Simple is as simple does........
"Steve: That is ultimate simplicity!"
http://www.diyaudio.com/forums/attachment.php?s=&postid=213170
Hmmm.......... A high inductive transformer winding and a highly capacitive mosfet gate with no damping or voltage clamps. Yep, that is about the simplest designer that I have seen here.
"Steve: That is ultimate simplicity!"
http://www.diyaudio.com/forums/attachment.php?s=&postid=213170
Hmmm.......... A high inductive transformer winding and a highly capacitive mosfet gate with no damping or voltage clamps. Yep, that is about the simplest designer that I have seen here.
Re: Simple is as simple does........
No damping? What, did Jensen discover room temperature superconductors and are using them to wind their transformers unbeknownst to everyone else?
And what highly inductive transformer winding? The only inductance the input capacitance of the MOSFET would see is the small amount of leakage inductance, which is in series with the non-resistance of those superconductive windings.
se
Fred Dieckmann said:
No damping? What, did Jensen discover room temperature superconductors and are using them to wind their transformers unbeknownst to everyone else?
And what highly inductive transformer winding? The only inductance the input capacitance of the MOSFET would see is the small amount of leakage inductance, which is in series with the non-resistance of those superconductive windings.
se
Where did supercondutors come into this?
"No damping? What, did Jensen discover room temperature superconductors and are using them to wind their transformers unbeknownst to everyone else?
And what highly inductive transformer winding? The only inductance the input capacitance of the MOSFET would see is the small amount of leakage inductance, which is in series with the non-resistance of those superconductive windings.
se"
"Not if they're loaded properly. The Jensen JT-13K7-A which is a 1:5 step-up has a flat response across its bandwidth, which is from 0.9Hz to 120kHz."
I am sorry I thought anyone would know I was taking about adding external damping resistor, especially you after the loading statment. BTW it would only see solely leakage inductance with the primary shorted Perhaps you tell us the resistive component of the leakage inductance ( actually it would be a a leakage impedance if it is an RL circuit as you stated) and tell us the resonant frequency it will from with the mosfet capacitance and the resulting Q of that circuit . It couldn't possibly be a high Q circuit that will ring like hell, could it? Pure idle speculation of my part I'm sure. Isn't this the same guy that puts a snubber on the secondary of his isolation transformers? The number times the recommendation for putting gate damping resistors on Fets has been posted here (usually by Mr. Pass) is in the double digits. Gate damper resistors are a good idea for even wiring inductance resonating with gate capacitance and oscillating. I thought you knew a lot about transformers and capacitive loading. I apologize, I won't make that mistake again. It will probably only oscillate or overshoot long enough to punch through the Mosfet gate dielectric and destroy the transistor and stop any oscillation, or amplification. Be sure to DC couple the output for further parts saving, you're speaker won't mind the resulting DC in a fault condition described above. I suppose this is just a matter of opinion though. You would never risk offering advice on a subject that was beyond your understanding and could result in an amplifier blowing up from building such a circuit.
I will however apologize sincerely for the statement:
"Yep, that is about the simplest designer that I have seen here."
It was supposed to read:
"Yep, that is about the simplest DESIGN that I have seen here."
It was sloppiness and haste. I had no intention of stating or even implying ( for the many that read between the lines on this forum) that you were a simpleton. Thanks for not taking offense at my carelessness.
mea culpa,
Fred
"No damping? What, did Jensen discover room temperature superconductors and are using them to wind their transformers unbeknownst to everyone else?
And what highly inductive transformer winding? The only inductance the input capacitance of the MOSFET would see is the small amount of leakage inductance, which is in series with the non-resistance of those superconductive windings.
se"
"Not if they're loaded properly. The Jensen JT-13K7-A which is a 1:5 step-up has a flat response across its bandwidth, which is from 0.9Hz to 120kHz."
I am sorry I thought anyone would know I was taking about adding external damping resistor, especially you after the loading statment. BTW it would only see solely leakage inductance with the primary shorted Perhaps you tell us the resistive component of the leakage inductance ( actually it would be a a leakage impedance if it is an RL circuit as you stated) and tell us the resonant frequency it will from with the mosfet capacitance and the resulting Q of that circuit . It couldn't possibly be a high Q circuit that will ring like hell, could it? Pure idle speculation of my part I'm sure. Isn't this the same guy that puts a snubber on the secondary of his isolation transformers? The number times the recommendation for putting gate damping resistors on Fets has been posted here (usually by Mr. Pass) is in the double digits. Gate damper resistors are a good idea for even wiring inductance resonating with gate capacitance and oscillating. I thought you knew a lot about transformers and capacitive loading. I apologize, I won't make that mistake again. It will probably only oscillate or overshoot long enough to punch through the Mosfet gate dielectric and destroy the transistor and stop any oscillation, or amplification. Be sure to DC couple the output for further parts saving, you're speaker won't mind the resulting DC in a fault condition described above. I suppose this is just a matter of opinion though. You would never risk offering advice on a subject that was beyond your understanding and could result in an amplifier blowing up from building such a circuit.
I will however apologize sincerely for the statement:
"Yep, that is about the simplest designer that I have seen here."
It was supposed to read:
"Yep, that is about the simplest DESIGN that I have seen here."
It was sloppiness and haste. I had no intention of stating or even implying ( for the many that read between the lines on this forum) that you were a simpleton. Thanks for not taking offense at my carelessness.
mea culpa,
Fred
There's been some hard feelings here About some words that were said
If I may reflect a bit on the subject at hand......
"You're looking at it backwards. A 1:9 step-up transformer doesn't reflect the load back to the source by a factor of 81, but a factor of 1/81 (looking back from the load to the source it's a step-down transformer). It's the source that gets reflected to the load by a factor of 81."
Reducing the impedance by a factor of the 81 is the the same thing as increasing the capacitance by the same factor.
http://w1.859.telia.com/~u85920178/begin/calc-00.htm
Reactance of a Capacitor
Formula: Reactance = 1 / (2 * Pi * F * C)
When you look into one winding the load on the secondary winding relflects by the square of the turns ratio. That's why the call them step up or step down transformers and why the distinction is
one of which side your looking into.
http://www.bowest.com.au/library/machines.html#01
The primary circuit impedance Z1 referred to the secondary circuit for an ideal transformer with N1 primary turns and N2 secondary turns is:
Z12 = Z1(N2 / N1)2
The secondary circuit impedance Z2 referred to the primary circuit for an ideal transformer with N1 primary turns and N2 secondary turns is:
Z21 = Z2(N1 / N2)2
As Paul Simon said "One man's ceiling is another mans floor"
Have I told you how much we have missed you. I usually have to pay money for entertainment and education like this. but the truth will set you free.
Yeah we tease him a lot cause we've got him on the spot, welcome back,
Fred
If I may reflect a bit on the subject at hand......
"You're looking at it backwards. A 1:9 step-up transformer doesn't reflect the load back to the source by a factor of 81, but a factor of 1/81 (looking back from the load to the source it's a step-down transformer). It's the source that gets reflected to the load by a factor of 81."
Reducing the impedance by a factor of the 81 is the the same thing as increasing the capacitance by the same factor.
http://w1.859.telia.com/~u85920178/begin/calc-00.htm
Reactance of a Capacitor
Formula: Reactance = 1 / (2 * Pi * F * C)
When you look into one winding the load on the secondary winding relflects by the square of the turns ratio. That's why the call them step up or step down transformers and why the distinction is
one of which side your looking into.
http://www.bowest.com.au/library/machines.html#01
The primary circuit impedance Z1 referred to the secondary circuit for an ideal transformer with N1 primary turns and N2 secondary turns is:
Z12 = Z1(N2 / N1)2
The secondary circuit impedance Z2 referred to the primary circuit for an ideal transformer with N1 primary turns and N2 secondary turns is:
Z21 = Z2(N1 / N2)2
As Paul Simon said "One man's ceiling is another mans floor"
Have I told you how much we have missed you. I usually have to pay money for entertainment and education like this. but the truth will set you free.
Yeah we tease him a lot cause we've got him on the spot, welcome back,
Fred
Re: Where did supercondutors come into this?
I know you were talking about an external damping resistor. But you said there was no damping. To wit:
A high inductive transformer winding and a highly capacitive mosfet gate with no damping or voltage clamps.
In order for that to be the case, the secondary would have to have been wound with some sort of resistanceless superconductive wire.
Point being, there's damping built-in in the form of the secondary's winding resistance, which in the case of the transformer I had in mind, is about 470 ohms.
The only other inductance it would see would be any inductance from the source, which would typically just be the inductance of the interconnects.
The resistive component would be the secondary winding resistance which again in this case is 471 ohms. Of course the source resistance also becomes part of this.
Don't see how it could be with such a small amount of inductance and a large amount of resistance. Certainly there will be some resonance, but I don't see how it could have a terribly high Q.
Well, I think you could stand to brush up a bit on transformer basics.
Yes, but it's optional. Some people freak if there's any deviation from objective perfection.
The damping is there. It's just not written out literally in the idealized schematic.
But in that case, the ratio of the inductance to wire resistance is much smaller. In the case of a transformer's secondary, the ratio of the winding resistance to the leakage inductance is far far greater.
And I thought you knew at least a little something about transformers. I apologize. I won't make that mistake again.
Christ. Get a grip on yourself, Fred.
I understand that the leakage inductance is quite small and winding resistance quite high. And I wasn't the the one effectively claiming that transformers have zero winding resistance and therefore no damping.
se
Fred Dieckmann said:
I know you were talking about an external damping resistor. But you said there was no damping. To wit:
A high inductive transformer winding and a highly capacitive mosfet gate with no damping or voltage clamps.
In order for that to be the case, the secondary would have to have been wound with some sort of resistanceless superconductive wire.
Point being, there's damping built-in in the form of the secondary's winding resistance, which in the case of the transformer I had in mind, is about 470 ohms.
The only other inductance it would see would be any inductance from the source, which would typically just be the inductance of the interconnects.
The resistive component would be the secondary winding resistance which again in this case is 471 ohms. Of course the source resistance also becomes part of this.
Don't see how it could be with such a small amount of inductance and a large amount of resistance. Certainly there will be some resonance, but I don't see how it could have a terribly high Q.
Well, I think you could stand to brush up a bit on transformer basics.
Yes, but it's optional. Some people freak if there's any deviation from objective perfection.
The damping is there. It's just not written out literally in the idealized schematic.
But in that case, the ratio of the inductance to wire resistance is much smaller. In the case of a transformer's secondary, the ratio of the winding resistance to the leakage inductance is far far greater.
And I thought you knew at least a little something about transformers. I apologize. I won't make that mistake again.
Christ. Get a grip on yourself, Fred.
I understand that the leakage inductance is quite small and winding resistance quite high. And I wasn't the the one effectively claiming that transformers have zero winding resistance and therefore no damping.
se
Re: There's been some hard feelings here About some words that were said
Aye. Oversight on my part.
Just returning the favor. I'd have never known that transformer manufacturers were now using room temperature superconducting wire. Thanks for bringing me up to speed on that one.
se
Fred Dieckmann said:
Aye. Oversight on my part.
Just returning the favor. I'd have never known that transformer manufacturers were now using room temperature superconducting wire. Thanks for bringing me up to speed on that one.
se
Thanks for bringing me up to speed on that one.
Not at all. For guy who wants talk about parasitics like leakage inductance, you should probably learn the basics equations for reflected impedance first. Leakage inductance is distributed parameter like winding capacitance and not accurately described by a simple lump element leakage inductance term in an idealized model of a transformer. The reflected impedance from the primary is complex impedance and a function of the source impedance and the interconnect which is a transmission line and not a simple inductance and capacitance. I have worked on pulse transformer design with compensation networks and EMI reduction on switching power supplies. But you would rather talk about superconductors and idealized models that you don't even have the basics down for. Thanks for the advice on transformer basics. You know I would even have a problem discussing this with someone who knew something about the subject, but to be second guessed someone who has barely dipped his toe into some pretty complicated design issues is just priceless......... A little knowledge is a dangerous thing. A very little knowledge with pretensions of being an expert, well that is true humor that I enjoy beyond description. That is why I read this forum, for just such moments of transcendental absurdity. I just pray the Jocko doesn't read this until the next time he comes over, so I can watch him roll around on the floor in hysterics. thanks for straighting me out on about ten years engineering misperceptions, including a failure to understand any of the half a dozen books I have read on magnetics, the computer models, and the time domain reflectometry measurements. Thanks again for your breathtaking efforts to entertain and educate us. I wouldn't have it any other way.
One of your biggest fans,
Fred
Not at all. For guy who wants talk about parasitics like leakage inductance, you should probably learn the basics equations for reflected impedance first. Leakage inductance is distributed parameter like winding capacitance and not accurately described by a simple lump element leakage inductance term in an idealized model of a transformer. The reflected impedance from the primary is complex impedance and a function of the source impedance and the interconnect which is a transmission line and not a simple inductance and capacitance. I have worked on pulse transformer design with compensation networks and EMI reduction on switching power supplies. But you would rather talk about superconductors and idealized models that you don't even have the basics down for. Thanks for the advice on transformer basics. You know I would even have a problem discussing this with someone who knew something about the subject, but to be second guessed someone who has barely dipped his toe into some pretty complicated design issues is just priceless......... A little knowledge is a dangerous thing. A very little knowledge with pretensions of being an expert, well that is true humor that I enjoy beyond description. That is why I read this forum, for just such moments of transcendental absurdity. I just pray the Jocko doesn't read this until the next time he comes over, so I can watch him roll around on the floor in hysterics. thanks for straighting me out on about ten years engineering misperceptions, including a failure to understand any of the half a dozen books I have read on magnetics, the computer models, and the time domain reflectometry measurements. Thanks again for your breathtaking efforts to entertain and educate us. I wouldn't have it any other way.
One of your biggest fans,
Fred
Re: Thanks for bringing me up to speed on that one.
I'm well aware of the basic equations for reflected impedance. My error was one of laziness and not thinking through and realizing that where a purely capacitive load is being considered, reducing the load impedance by a factor of 81 is the same as increasing capacitance by the same factor.
What!?
Lumped parameters are just fine until the system size starts approaching 1/4 wavelength of the shortest wavelength you plan on passing thruogh the system or the system is capable of passing. Seeing as we're talking about audio here, even considering an unusually sluggish propagation velocity of 50% C, at 20kHz we're talking about a wavelength of about 5 miles, 1/4 of that being a little over a mile.
And by the way, leakage inductance and winding capacitance aren't part of the idealized model. That's why it's called an idealized model.
Still boils down to the lumped inductance and capacitance at audio frequencies.
So? This is audio. We're not trying to pass narrow 100kHz pulses through the system.
You apparently don't even know what an idealized model is. Or for that matter what the word idealized means. When you start adding parasitics, it's no longer an idealized model. So when I'm talking about leakage inductance and winding resistance and capacitance, I'm clearly not talking about the idealized model.
se
Fred Dieckmann said:
I'm well aware of the basic equations for reflected impedance. My error was one of laziness and not thinking through and realizing that where a purely capacitive load is being considered, reducing the load impedance by a factor of 81 is the same as increasing capacitance by the same factor.
What!?
Lumped parameters are just fine until the system size starts approaching 1/4 wavelength of the shortest wavelength you plan on passing thruogh the system or the system is capable of passing. Seeing as we're talking about audio here, even considering an unusually sluggish propagation velocity of 50% C, at 20kHz we're talking about a wavelength of about 5 miles, 1/4 of that being a little over a mile.
And by the way, leakage inductance and winding capacitance aren't part of the idealized model. That's why it's called an idealized model.
Still boils down to the lumped inductance and capacitance at audio frequencies.
So? This is audio. We're not trying to pass narrow 100kHz pulses through the system.
You apparently don't even know what an idealized model is. Or for that matter what the word idealized means. When you start adding parasitics, it's no longer an idealized model. So when I'm talking about leakage inductance and winding resistance and capacitance, I'm clearly not talking about the idealized model.
se
fdegrove said:
It's a mutual thing. I remind Fred of things he's overlooked like winding resistance, that leakage inductance is small and not large (at least in a good quality transformer), that an idealized model doesn't include parasitics, and that lumped parameters are just fine for systems larger than 1/4 wavelength.
While he reminds me of things I've overlooked like capacitive reactance decreasing as frequency increases.
I've owned up to my error. When's Fred going to own up to his? He still hasn't even owned up to his claim that there's no damping in which he completely overlooked secondary winding resistance.
Not long ago you tried taking me to task for not owning up to even the slightest mistakes (of course that ended up being your own misreading of what I wrote), so why don't you also take Fred to task for not doing the same? Are you a professional toady and have a contract with Fred or something?
se
Fools Rush In Where Angels Fear to Tread
I never claimed zero winding resistance at any time. That was your usual tactic of inserting claims that were never there. Reading between the lines is an accusation you love to make about others concerning you post.
You have demonstrated your lack of understanding of the basic equations for the impedance transformation. Now you want to convince us that you understand the limitations of idealized models and can predict the influence of parasitics that people who write books on the subject, say to measure since they cannot be modeled or easily predicted. Breakdown of the gate dieletric from excessive gate to source voltage as been discussed to death. I believe it is the reason that real life designers put source resistors and zener clamps in mosfet circuits. Predicting the overshoot, Q, and oscillation of a resonant circuit (containing real parasistics and not lumped element approximations) to be at level that is of no problem, is a talent that no really good engineer I ever worked with, would dare to claim. Perhaps it is the 30 years of experience fixing circuits designed by people who made simplistic assumptions about circuit parameters, without the benefit of actual measurements. Your willingness to make assumptions based on the most basic simplified models with no math, no computer models, no measurements, and no real design experience; just defies comprehension. Or it would be if we hadn't seen you paint yourself in a corner on to many occasions to count. I know you want to learn and I respect that. You obviously read enough to know many of the basics. You work with amplifier designers who could answer many of these questions. Yet you want to read a few basics, skip the intermediate step of learning, and fly right to the expert level only to be shot down over and over and over again. I feel guilty for even intervening most of the time, but you misinform to a level that confuses and frustrates the questioner. It undermines the credibility of the entire forum. I cannot see that you entertain or enlighten in the majority of your post. This is not a giant chat room or BS session. People want to learn things that make better sounding and reliable audio projects. They do not want to build something that will blow up, work erratically, or cause endless problems because somebody was interest in seeing themselves in print with a damned near total indifference to the accuracy of his post. It takes one person post trash to undermine the efforts of ten people who really want to contribute to the art and learn something. Call me all the names you want and smear away, I can handle it. Keep posting questionable information, outright mistakes, and challenging statements of fact that can be easily checked with credible references (often in about 5 minutes on the web) and expect equal time from those who seriously disagree with your post and are not going to sit on their hands while you run amok. It is not only their right but their duty if they really care about this forum.
Watching and waiting,
Fred
Steve Eddy said:
I know you were talking about an external damping resistor. But you said there was no damping. To wit:
A high inductive transformer winding and a highly capacitive mosfet gate with no damping or voltage clamps.
In order for that to be the case, the secondary would have to have been wound with some sort of resistanceless superconductive wire.
Point being, there's damping built-in in the form of the secondary's winding resistance, which in the case of the transformer I had in mind, is about 470 ohms.
The only other inductance it would see would be any inductance from the source, which would typically just be the inductance of the interconnects.
The resistive component would be the secondary winding resistance which again in this case is 471 ohms. Of course the source resistance also becomes part of this.
Don't see how it could be with such a small amount of inductance and a large amount of resistance. Certainly there will be some resonance, but I don't see how it could have a terribly high Q.
Well, I think you could stand to brush up a bit on transformer basics.
Yes, but it's optional. Some people freak if there's any deviation from objective perfection.
The damping is there. It's just not written out literally in the idealized schematic.
But in that case, the ratio of the inductance to wire resistance is much smaller. In the case of a transformer's secondary, the ratio of the winding resistance to the leakage inductance is far far greater.
And I thought you knew at least a little something about transformers. I apologize. I won't make that mistake again.
Christ. Get a grip on yourself, Fred.
I understand that the leakage inductance is quite small and winding resistance quite high. And I wasn't the the one effectively claiming that transformers have zero winding resistance and therefore no damping.
se
I never claimed zero winding resistance at any time. That was your usual tactic of inserting claims that were never there. Reading between the lines is an accusation you love to make about others concerning you post.
You have demonstrated your lack of understanding of the basic equations for the impedance transformation. Now you want to convince us that you understand the limitations of idealized models and can predict the influence of parasitics that people who write books on the subject, say to measure since they cannot be modeled or easily predicted. Breakdown of the gate dieletric from excessive gate to source voltage as been discussed to death. I believe it is the reason that real life designers put source resistors and zener clamps in mosfet circuits. Predicting the overshoot, Q, and oscillation of a resonant circuit (containing real parasistics and not lumped element approximations) to be at level that is of no problem, is a talent that no really good engineer I ever worked with, would dare to claim. Perhaps it is the 30 years of experience fixing circuits designed by people who made simplistic assumptions about circuit parameters, without the benefit of actual measurements. Your willingness to make assumptions based on the most basic simplified models with no math, no computer models, no measurements, and no real design experience; just defies comprehension. Or it would be if we hadn't seen you paint yourself in a corner on to many occasions to count. I know you want to learn and I respect that. You obviously read enough to know many of the basics. You work with amplifier designers who could answer many of these questions. Yet you want to read a few basics, skip the intermediate step of learning, and fly right to the expert level only to be shot down over and over and over again. I feel guilty for even intervening most of the time, but you misinform to a level that confuses and frustrates the questioner. It undermines the credibility of the entire forum. I cannot see that you entertain or enlighten in the majority of your post. This is not a giant chat room or BS session. People want to learn things that make better sounding and reliable audio projects. They do not want to build something that will blow up, work erratically, or cause endless problems because somebody was interest in seeing themselves in print with a damned near total indifference to the accuracy of his post. It takes one person post trash to undermine the efforts of ten people who really want to contribute to the art and learn something. Call me all the names you want and smear away, I can handle it. Keep posting questionable information, outright mistakes, and challenging statements of fact that can be easily checked with credible references (often in about 5 minutes on the web) and expect equal time from those who seriously disagree with your post and are not going to sit on their hands while you run amok. It is not only their right but their duty if they really care about this forum.
Watching and waiting,
Fred
transformer damping
Fred and Steve: I think you both understand what you're saying and the battle is now over semantics. I think there's a semantics forum somewhere more appropriate to your current disagreements; I'll try to find you the link.
In any case, despite the unecessary acrimony, this has been an interesting discussion (the part about the transformers, not the part where you accuse each other of having shot Kennedy) because it's not a technology I've done much work with. And it's doubly helpful in that I've now got some order-of-magnitude numbers to plug into back-of-the-envelope calculations.
Fred and Steve: I think you both understand what you're saying and the battle is now over semantics. I think there's a semantics forum somewhere more appropriate to your current disagreements; I'll try to find you the link.
In any case, despite the unecessary acrimony, this has been an interesting discussion (the part about the transformers, not the part where you accuse each other of having shot Kennedy) because it's not a technology I've done much work with. And it's doubly helpful in that I've now got some order-of-magnitude numbers to plug into back-of-the-envelope calculations.
Re: Fools Rush In Where Angels Fear to Tread
You claimed there was no damping. In order for that claim to be true, there would have to be no winding resistance. So which is it? Were you incorrect in saying there was no damping or were you correct in that there's no winding resistance?
I demonstrated no such thing.
Here's what I said:
A 1:9 step-up transformer doesn't reflect the load back to the source by a factor of 81, but a factor of 1/81 (looking back from the load to the source it's a step-down transformer). It's the source that gets reflected to the load by a factor of 81.
You responded with:
When you look into one winding the load on the secondary winding relflects by the square of the turns ratio. That's why the call them step up or step down transformers and why the distinction is one of which side your looking into.
http://www.bowest.com.au/library/machines.html#01
The primary circuit impedance Z1 referred to the secondary circuit for an ideal transformer with N1 primary turns and N2 secondary turns is:
Z12 = Z1(N2 / N1)2
The secondary circuit impedance Z2 referred to the primary circuit for an ideal transformer with N1 primary turns and N2 secondary turns is:
Z21 = Z2(N1 / N2)2
That's entirely consistent with what I had said.
Z12 = Z1(N2 / N1)2 is the same as saying "It's the source that gets reflected to the load by a factor of 81."
Put another way, Z12 = Z1(N2 / N1)2 is the same as Z12 = Z1 x 81.
Z21 = Z2(N1 / N2)2 is the same as saying "1:9 step-up transformer doesn't reflect the load back to the source by a factor of 81, but a factor of 1/81 (looking back from the load to the source it's a step-down transformer)."
Put another way, Z21 = Z2(N1 / N2)2 is the same as Z21 = Z2 x (1/81).
In other words, for a 1:9 step up transformer, the source sees the load multiplied by a factor of 81 and the load sees the source multiplied by a factor of 1/81.
Here, let's work it out.
Since it doesn't matter how many actual turns there are (it's only the turns ratio that counts as far as the equations go) we can model the 1:9 step up transformer as a transformer with 1 turn on its primary and 9 turns on its secondary.
The source gets reflected to the load by Z12 = Z1(N2 / N1)2 = Z1(9/1)2 = Z1(9)2 = Z1 x 81.
The load gets reflected to the source by Z21 = Z2(N1 / N2)2 = Z2(1/9)2 = Z2(0.1111111)2 = Z2 x 0.0123457. And of course 1/81, which is the way I chose to put it, is equal to 0.0123457.
So, as I said originally, given a 1:9 step-up transformer, the load is reflected back to the source by a factor of 81 and the source gets reflected to the load by a factor of 1/81.
So where exactly have I demonstrated a lack of understanding of the basic equations for the impedance transformation of a transformer?
You've only demonstrated that you didn't bother to think about what I'd actually said, much like I didn't bother to think about the capacitance of a capcitive load having to effectively increase in order to result in a lower impedance as seen by the source.
What are you talking about? I only said that the leakage inductance in a good quality transformer is very small. Has nothing to do with predicting anything. It's simply the result of good design. And the transformer I had in mind (the Jensen JT-13K7-A which I had already mentioned previously) already exists and its parasitics are already known.
Well you certainly seem to be VERY selective in who you single out for making such criticisms. I don't recall your criticizing all the other circuits that had been presented in this thread which didn't use clamping diodes.
I guess Jean Hiraga must not be a real life designer. Don't see any clamping diodes in his Nemesis amplifiers here:
http://www.bonavolta.ch/hobby/en/audio/2sk135se.htm
Don't see any clamping diodes in tschrama's design here:
http://home.zonnet.nl/tschrama/owndesign.html
Don't see any clamping diodes in tom-vdl's schematic here:
http://www.diyaudio.com/forums/attachment.php?s=&postid=211477
Don't see any clamping diodes in the schamatics from TubeCAD Journal here:
http://www.tubecad.com/page11.html
And here:
http://www.tubecad.com/page12.html
Did I miss something? Or did you also criticize all these other schematics for not including any clamping diodes?
So who are these real life designers you're speaking of? How 'bout Nelson Pass? Is he a real life designer? Let's see. Here's the schematic for the Son of Zen:
Not a clamping diode in sight, unless my eyes are failing me. Do you see any there Fred? If so, could you please point them out to me? Or could you please point me to a post of yours where you took Nelson Pass to task for not including any clamping diodes in the schematic for the Son of Zen?
You're the one working off assumption here.
Let's go back and see exactly what I said when I first posted the schematic that's got your panties all in a bunch here.
Well I've never really looked at MOSFETs in depth, but why couldn't something like this work? Just four parts.
I didn't present it as any sort of complete design expecting people to run out and start building it. I offered it merely as an idea and asked "why couldn't something like this work?"
You responded by going on about no damping and high inductance vis a vis MOSFET input capacitance. I pointed out that there is indeed damping and that the inductance is just a small amount of leakage inductance. You've been dancing around like Fred Astaire ever since, going on about the difficulty of predicting parasitics when predicting the parasitics is a non-issue as the parasitics are already known and then this nonsense about having to use distributed elements rather than lumped.
What corner have I painted myself into here? You said there was no damping and high inductance. The facts are that there is damping and little inductance.
And you never misinform, right, Fred? Your claiming no damping and high inductance when indeed there is damping and low inductance doesn't misinform?
I've never claimed to be nor passed myself off as any sort of expert. I make mistakes same as everyone else, and when I do make them, I own up to them.
You on the other hand make mistakes and never seem to own up to them. You just go on spewing more BS to try and cover the previous mistakes.
Again, the schematic I posted WAS NEVER PRESENTED AS ANY SORT OF COMPLETE DESIGN. Why do you insist on making it seem as if I had?
And where is your criticism of all those other schematics which were posted as complete designs which lacked the clamping diodes you feel are so all important? Why aren't you berating Jean Hiraga and Nelson Pass for presenting a COMPLETE designs which do not include clamping diodes in the schematics?
se
Fred Dieckmann said:
You claimed there was no damping. In order for that claim to be true, there would have to be no winding resistance. So which is it? Were you incorrect in saying there was no damping or were you correct in that there's no winding resistance?
I demonstrated no such thing.
Here's what I said:
A 1:9 step-up transformer doesn't reflect the load back to the source by a factor of 81, but a factor of 1/81 (looking back from the load to the source it's a step-down transformer). It's the source that gets reflected to the load by a factor of 81.
You responded with:
When you look into one winding the load on the secondary winding relflects by the square of the turns ratio. That's why the call them step up or step down transformers and why the distinction is one of which side your looking into.
http://www.bowest.com.au/library/machines.html#01
The primary circuit impedance Z1 referred to the secondary circuit for an ideal transformer with N1 primary turns and N2 secondary turns is:
Z12 = Z1(N2 / N1)2
The secondary circuit impedance Z2 referred to the primary circuit for an ideal transformer with N1 primary turns and N2 secondary turns is:
Z21 = Z2(N1 / N2)2
That's entirely consistent with what I had said.
Z12 = Z1(N2 / N1)2 is the same as saying "It's the source that gets reflected to the load by a factor of 81."
Put another way, Z12 = Z1(N2 / N1)2 is the same as Z12 = Z1 x 81.
Z21 = Z2(N1 / N2)2 is the same as saying "1:9 step-up transformer doesn't reflect the load back to the source by a factor of 81, but a factor of 1/81 (looking back from the load to the source it's a step-down transformer)."
Put another way, Z21 = Z2(N1 / N2)2 is the same as Z21 = Z2 x (1/81).
In other words, for a 1:9 step up transformer, the source sees the load multiplied by a factor of 81 and the load sees the source multiplied by a factor of 1/81.
Here, let's work it out.
Since it doesn't matter how many actual turns there are (it's only the turns ratio that counts as far as the equations go) we can model the 1:9 step up transformer as a transformer with 1 turn on its primary and 9 turns on its secondary.
The source gets reflected to the load by Z12 = Z1(N2 / N1)2 = Z1(9/1)2 = Z1(9)2 = Z1 x 81.
The load gets reflected to the source by Z21 = Z2(N1 / N2)2 = Z2(1/9)2 = Z2(0.1111111)2 = Z2 x 0.0123457. And of course 1/81, which is the way I chose to put it, is equal to 0.0123457.
So, as I said originally, given a 1:9 step-up transformer, the load is reflected back to the source by a factor of 81 and the source gets reflected to the load by a factor of 1/81.
So where exactly have I demonstrated a lack of understanding of the basic equations for the impedance transformation of a transformer?
You've only demonstrated that you didn't bother to think about what I'd actually said, much like I didn't bother to think about the capacitance of a capcitive load having to effectively increase in order to result in a lower impedance as seen by the source.
What are you talking about? I only said that the leakage inductance in a good quality transformer is very small. Has nothing to do with predicting anything. It's simply the result of good design. And the transformer I had in mind (the Jensen JT-13K7-A which I had already mentioned previously) already exists and its parasitics are already known.
Well you certainly seem to be VERY selective in who you single out for making such criticisms. I don't recall your criticizing all the other circuits that had been presented in this thread which didn't use clamping diodes.
I guess Jean Hiraga must not be a real life designer. Don't see any clamping diodes in his Nemesis amplifiers here:
http://www.bonavolta.ch/hobby/en/audio/2sk135se.htm
Don't see any clamping diodes in tschrama's design here:
http://home.zonnet.nl/tschrama/owndesign.html
Don't see any clamping diodes in tom-vdl's schematic here:
http://www.diyaudio.com/forums/attachment.php?s=&postid=211477
Don't see any clamping diodes in the schamatics from TubeCAD Journal here:
http://www.tubecad.com/page11.html
And here:
http://www.tubecad.com/page12.html
Did I miss something? Or did you also criticize all these other schematics for not including any clamping diodes?
So who are these real life designers you're speaking of? How 'bout Nelson Pass? Is he a real life designer? Let's see. Here's the schematic for the Son of Zen:
An externally hosted image should be here but it was not working when we last tested it.
Not a clamping diode in sight, unless my eyes are failing me. Do you see any there Fred? If so, could you please point them out to me? Or could you please point me to a post of yours where you took Nelson Pass to task for not including any clamping diodes in the schematic for the Son of Zen?
You're the one working off assumption here.
Let's go back and see exactly what I said when I first posted the schematic that's got your panties all in a bunch here.
Well I've never really looked at MOSFETs in depth, but why couldn't something like this work? Just four parts.
I didn't present it as any sort of complete design expecting people to run out and start building it. I offered it merely as an idea and asked "why couldn't something like this work?"
You responded by going on about no damping and high inductance vis a vis MOSFET input capacitance. I pointed out that there is indeed damping and that the inductance is just a small amount of leakage inductance. You've been dancing around like Fred Astaire ever since, going on about the difficulty of predicting parasitics when predicting the parasitics is a non-issue as the parasitics are already known and then this nonsense about having to use distributed elements rather than lumped.
What corner have I painted myself into here? You said there was no damping and high inductance. The facts are that there is damping and little inductance.
And you never misinform, right, Fred? Your claiming no damping and high inductance when indeed there is damping and low inductance doesn't misinform?
I've never claimed to be nor passed myself off as any sort of expert. I make mistakes same as everyone else, and when I do make them, I own up to them.
You on the other hand make mistakes and never seem to own up to them. You just go on spewing more BS to try and cover the previous mistakes.
Again, the schematic I posted WAS NEVER PRESENTED AS ANY SORT OF COMPLETE DESIGN. Why do you insist on making it seem as if I had?
And where is your criticism of all those other schematics which were posted as complete designs which lacked the clamping diodes you feel are so all important? Why aren't you berating Jean Hiraga and Nelson Pass for presenting a COMPLETE designs which do not include clamping diodes in the schematics?
se
Hi,
Correct. Mr. Jean Hiraga does not consider himself a designer.
He's a journalist who's done formidable work as a go between and as such brought the French audio community in touch with the Japanese audio scene back in the seventies.
Most of the circuits published, including the copies shown at Mr. Claudio Bonavoltas' site won't work correctly and will never achieve their sonic goals unless accompanied by the corresponding articles which are all in French.
Now, while I agree that a xformer coupled FET can work, it certainly will require much more design work than what was shown here before, it looks simple but isn't.
Cheers,
Correct. Mr. Jean Hiraga does not consider himself a designer.
He's a journalist who's done formidable work as a go between and as such brought the French audio community in touch with the Japanese audio scene back in the seventies.
Most of the circuits published, including the copies shown at Mr. Claudio Bonavoltas' site won't work correctly and will never achieve their sonic goals unless accompanied by the corresponding articles which are all in French.
Now, while I agree that a xformer coupled FET can work, it certainly will require much more design work than what was shown here before, it looks simple but isn't.
Cheers,
fdegrove said:
Ok, so what's the excuse for Nelson?
Anyway, it's a bit odd don't you think that Fred singles out something which wasn't even offered as a completed design?
So what needs to be added? The MOSFETs I was thinking of using are BUZ901Ps, which have the protection diodes built in for those who like to play it as safe as possible (though it's cap coupled on the output so you're not going to blow up your speakers anyway).
There's nearly 500 ohms of damping resistance in the secondary windings (more than the usual 100-200 ohms) and just a small amount of leakage inductance so I don't see any oscillation issues.
So how much more complicated need it be? I mean, if you wan to keep Fred happy, just add a resistor.
se
Hi,
All in all I don't think it's odd for Fred to single it out, he just beat me to it for I would have done the exact same thing.
While I agree that when pointed to possible flaws, by Fred, you started to downtone the design, he's still absolutely correct in pointing out the fact that it's just not as simple as you present it.
As such it would have been better, in the context of this thread, one transistor clapping remember, that you wouldn't have posted the schematic as such.
By your own admission you're not knowledgeable enough to understand the risks involved for, and I absolutely agree with Fred here, the circuit is destructive.
From further reading you, I can only conclude you don't know the different types of FETs that are out there well enough to know what type can be used for what and so on.
Again misguided information to other members is added by you when you don't seem to be aware of what depletion mode FETs can be used for in audio etc.
So, why do you argue? Do you think that the resistance of the secondary will save you when we talk about the function of a gatestopper? Amongst a bunch of other things to consider?
I realise that this may all have stemmed from an initial misunderstanding between yourself and Fred but does it really help anyone arguing about it at nauseam?
Set the record straight as adults and move on.
Oh, and if you really feel confident, why don't you present a working circuit?
It can be done, you know...just not as econimically as you present it.
So folks, put the personality clashes aside and move on, please.
I am sure all members have second guessed you and had that FET in mind...
This is again sooo typical of your way of defending yourself that it's becoming a little too transparent to older members, you always have some twisted answer ready, haven't you?
But fine, let someone build that circuit with that FET and let's take it from there even though you argued those diodes wheren't necessary in the first place...see what we mean?
Cheers...
All in all I don't think it's odd for Fred to single it out, he just beat me to it for I would have done the exact same thing.
While I agree that when pointed to possible flaws, by Fred, you started to downtone the design, he's still absolutely correct in pointing out the fact that it's just not as simple as you present it.
As such it would have been better, in the context of this thread, one transistor clapping remember, that you wouldn't have posted the schematic as such.
By your own admission you're not knowledgeable enough to understand the risks involved for, and I absolutely agree with Fred here, the circuit is destructive.
From further reading you, I can only conclude you don't know the different types of FETs that are out there well enough to know what type can be used for what and so on.
Again misguided information to other members is added by you when you don't seem to be aware of what depletion mode FETs can be used for in audio etc.
So, why do you argue? Do you think that the resistance of the secondary will save you when we talk about the function of a gatestopper? Amongst a bunch of other things to consider?
I realise that this may all have stemmed from an initial misunderstanding between yourself and Fred but does it really help anyone arguing about it at nauseam?
Set the record straight as adults and move on.
Oh, and if you really feel confident, why don't you present a working circuit?
It can be done, you know...just not as econimically as you present it.
So folks, put the personality clashes aside and move on, please.
I am sure all members have second guessed you and had that FET in mind...
This is again sooo typical of your way of defending yourself that it's becoming a little too transparent to older members, you always have some twisted answer ready, haven't you?
But fine, let someone build that circuit with that FET and let's take it from there even though you argued those diodes wheren't necessary in the first place...see what we mean?
Cheers...
fdegrove said:
So what? You miss the point. Why is it that half a dozen other schematics were posted in this thread which didn't indicate any clamping diodes and there wasn't a peep out of either you or Fred. But when I post a schematic that doesn't indicate any clamping diodes and Fred goes into a tizzy?
In other words, what makes the lack of clamping diodes in my schematic any different than the lack of clamping diodes in the other schematics? I see no difference. So how do you explain my schematic being singled out other than you and Fred having some sort of obsession with me?
What I was addressing was his making out as if I had presented the design as a complete design. Something which I did not do.
With regard to the issues Fred brought up with regard to his claim of there being no damping and high inductance I don't downtone the design at all.
There is damping and there is very little inductance.
This isn't based on any sort of wild guesses or inadequate modeling as Fred keeps trying to misportray it. It's based on the realworld behavior of such transformers.
Fred simply overlooked the winding resistance and was under the mistaken notion that the secondary's inductance is what it would be if it weren't coupled to the primary winding (i.e. as if it were a stand-alone inductor).
Instead of just saying "Oops" and moving on, he had to try and save face so he started going off on tangents which were wholly irrelevant such as such as inadequate models, distributed elements, etc.
Why?
First, I didn't post it claiming it was a complete design. Second, it hasn't been shown that it won't function as-is. The only other bone of contenion are the protection diodes which would mean that none of those other half dozen schematics shouldn't have been posted either.
But you're not arguing that those shouldn't have been posted. So again, what's the difference?
I didn't just fall off the turnip truck, Frank. I'm well aware of the risks involved. I chose not to include a gate stopper resistor because of the damping provided by the transformer. I didn't include any clamping diodes because I didn't see them as an absolute necessity (as apparently neither does Nelson vis a vis his Son of Zen) particularly seeing as I was only trying to get the basic concept across.
As for it being destructive, the circuit's capacitively coupled to the load so you're not going to get DC across your speaker. And if you're worried about the possibility of oscillation, then you'd better go after those Gainclone schematics which haven't any high frequency bypass caps and output zobels.
Bottom line, if anyone else had posted the exact same schematic that I posted, neither you or Fred would have would have uttered a word.
What reading is that?
Perhaps you'd care to tell us, Frank.
What I said was:
Seems that depletion mode power MOSFETs are only good for switches. You aware of any depletion mode power MOSFETs that are suitable for audio?
I based that "seems" on my having searched the net and couldn't find any references of their being used for power audio purposes except for constant current sources by the bottleheads.
Hell, Curl wasn't even sure that 2SK60 parts he had were depletion mode until he ran 'em up on a curve tracer.
So tell us of some depletion mode power MOSFETs that are suitable for audio purposes other than current sources.
I didn't argue about depletion mode MOSFETs. I said that it SEEMS they're only good as switches and ASKED if the original poster was aware of any which would be suitable for audio. How is that an argument? If I were insisting that there were none suiable for audio I wouldn't have bothered to ask if he was aware of any.
No one has so far shown that it won't. Fred said there would be no damping, which clearly isn't the case. So perhaps you'd care to explain why it won't.
Which things are those? Or is this going to be yet another instance of your saying vague things and never providing any actual details?
I only argued against the BS claims made by Fred. He claimed there was no damping. I argued that there was. He claimed the secondary's inductance was high. I argued that it wasn't, that it was just a small amount of leakage inductance. Fred claimed that you had to use distributed element analysis. I argued that lumped elements were just fine given the small size of the system relative to the wavelengths involved.
Fred doesn't know how to act like an adult.
Well, so far no one's shown that the circuit can't work as-is.
Fred's argument was based on the erroneous assumptions of no damping and high inductance.
For what it's worth, I ran it by Rod Elliot and he said it would work.
How do you know? Tell me specifically why it won't.
It doesn't matter whether they second-guessed which MOSFET I had in mind. They're not absolutely necessary.
If they're so absolutely necessary, then where the hell are they in the Son of Zen schematic?
Here's what Nelson says in the Son of Zen article:
Mosfets are sensitive to static discharge, and should be handled carefully to avoid damaging the gate due to static. Elaborate precautions are not essential, just try to use some common sense when handling the devices. The Son of Zen is not provided with static input protection, and requires moderate care when hooking up the amplifier to a source. You can provide input protection if you want by placing back-to-back zener diode pairs at each input to ground. The ratings of the diodes should be 10 to 16 volts.
Now, Nelson actually DOES pass himself off as a designer AND as an authority on amplifier design. His Son of Zen article wasn't just two sentences long and included "why couldn't something like this work?" It was a full-length feature article formally published in Audio Amateur.
So here's what you do, Frank. First you go rip Nelson a new one for publishing such a slipshod article which doesn't include the necessary protection diodes in the schematic and takes a cavallier "if you want to" approach to protection and that he should either modify the article or remove it from his web site because it's just too irresponsible.
THEN you can come chew my *** out and at least have a shred of credibility.
se
Steve,
You know a lot about audio. No question. Your comments about protection diodes clearly show your passionate belief that they should be there to prevent outputs laying down their lives. Truth is, SOARs have improved so much in the last 15 years that a lot of designers don't bother. Very few reactive loads threaten modern output stages these days. Who really cares given the relentless pressure to keep the component count down? Actually, I have included clamping diodes across the outputs on my latest high power design, but I'm still tossing up whether to bother as I'm using 230V transistors on 50V rails.
To go on.... Why stand around at an auto accident? Why yell in public? Who cares who's right or wrong? What's it prove? That you can express yourself loudly, and well?
Steve, why defend your position with such verbal violence? No one is sleeping with your wife, fer crissakes....... No honor has been trampled, no dignity exposed.
And Frank, Fred, why get so upset?
Walk away from the fight, and realize that most people are just here to learn something, not to witness spars. Bloody ridiculous!!
Cheers,
Hugh
You know a lot about audio. No question. Your comments about protection diodes clearly show your passionate belief that they should be there to prevent outputs laying down their lives. Truth is, SOARs have improved so much in the last 15 years that a lot of designers don't bother. Very few reactive loads threaten modern output stages these days. Who really cares given the relentless pressure to keep the component count down? Actually, I have included clamping diodes across the outputs on my latest high power design, but I'm still tossing up whether to bother as I'm using 230V transistors on 50V rails.
To go on.... Why stand around at an auto accident? Why yell in public? Who cares who's right or wrong? What's it prove? That you can express yourself loudly, and well?
Steve, why defend your position with such verbal violence? No one is sleeping with your wife, fer crissakes....... No honor has been trampled, no dignity exposed.
And Frank, Fred, why get so upset?
Walk away from the fight, and realize that most people are just here to learn something, not to witness spars. Bloody ridiculous!!
Cheers,
Hugh
Well, I had a bad feeling when Fred welcomed Mr. Curl onboard describing the forum as the fight club. In fact, that is NOT the general vibe here at DIY Audio. In fact that is usually the first comment of newbies- that this place has a level of maturity above that of AA. Then you (3) guys make a special effort t o show that this isn't so.
It's understandable when we get an 18 year old onboard with his "in your face" style. It's been quite heartening that a lot of these come around to civilized discourse. But what is are your excuses for the 18 year old maturity level? You assume that since you in fact have the most to contribute, you are entitled to special consideration of your pathetic tantrums. Your guys are an embarassment to the forum.
Thank You,
Mark
It's understandable when we get an 18 year old onboard with his "in your face" style. It's been quite heartening that a lot of these come around to civilized discourse. But what is are your excuses for the 18 year old maturity level? You assume that since you in fact have the most to contribute, you are entitled to special consideration of your pathetic tantrums. Your guys are an embarassment to the forum.
Thank You,
Mark
- Status
- Not open for further replies.