 or even 10 Ohm output impedance, just that chance of encountering one such is quite low.
So we can save ourselves all your theoretical waffle about what would happen if tube amplifiers had such output impedances, for the mainstream they do not have them and the whole line of enquiry is moot.
It is a complete mystery how you could possibly come up with the notion that
"In many tested tube amps, the output impedance is in the range 5-10 ohm."
It would take magic.
First, it news only in the sense in which what you read in the tabloids is "news" (as in made up on the spot in the whole cloth or if not blown way out of proportion by holding extreme examples as an illustration of what is common, when it is not), nor is it new, as current drive was investigated in detail at least as far back as the 80's (Quarter of a century ago) when it came to my attention.
Pot. Kettle. Black!
Apologies, my mistake, of course you where refering to "THE SECRET", not secrets in general, as in you where revealing the main and only worthwhile secret. I stand corrected.
I do not. Again. Pot. Kettle. Black!
Correct. I in turn remarked that while such a tendency is indeed observable, the actual absolute magnitude of the impedances was minuscule, compared for example to the output impedance of even a well designed tube amplifiers or even the ESR of a crossover capacitor and would be likely swamped by other factors in a complete music replay system.
I did not bring it up, you did.
I merely attempted to put the effects you attached such a great importance to in your post in an actual practical context.
Ciao T
Last edited:
Last time, I was only substantially ignorant. Now I have complete ignorance. I'm making progress.ThorstenL said:
Too obviously, the end has here justified the means. In high-efficiency speakers, the problem of current distortion is not reduced, in fact it is worsened because the relative proportion of the electromotive forces acting in the circuit is increased. Also, the distortion produced by single-ended amplifiers is mainly 2nd harmonic, which is of the same note as the fundamental, and the levels are often very moderate, nothing to justify the exclusion of the whole category. Interestingly, you rule out even your own preferences from the study. Why does anybody prefer poorly designed amps?ThorstenL said:
I made a survey on the output impedances of all tube amps measured at Soundstage from 2002 to the present, without any biasing filters. The impedance in the middle range was taken. I divided the values into three bins: 0 < Z < 3, 3 =< Z < 5, and Z >= 5. In case an amp had options for the output impedance, the values were treated as separate samples if they fall into different bins. If they fall into the same bin, only one sample was recorded.ThorstenL said:
Here is the result of 34 samples thus collected.
5 ohm or over: 6 of 34 (18%)
3 ohm or over: 11 of 34 (32%)
Minimum: 0.42
Maximum: 17.3
Median: 2.1
6 samples out of 34 exceeded 5 ohms, which is already quite many; and most of the amps reviewed didn't have measurements. About third of the samples were at least 3 ohms, that is yet a significant level.
Mainly due to shifts in frequency balance, it is not easy to conduct studies on the audibility threshold of the current distortion, but if we assume quite realistically it to be near 2 ohms, at least half of the measured amps should more or less benefit from the current purification.
The title does not mean that all tube amplifiers represent the 'secret' in question. Rather, it implies that there exists among tube amplifiers an unexplained beneficial factor, that is now being disclosed. I don't think that there even exists any generic 'tube sound' that would fit to all amplifiers. Instead, as I see it, those 'well designed' ones often don't seem to offer much distinction.
Why is there nowadays so remarkable interest towards many higher-impedance designs and unorthodox paradigms such as 'power drive'? What is it that appeals in them despite their limited power, despite their inferior THD, and despite the prevalent conception of the omnipotence of the 'damping factor'? That also calls for answer.
The title is also applicable to guitar amplifiers and the like. It is difficult to find any actual data on their output impedances, but there is indications that they are at least comparable to the recommended load. I wonder if there is any guitarists reading this that could perform such measurements or point out some sources.
Mills & Hawksford mainly concentrated on developing their extensive amplification system. They did little to address the actual flaws of voltage drive. Even so, what number of DIYers has seen even that paper and related it to tube amps?ThorstenL said:
Quite a few. Besides the distribution and popularization from Nelson Pass, it's been linked to numerous times on this forum, mentioned several times in this thread, and one of the articles on this site (Morgan Jones's Arpeggio loudspeaker) made explicit use of this from a tube amp.
Hi,
It is a mystery why you would say that.
The motors used in "normal" and "high efficiency" speakers are fundamentally similar. Most so-called HE systems achieve the increase in efficiency by reducing moving mass (which leaves the eddy current losses unaffected) and by adding waveguides (horns etc.), so the distortion for a given amount of power applied to the voicecoil is retained, HOWEVER, the SPL at this power is significantly increased.
So for a given SPL a typical HE system will have less problems with eddy current induced distortion and as the power for a given SPL is reduced also the thermal compression is reduced.
I actually covered this on previous occasions where current drive was discussed here.
Again, you clearly show ignorance of the facts.
The distortion in the output stage of a single-ended amplifier is mainly 2nd and 3rd harmonics, with the distribution at "full power" depending on complex factors. You can find examples on the WE Site here:
Western Electric
For example, the most common operating condition for the 300B (350V/80mA and 2-3K load) will produce 5% 2nd HD and over 1% 3rd HD at 9.6 Watt output. I would not call these levels "moderate" even next to speakers...
Once we add a driver stage, which of course also has some amount of distortion, we will reduce the 2nd harmonic by cancellation and increase the 3rd harmonics, plus we will have added 4th and 5th...
I actually did some work on how different driver stages interact with the output stage and what hind of harmonic profiles we get as result:
Harmonic Distortion Profiles for Different Driverstages driving the 300B
That is a good question. Why do some guys prefer slim brunettes and other volouptous blondes. De Gustibus my friend...
And you exclude all measurements by stereophile. Soundstage does not test and review much mainstream items.
And you simply state your results without giving others access to your data. as I did. Got any agenda, perhaps?
On this page at soundstage I count less than 34 Tube amplifiers.
Welcome to www.SoundStageAV.com
So where did you get test data for 34 Amplifiers from? Not soundstage I think? What bias did you apply but NOT state and instead hide by not making your actual source data available?
As I stated, my aim was to illustrate a "typical" tube amplifier, such as one may expect to readily encounter in some-one's system and I made it clear that I was well aware of certain excotic constructs that showed a much higher source impedance.
Your data does not refute this in the least, it merely confirms it.
You get a median of 2 Ohm, I got one of 1 Ohm, in either case the median impedance is much lower than what you claimed earlier, even after including a full sample with many "non-mainstream" items.
With a median level of output impedance as low as mine or even your results you cannot support the statement that:
"in many tested tube amps, the output impedance is in the range 5-10 ohm"
Of course, as "many" is indefinite you can weasel out of this by claiming that even only two cases (or the 6 out of 34 you got) would constitute "many".
First, you omit to even illustrate the degree by which the current distortion is reduced when increasing the source impedance from zero to 2 Ohm, nor do you include any consideration of the 3hd from the Tube amplifiers themselves.
Secondly, you fail to illustrate how the current distortions relates to the other distortion mechanisms in speakers that cause 3rd HD.
And third then you make vacillating statements about the audibility treshold for such changes claiming first it is difficult to conduct studies and offering no evidence whatsoever, but then following up by claiming positively that this threshold you just claim was very difficult to establish co-incides with the median value of your cursory, limited and biased (biased by using only soundstage as source) enquiry.
So basically you make empty statements, backed by zero evidence.
Especially considering that eddy current distortion is only one of the contributing factors to 3rd HD generation in drivers I would instead propose that the difference in observable 3rd HD due to the insertion of 2 Ohm Source Impedance is quite small and can be called insignificant, realistically speaking, while offering at least as much proof as you do for my position.
Well, first of all, as I pointed out, the factor is not unexplained and sec ondly, it exists only in a small minority of cases with amplifiers whose design would not be universally agreed to be competent.
So, you actually compared, in objective and subjective analysis a substantial number of well designed tube amplifiers and well designed solid state amplifiers and found "not much distinction"?
Or do ypu base this "not much distinction" on the fact that they in actual fact have low output impedances only and hence do not support your contention?
Same reason why there was interest in that in the 1960's when a number of companies offered amplifiers with adjustable output impedance and in the 80's when current drive with solid state amplifiers first came to my attention (and I explored it's use).
It is primarily because a substantial number of speaker designs exist that will show a severely attenuated bass response with low impedance amplifiers due to an overdamped design. Increasing the source impedance rights these very obvious flaws in the bass response.
The reduction in 3rd HD I observed then is more incidental.
I answered it - correct bass response with overdamped drivers.
Yes. Guitar amplifiers have high source impedances but also quite intentionally very high levels of distortion.
You seem to be clutching at more straws as you cannot illustrate the claim in your heading. I suggest that you may be better off stopping it, before even more of your ignorance is revealed.
Dunno. I am familiar with the paper and I have related it to tube amplifers.
In fact, I repeatedly commented on the late 1930's german studio monitoring system which combined zero-feedback pentode driven amplifiers (essentially very high source impedance) with a coaxial/coincidental speaker employing a series crossover and mechanical means to damp the fundamental resonance, which resulted in a remarkably flat electrical impedance.
And your point is?
My point is that you are unaware of how MOST or THE MAJORITY of tube amplifiers work and what their performance parameters are regarding HD and output impedance (and others) and that despite this you made claims to the contrary in your introduction to this thread, after (as you admitted yourself) you failed to attract much attention using more realistic and topical titles.
And I feel that this was a poor choice, as it detracts from your quite interesting and substantially accurate work regarding "current drive" for speakers.
I would suggest you change the title of the thread to something that is topical and ask the moderators to remove our debate regarding it at the same time, so what stands is the topical discussion of current drive and not your poor judgement.
Ciao T
It is a mystery why you would say that.
The motors used in "normal" and "high efficiency" speakers are fundamentally similar. Most so-called HE systems achieve the increase in efficiency by reducing moving mass (which leaves the eddy current losses unaffected) and by adding waveguides (horns etc.), so the distortion for a given amount of power applied to the voicecoil is retained, HOWEVER, the SPL at this power is significantly increased.
So for a given SPL a typical HE system will have less problems with eddy current induced distortion and as the power for a given SPL is reduced also the thermal compression is reduced.
I actually covered this on previous occasions where current drive was discussed here.
Again, you clearly show ignorance of the facts.
The distortion in the output stage of a single-ended amplifier is mainly 2nd and 3rd harmonics, with the distribution at "full power" depending on complex factors. You can find examples on the WE Site here:
Western Electric
For example, the most common operating condition for the 300B (350V/80mA and 2-3K load) will produce 5% 2nd HD and over 1% 3rd HD at 9.6 Watt output. I would not call these levels "moderate" even next to speakers...
Once we add a driver stage, which of course also has some amount of distortion, we will reduce the 2nd harmonic by cancellation and increase the 3rd harmonics, plus we will have added 4th and 5th...
I actually did some work on how different driver stages interact with the output stage and what hind of harmonic profiles we get as result:
Harmonic Distortion Profiles for Different Driverstages driving the 300B
That is a good question. Why do some guys prefer slim brunettes and other volouptous blondes. De Gustibus my friend...
And you exclude all measurements by stereophile. Soundstage does not test and review much mainstream items.
And you simply state your results without giving others access to your data. as I did. Got any agenda, perhaps?
On this page at soundstage I count less than 34 Tube amplifiers.
Welcome to www.SoundStageAV.com
So where did you get test data for 34 Amplifiers from? Not soundstage I think? What bias did you apply but NOT state and instead hide by not making your actual source data available?
As I stated, my aim was to illustrate a "typical" tube amplifier, such as one may expect to readily encounter in some-one's system and I made it clear that I was well aware of certain excotic constructs that showed a much higher source impedance.
Your data does not refute this in the least, it merely confirms it.
You get a median of 2 Ohm, I got one of 1 Ohm, in either case the median impedance is much lower than what you claimed earlier, even after including a full sample with many "non-mainstream" items.
With a median level of output impedance as low as mine or even your results you cannot support the statement that:
"in many tested tube amps, the output impedance is in the range 5-10 ohm"
Of course, as "many" is indefinite you can weasel out of this by claiming that even only two cases (or the 6 out of 34 you got) would constitute "many".
First, you omit to even illustrate the degree by which the current distortion is reduced when increasing the source impedance from zero to 2 Ohm, nor do you include any consideration of the 3hd from the Tube amplifiers themselves.
Secondly, you fail to illustrate how the current distortions relates to the other distortion mechanisms in speakers that cause 3rd HD.
And third then you make vacillating statements about the audibility treshold for such changes claiming first it is difficult to conduct studies and offering no evidence whatsoever, but then following up by claiming positively that this threshold you just claim was very difficult to establish co-incides with the median value of your cursory, limited and biased (biased by using only soundstage as source) enquiry.
So basically you make empty statements, backed by zero evidence.
Especially considering that eddy current distortion is only one of the contributing factors to 3rd HD generation in drivers I would instead propose that the difference in observable 3rd HD due to the insertion of 2 Ohm Source Impedance is quite small and can be called insignificant, realistically speaking, while offering at least as much proof as you do for my position.
Well, first of all, as I pointed out, the factor is not unexplained and sec ondly, it exists only in a small minority of cases with amplifiers whose design would not be universally agreed to be competent.
So, you actually compared, in objective and subjective analysis a substantial number of well designed tube amplifiers and well designed solid state amplifiers and found "not much distinction"?
Or do ypu base this "not much distinction" on the fact that they in actual fact have low output impedances only and hence do not support your contention?
Same reason why there was interest in that in the 1960's when a number of companies offered amplifiers with adjustable output impedance and in the 80's when current drive with solid state amplifiers first came to my attention (and I explored it's use).
It is primarily because a substantial number of speaker designs exist that will show a severely attenuated bass response with low impedance amplifiers due to an overdamped design. Increasing the source impedance rights these very obvious flaws in the bass response.
The reduction in 3rd HD I observed then is more incidental.
I answered it - correct bass response with overdamped drivers.
Yes. Guitar amplifiers have high source impedances but also quite intentionally very high levels of distortion.
You seem to be clutching at more straws as you cannot illustrate the claim in your heading. I suggest that you may be better off stopping it, before even more of your ignorance is revealed.
Dunno. I am familiar with the paper and I have related it to tube amplifers.
In fact, I repeatedly commented on the late 1930's german studio monitoring system which combined zero-feedback pentode driven amplifiers (essentially very high source impedance) with a coaxial/coincidental speaker employing a series crossover and mechanical means to damp the fundamental resonance, which resulted in a remarkably flat electrical impedance.
And your point is?
My point is that you are unaware of how MOST or THE MAJORITY of tube amplifiers work and what their performance parameters are regarding HD and output impedance (and others) and that despite this you made claims to the contrary in your introduction to this thread, after (as you admitted yourself) you failed to attract much attention using more realistic and topical titles.
And I feel that this was a poor choice, as it detracts from your quite interesting and substantially accurate work regarding "current drive" for speakers.
I would suggest you change the title of the thread to something that is topical and ask the moderators to remove our debate regarding it at the same time, so what stands is the topical discussion of current drive and not your poor judgement.
Ciao T
Hi,
> In HE Speakers the problem of current distortion is already
> drastically reduced
As a rule, the current distortion follows the cube of the current passed through the voice coil. As there is very limited scope in the magnet system, to increase efficiency (they are usually all maxed out anyway) a speaker with higher efficiency will have less current flowing in the voice coil for a given SPL than a lower efficiency one, reducing the current distortion accordingly.
Due to the cubic function in the current distortion the reduction for a 10dB increase in efficiency is quite dramatic.
> while these kind of amplifiers usually also show very high distortion
> levels (with 3rd HD not dissimilar to that in voltage fed speakers)
SE Amplifiers have high levels of distortion. Their 3rd HD reaches levels that are comparable to those of most speaker drivers at around 1 Watt or less.
So, my point for excluding SE Amplifiers was that:
1) The Speakers they are commonly used with usually show less of an issue with current distortion than those common in the HiFi/High-End Mainstream.
2) The Amplifiers themselves have so much HD that the reduction of HD in the speaker may very well swamped by the amplifiers distortion.
3) Waht I did not say was that the 2nd HD from the Amplifier is also in similar levels to that of the driver and tracks very well over a wide range of power, so that it is possible to use the high 2nd HD from the SE Amplifier to reduce the 2nd HD in the speaker, yielding an system that is overall more linear.
Ciao T
Could you kindly elaborate more on these seemingly confused statements?
> In HE Speakers the problem of current distortion is already
> drastically reduced
As a rule, the current distortion follows the cube of the current passed through the voice coil. As there is very limited scope in the magnet system, to increase efficiency (they are usually all maxed out anyway) a speaker with higher efficiency will have less current flowing in the voice coil for a given SPL than a lower efficiency one, reducing the current distortion accordingly.
Due to the cubic function in the current distortion the reduction for a 10dB increase in efficiency is quite dramatic.
> while these kind of amplifiers usually also show very high distortion
> levels (with 3rd HD not dissimilar to that in voltage fed speakers)
SE Amplifiers have high levels of distortion. Their 3rd HD reaches levels that are comparable to those of most speaker drivers at around 1 Watt or less.
So, my point for excluding SE Amplifiers was that:
1) The Speakers they are commonly used with usually show less of an issue with current distortion than those common in the HiFi/High-End Mainstream.
2) The Amplifiers themselves have so much HD that the reduction of HD in the speaker may very well swamped by the amplifiers distortion.
3) Waht I did not say was that the 2nd HD from the Amplifier is also in similar levels to that of the driver and tracks very well over a wide range of power, so that it is possible to use the high 2nd HD from the SE Amplifier to reduce the 2nd HD in the speaker, yielding an system that is overall more linear.
Ciao T
Could you kindly elaborate more on these seemingly confused statements?
Hi Thorsten,
you can`t increase efficiency without giving up linearity. Horn speakers exhibit vastly high distortion. For lowest distortion you must employ zero compression.
The harmonic distortion produced by SE amplifiers is high indeed. All harmonics should be amplified equally, however, no amplification device is capable to do that. A handful of directly heated triodes from the thirties come the closest to it, the harmonic order magnitudes pretty evenly falling with rising frequency, when no feedback applied. Since it largely coincides with the ear´s distortion pattern, the distortion will be inaudible.
Devices in balanced configuration do not amplify even order harmonics, resulting in spectral discontinuity and, of course, an irrelevant lower measured distortion. Feedback upsets the natural harmonic distribution, marring the unique transparency and sound stage of linear triodes - as a price of the reduced output impedance.
THD is completely meaningless, referring to levels of detached harmonic orders is completely meaningless, only a detailed account of the entire spectrum, including all the high orders is meaningful.
Your simplistic view on distortions is in need of deepening.
Your criticism of Esa lacks any basis.
you can`t increase efficiency without giving up linearity. Horn speakers exhibit vastly high distortion. For lowest distortion you must employ zero compression.
The harmonic distortion produced by SE amplifiers is high indeed. All harmonics should be amplified equally, however, no amplification device is capable to do that. A handful of directly heated triodes from the thirties come the closest to it, the harmonic order magnitudes pretty evenly falling with rising frequency, when no feedback applied. Since it largely coincides with the ear´s distortion pattern, the distortion will be inaudible.
Devices in balanced configuration do not amplify even order harmonics, resulting in spectral discontinuity and, of course, an irrelevant lower measured distortion. Feedback upsets the natural harmonic distribution, marring the unique transparency and sound stage of linear triodes - as a price of the reduced output impedance.
THD is completely meaningless, referring to levels of detached harmonic orders is completely meaningless, only a detailed account of the entire spectrum, including all the high orders is meaningful.
Your simplistic view on distortions is in need of deepening.
Your criticism of Esa lacks any basis.
Hi,
Now why would make such a ludicrous claim like this?
The quintessential Horn Speaker, Avantgarde was found by Martin Colloms to have very low distortion when he reviewed it, so much so he found it worth mentioning because it was so much lower than speakers he normally tested (Stereophile June 1998).
I also have with the measured distortion of a Tannoy Autograph Speaker originating with MJ at around 100dB/1m SPL (1W power). Interestingly it shows around equal levels of 2nd and 3rd HD (3rd HD slightly lower) for the cone driver part between 100Hz and 1KHz (also quite flat distribution) at around 0.3%.
Above 1KHz where the compression driver takes over 2nd HD is increased (as one would expect from the non-linearity of the air in the compression chamber, to around 0.6% (eyballing), but 3rd HD (the one that is reduced by current drive) is dropped dramatically to well below 0.1%.
I am unaware of conventional 2-Way speakers with median efficiency (87dB/1W/1m) that come even close to similar distortion levels at 100dB/1m SPL even with current drive.
So it seems we must consign the "Horns have high distortion" Myth to the same dustbin of history as the "Tube amplifiers have 5-10 Ohm Output Impedance" has already been slam dunked into.
Note, I am not saying that there are no horns with high distortion (or dome tweeters or cone drivers for that), just as I am not saying that there no tube Amplifiers with high output impedance. I am however noting that these are extreme outliers, not the norm.
Apart from that, it is not necessary to employ horns or compression drivers for high efficiency.
Well, HD in SE Amplifiers will tend to change the original notes harmonic spectrum, however this will be mainly a preponderance of 2nd HD and the change will be small. Still, I found that cancelling 2nd HD in speakers with the SE Amplifiers 2nd HD (needs to be correct phase relationship, so measurements are needed) sounded notably cleaner.
I see you read my old article about SE Amplifiers where pointed all this out.
This is untrue. Devices in balanced configuration (or with even order HD cancellation between multiple stages add LESS 2nd HD (ideally non) to the signal, but the original harmonics of the signal are correctly amplified (and of course additionally distorted).
You focus only on the amplifier, please consider the whole system including the speaker.
Not as such. Feedback lowers the lower order harmonics at the expense of a multiplication process that generates higher order products, which have a much greater audibility.
The output impedance of triode Amplifiers can be reasonably low without feedback. My own approaches to SE Amplifiers operates the Tubes at somewhat non-conventional operating conditions which as a side-effect also lowers output impedance, at the cost of a slight reduction in power.
Funny, I have been saying this for years.
My view is NOT simplistic. The issue debated here is "current purification" caused by higher drive impedances, specifically with regards to the eddy current distortion in the pole piece of the driver enclosed within the voice coil (BTW, the driver in one of my commercial speakers gets around this by placing the non-conductive magnet inside the voice coil and the solid steel part forming the magnetic circuit outside, a very efficient way to negate the issue in systems with voltage drive).
As I have written years ago, this whole eddy current malarkey effects only the odd order harmonics (as the distortion follows a cubic function).
The highest level harmonic caused by this is 3rd HD (the rest can be largely be implied by understanding the mechanisms and knowing the 3HD levels) and is often measured for speakers (whereas higher harmonics rearely are).
This is why I am mainly writing about 3rd HD.
Really? My criticism is that he made claims about tube amplifiers in what is essentially advertising his book, despite a near complete ignorance of how tube amplifiers really function.
Do you suggest that he has shown any appreciable indication that he does understand how tube amplifiers actually function and what kind of performance parameters one would expect from a tube amplifier?
Ciao T
Now why would make such a ludicrous claim like this?
The quintessential Horn Speaker, Avantgarde was found by Martin Colloms to have very low distortion when he reviewed it, so much so he found it worth mentioning because it was so much lower than speakers he normally tested (Stereophile June 1998).
I also have with the measured distortion of a Tannoy Autograph Speaker originating with MJ at around 100dB/1m SPL (1W power). Interestingly it shows around equal levels of 2nd and 3rd HD (3rd HD slightly lower) for the cone driver part between 100Hz and 1KHz (also quite flat distribution) at around 0.3%.
Above 1KHz where the compression driver takes over 2nd HD is increased (as one would expect from the non-linearity of the air in the compression chamber, to around 0.6% (eyballing), but 3rd HD (the one that is reduced by current drive) is dropped dramatically to well below 0.1%.
I am unaware of conventional 2-Way speakers with median efficiency (87dB/1W/1m) that come even close to similar distortion levels at 100dB/1m SPL even with current drive.
So it seems we must consign the "Horns have high distortion" Myth to the same dustbin of history as the "Tube amplifiers have 5-10 Ohm Output Impedance" has already been slam dunked into.
Note, I am not saying that there are no horns with high distortion (or dome tweeters or cone drivers for that), just as I am not saying that there no tube Amplifiers with high output impedance. I am however noting that these are extreme outliers, not the norm.
Apart from that, it is not necessary to employ horns or compression drivers for high efficiency.
Well, HD in SE Amplifiers will tend to change the original notes harmonic spectrum, however this will be mainly a preponderance of 2nd HD and the change will be small. Still, I found that cancelling 2nd HD in speakers with the SE Amplifiers 2nd HD (needs to be correct phase relationship, so measurements are needed) sounded notably cleaner.
I see you read my old article about SE Amplifiers where pointed all this out.
This is untrue. Devices in balanced configuration (or with even order HD cancellation between multiple stages add LESS 2nd HD (ideally non) to the signal, but the original harmonics of the signal are correctly amplified (and of course additionally distorted).
You focus only on the amplifier, please consider the whole system including the speaker.
Not as such. Feedback lowers the lower order harmonics at the expense of a multiplication process that generates higher order products, which have a much greater audibility.
The output impedance of triode Amplifiers can be reasonably low without feedback. My own approaches to SE Amplifiers operates the Tubes at somewhat non-conventional operating conditions which as a side-effect also lowers output impedance, at the cost of a slight reduction in power.
Funny, I have been saying this for years.
My view is NOT simplistic. The issue debated here is "current purification" caused by higher drive impedances, specifically with regards to the eddy current distortion in the pole piece of the driver enclosed within the voice coil (BTW, the driver in one of my commercial speakers gets around this by placing the non-conductive magnet inside the voice coil and the solid steel part forming the magnetic circuit outside, a very efficient way to negate the issue in systems with voltage drive).
As I have written years ago, this whole eddy current malarkey effects only the odd order harmonics (as the distortion follows a cubic function).
The highest level harmonic caused by this is 3rd HD (the rest can be largely be implied by understanding the mechanisms and knowing the 3HD levels) and is often measured for speakers (whereas higher harmonics rearely are).
This is why I am mainly writing about 3rd HD.
Really? My criticism is that he made claims about tube amplifiers in what is essentially advertising his book, despite a near complete ignorance of how tube amplifiers really function.
Do you suggest that he has shown any appreciable indication that he does understand how tube amplifiers actually function and what kind of performance parameters one would expect from a tube amplifier?
Ciao T
May I add to your discussion? (not an expert on that topic...)
We have in France a company that launched succesfully a serie of amplifier using a knob to adjust the output impedance of the amplifier.
Having discussed with one of the designer, it appears that the main argument to promote this was that, for a specific output impedance, the shape of the power signal injected on the speaker was the same as the input voltage of the amplifier.
So as a voltage drive gives the same voltage shape, current drive provides the same current shape.
What is the influence of having the power shape correct?
My 2 cents is that having that shape and supposing that accoustic power is proportionnal to power injected to speaker, the shape of the radiated pressure will be the correct shape? We can now radiate a nice rectangular pulse?
The lateral argument was that driving speakers this way improved imaging by reducing phase error...
Do you have any feedback for such an assertion?
We have in France a company that launched succesfully a serie of amplifier using a knob to adjust the output impedance of the amplifier.
Having discussed with one of the designer, it appears that the main argument to promote this was that, for a specific output impedance, the shape of the power signal injected on the speaker was the same as the input voltage of the amplifier.
So as a voltage drive gives the same voltage shape, current drive provides the same current shape.
What is the influence of having the power shape correct?
My 2 cents is that having that shape and supposing that accoustic power is proportionnal to power injected to speaker, the shape of the radiated pressure will be the correct shape? We can now radiate a nice rectangular pulse?
The lateral argument was that driving speakers this way improved imaging by reducing phase error...
Do you have any feedback for such an assertion?
The reason why the number of samples is greater than of the amps becomes apparent from my text (Some amplifiers had options.); still you decided to raise suspicions. The data is from here. I have had to omit many other sources too, as my time for this is limited. The choice of Soundstage was random. I wouldn't call any tube amplifiers 'mainstream' - curiosities they are all.ThorstenL said:
I got a median of 2.1 and I have made no earlier claims about it. Your quibbling doesn't help.ThorstenL said:
I haven't 'claimed'. I made a realistic assumption. It is left to the reader to assess whether the assumption is valid.ThorstenL said:
I think I have now found an explanation why you are talking what you are talking. You somehow consider the 'current distortion' to mean only or mostly the 3rd HD due to eddy currents, whereas I have considered all of the nonharmonic, harmonic and other, unclassified, effects of distortion and interference. This may explain much of your objection of the title and claims of ignorance.ThorstenL said:
I have always known that tube amps produce 3rd HD along with other harmonics if that is the great knowledge you thought I was lacking. I am also fully aware that harmonics originating from different stages in the signal path may strengthen or weaken each other. However, the whole issue of 3rd HD is rather secondary in the whole scope of the current distortion effects, that are largely nonharmonic or phase muddling or even indefinite by nature. This debate has thus gone somewhat on the side track about the harmonic content of the systems, that is though part of the picture but yet secondary in this context.
Of course, if the 3rd HD due to eddy currents and the power compression are the only effects that one can comprehend of the flaws of voltage drive, the title may remain odd; but when the whole picture of the current-degrading phenomena is considered, things can be seen in a better light. I have tried to bring these effects to your attention already in post #193, but seemingly you haven't taken any heed in this. Instead, from your position of ignorance, you choose to deliver your judgment.
It's not only the bass. There also seems to happen something in the other ranges. No, I don't have a doctoral thesis to show for proof; it's just my general observation.ThorstenL said:
Hi,
They used to have such knobs on a few Amplifiers in the 1950's as well.
Well, it is some way between voltage and current drive. So it comes with the drawbacks of both and non of the advantages of either.
It is very clear that at the most basic basic physical level the speakers reaction is to current and current only, not voltage or power.
Nope, cannot, does not. The situation will improve compared to voltage drive and we will have more damping than current drive, so in the real world it may be a reasonable compromise.
About the assertion that reducing phase error improves imaging? There insufficient data IMHO to draw such a conclusion and some indications to the contrary.
Or that driving a speaker with increased source impedance improves phase-response? It does not as such, in any reliable way, it will alter the frequency response and as speakers are minimum phase systems flattening frequency response reduces phase errors as well, increasing frequency response errors also increases phase errors. Without being able to measure how the specific output impedance impacts on the frequency response (and phase response) it is impossible to say.
Ciao T
They used to have such knobs on a few Amplifiers in the 1950's as well.
Well, it is some way between voltage and current drive. So it comes with the drawbacks of both and non of the advantages of either.
It is very clear that at the most basic basic physical level the speakers reaction is to current and current only, not voltage or power.
Nope, cannot, does not. The situation will improve compared to voltage drive and we will have more damping than current drive, so in the real world it may be a reasonable compromise.
About the assertion that reducing phase error improves imaging? There insufficient data IMHO to draw such a conclusion and some indications to the contrary.
Or that driving a speaker with increased source impedance improves phase-response? It does not as such, in any reliable way, it will alter the frequency response and as speakers are minimum phase systems flattening frequency response reduces phase errors as well, increasing frequency response errors also increases phase errors. Without being able to measure how the specific output impedance impacts on the frequency response (and phase response) it is impossible to say.
Ciao T
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.