I'm here
.Maybe what you are wondering about is how did the DF get to what it is. Let's forget about negative DF for the moment and consider only those things that increase the output impedance as seen by the loudspeaker (i.e., those which decrease DF).
The first way is with a passive resistor simply in series with an amplifier that intrinsically has a very high DF. This is easy to do, and is the most benign, but wastes some power.
The second way is to just start out with an amplifier with a naturally low DF. A tube amp or a solid state amp with no negative feedback might be an example of this.
The third way is synthetic output impedance where some kind of feedback sets the output impedance (this is the approach that can give you positive or negative output impedance, and in which no power is necessarily wasted).
In principle, the speaker should not care how the impedance originated. In practice, there may be differences due to execution. The second two approaches, for example, might be accompanied by distortion.
There is another effect that could be in play. That is amplifier mis-behavior. Some amps don't like some loads. I believe that sometimes when people hear amplifier differences, those differences are a result of the amplifiers mis-behaving differently. One example might be how cleanly they clip. Another might be whether the amplifier briefly breaks into a parasitic oscillation on a certain part of the waveform as a result of the amplifier interaction with the load. If you use the first approach to lowering DF, the presence of the resistor isolates slightly the amplifier from the speaker cable and loudspeaker load. In some cases it might reduce misbehavior.
Cheers,
Bob
Hmm,, If using resistor is the most benign, it must be my loudspeaker nonlinearity or something else. I am not using HiEnd class loudspeaker, but some professional class loudspeakers in stack-up.
About that two different type of nice sound, at the first may because loudspeaker nonlinearity is recovered by amplifier and the sound is less distorted (looks like Hi-grade Head phone sound while driven by good opamps).
The second type sound may because it is at lowest mis behavior. If I could reduce mis behavior to its lowest at the first configuration, it will have best loudspeaker linearity at lowest amplifier mis behavior. Is it like this?
Hi bob,
And what is correlated to this mis behavior, it is a bit too general, but it is OK I will try to determine it.
Here some other way to maintain DF that has capability to transfer the input stage characteristics (like tubes) to the output.
About that two different type of nice sound, at the first may because loudspeaker nonlinearity is recovered by amplifier and the sound is less distorted (looks like Hi-grade Head phone sound while driven by good opamps).
The second type sound may because it is at lowest mis behavior. If I could reduce mis behavior to its lowest at the first configuration, it will have best loudspeaker linearity at lowest amplifier mis behavior. Is it like this?
Hi bob,
And what is correlated to this mis behavior, it is a bit too general, but it is OK I will try to determine it.
Here some other way to maintain DF that has capability to transfer the input stage characteristics (like tubes) to the output.
Attachments
There are many ways in which amplifiers can misbehave besides parasitic oscillation bursts, especially when driving difficult loads. For example, they can run out of current (current clipping), their protection circuits can fire or limit current, their clipping behavior can be altered for the worse by current demands from the output stage on the VAS (especially for amps with only a double-EF), etc. I'm sure that there are many more that I have not thought to mention now. Most of these will not show up on common bench testing into resistive dummy loads.
Cheers,
Bob
I might, actually. I've always felt that commercial hi-fi amps are missing usefull adjustments (DF for example). But because i don't like the idea of global NFB I shall need to investigate the issue more thoroughly.
It would also be great to have both tube and the ss implementations of the same concept...
I might, actually. I've always felt that commercial hi-fi amps are missing usefull adjustments (DF for example). But because i don't like the idea of global NFB I shall need to investigate the issue more thoroughly.
It would also be great to have both tube and the ss implementations of the same concept...
Hi Magick,
Bear in mind that if all you want to do is lower DF, you need only put a small resistor in series with the output of the amplifier and waste a little power. This works fine for an SS amp with or without NFB. However, it is rather unlikely that you have a tube amp for which you wish to lower an already low DF.
Cheers,
Bob
Hi Bob,
Thanks for your reply. I know that sometimes I'm too anal for my own good, but I just have to ask... Adding resistor to a signal path adds (thermal?) noise and changes frequency response, as (any) resistor has slightly different resistance at different frequencies.
Resistor is certainly lesser evel of any passive components, but is there any other (active) way of changing DF (without NFB involvement)?
And another thing... If one desires (for some peculiar reason) to reduce DF of his tube amplifier, where should he/she put a resistor in relation to an output transformer? Wouldn't it influence the transformer's performance (operational condition)?
Regards,
Marcus
Thanks for your reply. I know that sometimes I'm too anal for my own good, but I just have to ask... Adding resistor to a signal path adds (thermal?) noise and changes frequency response, as (any) resistor has slightly different resistance at different frequencies.
Resistor is certainly lesser evel of any passive components, but is there any other (active) way of changing DF (without NFB involvement)?And another thing... If one desires (for some peculiar reason) to reduce DF of his tube amplifier, where should he/she put a resistor in relation to an output transformer? Wouldn't it influence the transformer's performance (operational condition)?
Regards,
Marcus
Hi Marcus,
While any component added has the potential of causing some degradation, the resistor value you would add to reduce DF to 20 would be only 0.4 ohms, and you can buy a good one that will dissipate fairly little power and will have negligible inductance.
On a tube amplifier, just add the small resistance in series with the loudspeaker output, just as you would with a SS amp.
Cheers,
Bob
While any component added has the potential of causing some degradation, the resistor value you would add to reduce DF to 20 would be only 0.4 ohms, and you can buy a good one that will dissipate fairly little power and will have negligible inductance.
On a tube amplifier, just add the small resistance in series with the loudspeaker output, just as you would with a SS amp.
Cheers,
Bob
Hi, BOB,
I know that was another difficult things, just like when DUO asking me the other similar things. It will troubling me if I asking the same.
I have another question that may help me and may be DUO too. Is there really two different kind of nice sound or it is only happened in specific condition? The other adjustment with other amplifier topology I cannot find any good sound at all.
If it only one kind, then I will try to do something to make them together, and this sound already very nice, I hoped it will be like real cricket or mug+spoon sound out from my loudspeaker. Actually, I am not wishing that good, current sounding is enough for me.
I might, actually. I've always felt that commercial hi-fi amps are missing usefull adjustments (DF for example). But because i don't like the idea of global NFB I shall need to investigate the issue more thoroughly.
It would also be great to have both tube and the ss implementations of the same concept...
No single commercial amplifier using it, they must have big reasons. If we know it it must be something helpful.
I brought home my H6660 V+I amplifier proto. It is now in my room, and I will start to do something at it this day. I hope I find something new.
Hi DUO,
DF is a good starting, but not for good ending. DF is not the matter after you found the other.
You may ask another than me, why he not placing DF knob at his amps.
Thanks for this thread,
Thanks BOB, but anyway I still not agree with placing resistor in series at the output.
Hi DUO,
DF is a good starting, but not for good ending. DF is not the matter after you found the other.
You may ask another than me, why he not placing DF knob at his amps.
Thanks for this thread,
Thanks BOB, but anyway I still not agree with placing resistor in series at the output.
Last edited:
It seems to be that people just need to try this idea out instead of mulling over the details; especially if they don't understand how it works.
I will be releasing kits within the year and hope that some will try them. Those who have used this amplifier, including myself, find that it has some great benefits over a traditional non-adjustable amplifier.
The effect is very simple, very real, and is extremely easy to understand if you have a backing in electronics design theory. Even if you don't understand the theory, the effect is readily apparent in most systems.
I will be releasing kits within the year and hope that some will try them. Those who have used this amplifier, including myself, find that it has some great benefits over a traditional non-adjustable amplifier.
The effect is very simple, very real, and is extremely easy to understand if you have a backing in electronics design theory. Even if you don't understand the theory, the effect is readily apparent in most systems.
I'm here
Maybe what you are wondering about is how did the DF get to what it is. Let's forget about negative DF for the moment and consider only those things that increase the output impedance as seen by the loudspeaker (i.e., those which decrease DF).
The first way is with a passive resistor simply in series with an amplifier that intrinsically has a very high DF. This is easy to do, and is the most benign, but wastes some power.
The second way is to just start out with an amplifier with a naturally low DF. A tube amp or a solid state amp with no negative feedback might be an example of this.
The third way is synthetic output impedance where some kind of feedback sets the output impedance (this is the approach that can give you positive or negative output impedance, and in which no power is necessarily wasted).
In principle, the speaker should not care how the impedance originated. In practice, there may be differences due to execution. The second two approaches, for example, might be accompanied by distortion.
There is another effect that could be in play. That is amplifier mis-behavior. Some amps don't like some loads. I believe that sometimes when people hear amplifier differences, those differences are a result of the amplifiers mis-behaving differently. One example might be how cleanly they clip. Another might be whether the amplifier briefly breaks into a parasitic oscillation on a certain part of the waveform as a result of the amplifier interaction with the load. If you use the first approach to lowering DF, the presence of the resistor isolates slightly the amplifier from the speaker cable and loudspeaker load. In some cases it might reduce misbehavior.
Cheers,
Bob
The last mentioned effect was often overlooked. Especially the kind of the power buffer stage makes great influence of the whole behaviour of the whole amp in combination of certainly complex loads (darlington follower, Harman's T-circuit/Triple follower, CPF/Sziklai darlington with and without third transistor pair so as various mixtures often to find in two- and three stages quasi complementary topologies like Quad's 303 power output buffer)
independend, whether the output buffer stage is in the global NFB loop or not.
Without CAD supporting it is actually impossible to develope a good sounded amp with output buffer in the global NFB loop, which works stable (without oscillation) over a wide range of various loads at the same time (mainly because the ft's fare more than 10MHz, even by the power BjT's/MOSFET's).
If I use loudspeakers with no complex impedance character, the follow statements are very interesting:
The revolutionary Ultrasound amplifiers
Current driving
Last edited:
Hi, DUO.
Mine almost done. I think, I found something, yes. The sound is lighter and lighter. But I have to compensating the output through misbehave caused by Amp lagging, it is what i told about destructive force. I am not wrong about this, and found this kind of sound after designing my dynamic audio compressor for almost single year. If you make or use any low distortion compressor with a bit fast release time (not too fast). Disgusting sound will appear through release time. I need to fasten the Output stage to remove compensator that I've placed.
How far your progress?
Hi thief,
what is CAD?
And I don't understand why revolutionary still using tubes.
Mine almost done. I think, I found something, yes. The sound is lighter and lighter. But I have to compensating the output through misbehave caused by Amp lagging, it is what i told about destructive force. I am not wrong about this, and found this kind of sound after designing my dynamic audio compressor for almost single year. If you make or use any low distortion compressor with a bit fast release time (not too fast). Disgusting sound will appear through release time. I need to fasten the Output stage to remove compensator that I've placed.
How far your progress?
Hi thief,
what is CAD?
And I don't understand why revolutionary still using tubes.
i think i've stumbled on a direct reading method of measuring output impedance. a while back i was using the method of feeding a 10V sine wave (using the output of another amplifier) into the output of an amp through a 10 ohm resistor. it's ok as long as the output impedance is less than 0.5 ohms (5% error). in this case the error in the measurement is Zo/Rseries. i found a better way to do it. using a Howland current pump fed by an oscillator, i can feed 1 A (peak) AC into the output of the amplifier and read the residual with an oscope. not only can i see the impedance at any frequency, but i can also tell it's phase (and if the phase is 180 degrees, i also know it's a negative impedance).
CAD means "Computer-Aided Design"; software for circuit simulations is e. g. "ORCAD", "LT-Spice" or "circuitmaker".
Contrary to Mr. John Curl, I am convinced that without the support of this software it is not really possible to get a good design for a stable (i. e. oscillation free) and great sounded amplifier. Even not by simple circuits.
Hi tief,
I am using two independent stage(Voltage Follower for output stage) in single global FB (current sense). I cannot reach many advantage from single stage like commonly used.
Hi DUO,
I will build another proto. This one reach its limit. And another proto is in use.
With big difference caused by DF adjustment, placing DF knobs may not a bad idea. You will show people how DF affecting the sound quality. But be sure of so called mis behavior. Mis behavior is too general, why people not specify it too.
I have found something new again. This day, I am using some type loudspeaker, including 4"spk, with and without box, also resistor in series up to 20ohm. And there still something missing.
I am using two independent stage(Voltage Follower for output stage) in single global FB (current sense). I cannot reach many advantage from single stage like commonly used.
Hi DUO,
I will build another proto. This one reach its limit. And another proto is in use.
With big difference caused by DF adjustment, placing DF knobs may not a bad idea. You will show people how DF affecting the sound quality. But be sure of so called mis behavior. Mis behavior is too general, why people not specify it too.
I have found something new again. This day, I am using some type loudspeaker, including 4"spk, with and without box, also resistor in series up to 20ohm. And there still something missing.
with some amp markets, it may be a good idea not to have the knob on the front panel where somebody can mess with it. i'm thinking mostly of PA amps, where the ability to change the DF may be useful to some extent, but have it inside a cover with a panel screw securing it so it doesn't get indiscriminately "tweaked". this way it can be set for the amp to work well or sound good with a particular set of speakers by a band's sound man, and not get it messed with by a roadie or an assistant. having the range to go into the negative impedance region might not be such a good idea except for experimental amps, since the amp can easily become an oscillator.
one thing that might be worth investigating, and it's just a question that came to mind. amplifiers exhibit somewhat similar behavior to an inductance. if the amplifier is operated in negative impedance mode, does it then behave like a capacitance? probably not, but but i think i'll look into it anyway.
one thing that might be worth investigating, and it's just a question that came to mind. amplifiers exhibit somewhat similar behavior to an inductance. if the amplifier is operated in negative impedance mode, does it then behave like a capacitance? probably not, but but i think i'll look into it anyway.
I think the inductive effect is more from the fact that the amplifier getting closer to its gain-bandwidth product has less open loop gain at higher frequencies and this the amount of feedback is dropping accordingly. This means that the feedback's effect on output impedance is going to lift with rising frequency and so up goes output impedance. I'm speaking only of typical amplifiers with lots of open loop gain and a global feedback loop. The effects are different for different topologies.
If you were you make a negative impedance amplifier it would depend again on the characteristics of your particular topology. If this were effected using positive feedback around the amplifier then the same rule applies; closer to the gain-badwidth product means less open loop gain, less feedback, impedance rises (generally). I'm sure there are exceptions to this very simple rule.
If you were you make a negative impedance amplifier it would depend again on the characteristics of your particular topology. If this were effected using positive feedback around the amplifier then the same rule applies; closer to the gain-badwidth product means less open loop gain, less feedback, impedance rises (generally). I'm sure there are exceptions to this very simple rule.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.