This is 15V zener. On cold start 15,2V, when thermally stabilized 14,8V, resulting in thermal compensation of input injection current. Of course BF and TL must be thermally coupled.
Agree completely, that's why an unfiltered amp must have as high bandwidth as possible, meaning it is capable to settle its intrinsic instability (fight as you've mentioned) much faster, at least 100 times faster than the audio bandwith, thus not interfering with audio signals. So in a way it has to be 100 times "smarter" than a complexity of a problem it is dealing with.
Regards, Andrej
Let us explain this for some readers.
If an amp using counter reaction is not fast enough in open loop (means by itself... its slew rate) to follow fast transient signals, the signal will not be the same in his output.
A square wave will have less vertical edges and rounded corners.
There is a point in the input stage where the CR signal coming from output is subtracted to the original input signal, and the result is an error signal. During this transients, this error signal will have a both high levels and fast slopes. In this situation, this error signal will overdrive the active device and the result will be very bad distortion.
Trying to set-up very fast amps is not a matter of nice numbers, but the only way to reduce those distortions, unless you avoid global feedback.
That is the reason why global feedback has a bad reputation near audiophiles.
But this is unfair. C.R. brings a lot of advantages: It increase the bandwidth and reduce the distortion.... as long as the amp is fast enough.
In that matter, so called current feedback topology gives a lot of advantages, comparing to voltages feedback. In classical voltage feedback amps, we use a differential input stage, and the output CR signal is applied to a different transistor before they are compared. This transistor will add its own pole (bandwidth limitation, phase turns) and distortion. On the contrary, Current feedback topology apply the CR signal to the same input device. (the first input transistor). That is the reason why we can manage *with the same components* a 5X speed factor (bandwidth and slew rate) using current feedback topology instead of voltage one. And why the result will be a more transparent and smoother amp.
And removing a pole will increase stability.
As this speed race is not for the beauty of a bandwidth we do not care about -as long we are dealing with audio signal- but a search for low distortion, it will always be a good idea to add a passive filtering to the input stage, before the input signal is applied at the first active device.
This will reduce all unwanted transient signals, as well as parasitic Hf signals, reducing further this transient distortion. On a voltage feedback amp, where this filtering is most important, we are very limited, as the bandwidth is limited, and obliged to set a low pass filter, if we want to keep the bandwidth so low (far enough from the amp capability) that it will bring phase turns and level decrease at the highest frequencies. With current feedback amp, we can set this low pass filter high enough to have a flat response curve at 20 000 with a very little phase turn.
Better to try, with your own loudspeakers, to set this low pass frequency as low as possible, in order to keep the input signal very slow comparing th the amp capacity, to keep the transient distortion as low as possible.
On an educational point of view, i had compared a voltage feedback amp with *the same one*, SSA modified. The result was amazing. -30db in distortion. > X5 in both bandwidth and slew rate. and my ears can feel the difference.
Who want to deal any more with voltage feedback amps ?
Last edited:
I persist to say SSA is a "so called" current feedback.
This unappropriated name was given by Analog device, but used as generic, now.
An other way to build a current feedback amp is to use a preamp, send the CR to the output of it, and take the signal to the next stage in sensing the current feeding the amp across a current mirror, but the idea is the same.
Here, we use the input transitor in two topologies in the same time, emitter common for the signal, base common for the CR. We can consider emiter as the output of the first preamp, and collector's resistance as the current mirror....
Classical.
What is not so classical (but yet published) is the way the CR is divided in two paths, compensating bu itself the emitters offsets, and allowing the base of the input transistor to be polarized at the ground level, allowing direct coupling. Clever and elegant.
Where is my mistake ? ;-).
The name in itself has no sense, referring to the ohm law: Where is current, there is voltage, indeed.
Last edited:
Maybe its just semantics between us. Sampling voltage on first control level is voltage feedback I guess for classification. But me not being a specialist I would wait for the amp's author Andrej or more of you well versed in theoretical electronics guys to ''title'' the topology officially. Happy to be wrong myself.
I do not know why, but ir is easier for our human brains to think in voltages. May-be because we are learned this strong trio V=RI that way ?
When we are dealing with transistor, it is better to thing "current".
Yes, it is semantics, and this "current feedback" appellation was not the best idea from A.D.
I believe it is because they used, sometimes, this kind of funny architecture:
http://www.analog.com/library/analogDialogue/Anniversary/Graphics/figure95lg.gif.
Esperado you did it completely, all the way to the last detail, in theory and in practice, you're our man, we need people like you on this forum to acquaint masses with positive example. Your explanation, better said theoretic elaborate about speed, bandwidth, related distortions, current feedback and transparency, many times so overlooked but to mine opinion most important amp's features, is exemplary and I am sure many will be enlightened by your contribution.
The answer: people who up to this day didn't had opportunity to listen to a good current feedback amplifier, after experienced them, nothing will be ever the same, no way back syndrom.
Last edited:
I agree with Esp completely.
A current fb amp is the way to go. I use the concept too in all my NAKSAs, and the FetZilla is cfb.
The high speed, however, must be carefully managed, but OTOH the phase shift is low, so in the case of a fet output stage the pole created by the gate stopper and the input capacitance will bring the amp to stability generally without lag compensation.
CFB opamps have been available for a very long time and display similar performance, often used in video applications.
Excellent design, Andrej.
Hugh
A current fb amp is the way to go. I use the concept too in all my NAKSAs, and the FetZilla is cfb.
The high speed, however, must be carefully managed, but OTOH the phase shift is low, so in the case of a fet output stage the pole created by the gate stopper and the input capacitance will bring the amp to stability generally without lag compensation.
CFB opamps have been available for a very long time and display similar performance, often used in video applications.
Excellent design, Andrej.
Hugh
Educational scientific literature uses well deserved name current feedback amplifiers, certainly not just to distinguish it from more common voltage feedback, but definitely for a good number of reasons. Here's one very straightforward explanation according to Mr. Erik Barnes, Analog Devices engineer:
"Before looking at any circuits, let’s define voltage feedback, current feedback, and transimpedance amplifier. Voltage feedback, as the name implies, refers to a closed-loop configuration in which the error signal is in the form of a voltage. Traditional op amps use voltage feedback, that is, their inputs will respond to voltage changes and produce a corresponding output voltage. Current feedback refers to any closed-loop configuration in which the error signal used for feedback is in the form of a current. A current feedback op amp responds to an error current at one of its input terminals, rather than an error voltage, and produces a corresponding output voltage. Notice that both open-loop architectures achieve the same closed-loop result: zero differential input voltage, and zero input current. The ideal voltage feedback amplifier has high-impedance inputs, resulting in zero input current, and uses voltage feedback to maintain zero input voltage. Conversely, the current feedback op amp has a low impedance input, resulting in zero input voltage, and uses current feedback to maintain zero input current.
Current feedback amplifiers have excellent slew-rate capabilities. While it is possible to design a voltage-feedback amplifier with high slew rate, the current-feedback architecture is inherently faster. A traditional voltage-feedback amplifier, lightly loaded, has a slew rate limited by the current available to charge and discharge the internal compensation capacitance. When the input is subjected to a large transient, the input stage will saturate and only its tail current is available to charge or discharge the compensation node. With a current-feedback amplifier, the low-impedance input allows higher transient currents to flow into the amplifier as needed. The internal current mirrors convey this input current to the compensation node, allowing fast charging and discharging-theoretically, in proportion to input step size. A faster slew rate will result in a quicker rise time, lower slew-induced distortion and nonlinearity, and a wider large-signal frequency response."
At SSA we have splitted current feedback in a bridge configuration where differential gain devices are positioned in a middle.
Last edited:
Salas,
Here's an even simpler explanation.
On a working bipolar transistor, or a jfet or fet, the voltage at the emitter/source is determined solely by the voltage at the base/gate.
Therefore any input to the emitter/source can only be in the form of a current, since the voltage at that point is already dictated by the voltage at the base/gate.
So, if we take the voltage source at the output of a CFB amplifier, and connect it by a resistor to the emitter or source of the fb node, then that resistor becomes a transconductance device since it converts the voltage at the output end to a current at the fb node.
You are right that the fb starts out as voltage feedback. But the series fb resistor converts that voltage to a current - there is no option, as the voltage at the left, fb node side of the converting resistor is determined by the voltage at the base/gate of the fb device, as determined by the input signal in this case.
Clear as mud?
Hugh
Here's an even simpler explanation.
On a working bipolar transistor, or a jfet or fet, the voltage at the emitter/source is determined solely by the voltage at the base/gate.
Therefore any input to the emitter/source can only be in the form of a current, since the voltage at that point is already dictated by the voltage at the base/gate.
So, if we take the voltage source at the output of a CFB amplifier, and connect it by a resistor to the emitter or source of the fb node, then that resistor becomes a transconductance device since it converts the voltage at the output end to a current at the fb node.
You are right that the fb starts out as voltage feedback. But the series fb resistor converts that voltage to a current - there is no option, as the voltage at the left, fb node side of the converting resistor is determined by the voltage at the base/gate of the fb device, as determined by the input signal in this case.
Clear as mud?
Hugh
Practically yes but, hehe there's always some but, things are not that simple. The global CFB input node has to accept certain level of current resulting in a corresponding gain, meaning the ratio of a current change on that chatode resistor has to have certain voltage drop on resistor in anode connection to produce an adequate gain factor. Also current feedback should be DC coupled to the output via resistor, no serial caps alowed. Otherwise you will just get some sort of compensation to the local feedback not the global as it should be.
''Conversely, the current feedback op amp has a low impedance input, resulting in zero input voltage, and uses current feedback to maintain zero input current."
At SSA we have splitted current feedback in a bridge configuration where differential gain devices are positioned in a middle.
Hi Andrej (and others), thanks for the explanations. Does SSA have a low impedance input though?
Also current feedback should be DC coupled to the output via resistor, no serial caps alowed. Otherwise you will just get some sort of compensation to the local feedback not the global as it should be.
No series cap, that tube amp shows 235kHz -3dB at 1W on transformer's output.
Yes, current feedback input is at emitter pin, where low voltage change (zero voltage) transforms in a large current change, thus the impedance present at this input is low. On contrary, modulating signal input is at base pin, where low current change (zero current) results in a large voltage change, thus the impedance present at this input is high.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.