I have been reading the forums, and the word "servo" keeps
poping up, I must have fell asleep when that topic was covered,
or I learned it by another name. I have been reading the "op amp
book", I have attached a small quote from chapter 1. Op amps
are very important, and I want to eliminate any confusion I may
have, please explain the servo circuit.
Thank You......
poping up, I must have fell asleep when that topic was covered,
or I learned it by another name. I have been reading the "op amp
book", I have attached a small quote from chapter 1. Op amps
are very important, and I want to eliminate any confusion I may
have, please explain the servo circuit.
Thank You......
Please explain
What is DC output offset please,
Please elaborate, where is it used? I know that DC should
never reach the speakers, how is the "servo" used in a pre amp?
is it an op amp? can you show a circuit?
What is DC output offset please,
Please elaborate, where is it used? I know that DC should
never reach the speakers, how is the "servo" used in a pre amp?
is it an op amp? can you show a circuit?
DC is the average of a signal. it is a constant output. a DC offset is a constant (unchanging) signal present of the output of an amplifier (opamp, power amp, ect).
DC offset is affected by a few things. if you hook the opamp up with input shorted, you'll get some baseline offset. using feedback you can reduce this.
a servo uses an opamp in the feedback path of another amplifier to increase the DC gain of the feedback loop to force this error closer to a zero.
DC offset is affected by a few things. if you hook the opamp up with input shorted, you'll get some baseline offset. using feedback you can reduce this.
a servo uses an opamp in the feedback path of another amplifier to increase the DC gain of the feedback loop to force this error closer to a zero.
ppcblaster said:
A "servo" system is an "automatic control system", using negative feedback.
In EE (electrical engineering), at least back when I was an EE student at Purdue in the second half of the 1970s, it would have been an entire third-year (out of of four years) undergrad course, called something like "Control Systems", or "Automatic Control Theory", covering analysis and synthesis of "classical" (as opposed to "modern") negative-feedback-based automatic control systems.
If you remember anything about Laplace Transforms, Differential Equations, Bode Plots, et al, you shouldn't have too much trouble learning or re-learning all of the essentials of automatic control theory.
Basically, what was covered in the undergrad "control systems" course enabled one to start with the transfer function of an open-loop "plant" and then, after determining a desired overall transfer function, derive the transfer function needed for the feedback path, such that the closed-loop transfer function would become the desired overall transfer function. Of course, many kinds of system-analysis methods, performance criteria, design methods, and lots of other related stuff, were also covered.
The basic block diagram of the simplest form of a feedback control system has an input, then a "summing junction" that subracts the feedback from the input, then the "plant" block with its associated transfer function in the forward path, then the output, from which a feedback path branches back to the "minus" input of the summing junction near the input. In the more-general case, the feedback path also contains a block with its own transfer function.
"Classical" feedback control systems basically take a desired output and use feedback to generate what is essentially an error signal (i.e. the output of the summing junction; the difference between the desired output and the actual output of the plant (but, more generally, the output as processed by the feedback block's transfer function)), which is applied back into the plant's input. In a properly-designed and well-behaved control system, this has the effect of forcing the output to change, in a well-behaved way, to become equal to the desired output.
When using opamps to implement such a feedback control system, a "difference amplifier" opamp circuit could, for example, function as the "summing junction", to subtract the feedback of the output from the desired output. (See the well-known classic application notes AN-31 and AN-20, at http://www.national.com , for straight-to-the-point examples of the basic opamp amplifier topologies.)
But if the feedback path contains anything other than a constant (i.e. pure gain, i.e. "proportional" feedback), such as integration and/or differentiation for example, then often, in opamp-based circuits, the "summing junction" functionality is combined with one of the other operations, if possible, to lower the number of opamps needed for implementation of the system as a real circuit. An example of that is shown in the "DC SERVO" circuit I posted in another thread, very recently (the Chipamps thread named "PA100 with DC SERVO?"), which has now been posted at
http://www.fullnet.com/~tomg/gooteesp.htm
In that DC SERVO example circuit, the integrator opamp used in the feedback path is ALSO performing part of the subtraction of the output from the desired output, and then integrating the resulting "error" signal. Note that the "desired output", in that case, is "ground", i.e. zero volts, since that control system's purpose is to try to force the DC component of the main amplifier's output to always be zero. (And also note that integrating or low-pass filtering an AC signal results in the mean (average) of the signal, which is just its DC component, or "offset".) i.e. If you were viewing the main amp's output on an oscilloscope: The servo loop keeps the average value of the amplifier's audio output waveform centered, vertically, with respect to the horizontal axis (zero volts).
Real circuits can sometimes look quite a bit different than the nice, neat block diagrams of feedback control systems. You might notice, in the DC SERVO example circuit, that the sign of the subtraction of the output from the desired output seems to be reversed, with the desired output being subtracted from the output, instead. It is. But that's because the feedback loop eventually goes to the main amplifier's inverting input (the main/final "summing junction"), reversing the sign again.
---------
Fully understanding opamp amplifier circuits, in general, also necessitates understanding classical feedback control system theory, since virtually all opamps used as amplifiers are used with a negative feedback loop (i.e. from output to inverting input). Using classical feedback control system theory, virtually all aspects of opamp amplifier performance and stability, etc, can be fully and rigorously analyzed, predicted, designed, etc.
- Tom Gootee
http://www.fullnet.com/~tomg/index.html
Anyone listen to the discussion Nelson has in this Burning amp vidio around the 20 min. mark? In it he mentions some very strange findings and he admits they sound crazy, absolutely crazy. That is in some cases applying some dc at the amp's input so you have a substantial dc off set voltage at the output and dc current in the speaker voice coil may lower it's distortion a lot. Ok it sounds crazy but you may want to hear papa discus his findings.
A servo is an integrator built around an opamp. How does it work? The input of the servo normally comes from the amp output. That amp output has the signal ac voltage plus some DC, the offset. You don't want the DC offset.
The servo 'integrates' the ac + DC signal and the output is only the DC (because the integrated signal averages to zero over time).
Then you take that servo DC output, representing the amp output offset, and inject it back into the amp input in such a polarity that it counters the output offset, thus nulling it.
Check out 'opamp integrator'.
Jan
Another way to look at a servo stage is it increases the feedback path gain at very low frequencies (well below audio), and increasing the feedback gain reduces the amplifier's overall closed loop gain at these low frequencies.
Integrator stages are typically used because they are simple and stable in this situation. Typically they only start to have an effect after a couple of seconds. They can only achieve an offset as low as the opamp's input offset voltage plus the IR voltage drop across the high value resistors
used (megaohms) - this means in practice they have to be made using JFET or CMOS input opamps, not bipolar.
Servo stages are an alternative to having an electrolytic cap in the bottom arm of the feedback network, which is a commonly used technique to reduce output offsets in power amps. However this cap can only reduce the amp gain to unity at DC, whereas a servo can reduce the amp gain to zero at DC.
Integrator stages are typically used because they are simple and stable in this situation. Typically they only start to have an effect after a couple of seconds. They can only achieve an offset as low as the opamp's input offset voltage plus the IR voltage drop across the high value resistors
used (megaohms) - this means in practice they have to be made using JFET or CMOS input opamps, not bipolar.
Servo stages are an alternative to having an electrolytic cap in the bottom arm of the feedback network, which is a commonly used technique to reduce output offsets in power amps. However this cap can only reduce the amp gain to unity at DC, whereas a servo can reduce the amp gain to zero at DC.