With regret i discovered that a lot few here deal in really with the DIY. Isn't this the aim of this web site? Or i have perceived error? This means that vast majority or has bought ready appliances what has damage and searches how him it repairs or it deals with theories without it applies something in the practice through them. I place this question for the Zin and wait for answers and opinions. With regard to me, I have already ready my answer as experimented a long time with this subject. Thus I decided to make for a time the opposite from usual. I leave to be exposed first you, and afterwards I comment your opinions in spite of search 1.500.000 posts in all site. Be careful because it is so much critical subject as the output of amplifier. She is the one from the two doors of contacting him with the outside world.
anywhere from 10k on up works just fine for me. it's good to have a predictable value that you can base calculations on. you will probably notice that many amps have a 10k, 20k, 47k, or 100k resistor across their inputs. many years ago, when audio engineering was influenced by telephone system design, there were many amps on the market with 600 ohm inputs, but for the most part, impedance matching has been replaced by voltage matching. most preamps and other line level devices have an output impedance of around 1k, and most devices such as amps have very high input impedance, so there is not a lot of voltage loss in the cables.
10.000 Ohms.
If this is too low for CD-player output (or other intended source)
then I can accept 22.000 Ohms. But not any higher
When comes to using sources with weak outputs,
that require higher input impedance
than 47 kohm
I strongely recommend:
The use of a good pre-amplifier or a good pre-buffer.
lineup
Lineup Audiko Lab
If this is too low for CD-player output (or other intended source)
then I can accept 22.000 Ohms. But not any higher
When comes to using sources with weak outputs,
that require higher input impedance
than 47 kohm
I strongely recommend:
The use of a good pre-amplifier or a good pre-buffer.
lineup
Lineup Audiko Lab
How to calculate Zout?
How to calculate Zout?
I wanted to know what is the output impedance of a common emitor stage without any resistors referenced to ground or rails.
What is the output impedance of collector of VAS stage?
Also current mirrors, what is the output impedance of the collector of current mirror (that is used for differential pair load)?
How to calculate Zout?
I wanted to know what is the output impedance of a common emitor stage without any resistors referenced to ground or rails.
What is the output impedance of collector of VAS stage?
Also current mirrors, what is the output impedance of the collector of current mirror (that is used for differential pair load)?
lineup said:10.000 Ohms.
If this is too low for CD-player output (or other intended source)
then I can accept 22.000 Ohms. But not any higher
This is the advantage of building your own preamp: you can drive whatever amplifier you like. I'm setting the amp input Z between 2k and 3.3 k, which is high enough for an opamp to be able to drive it without too many problems. I like lower Z than usual because some subtle cable variations seem to go away at lower impedances.
Re: How to calculate Zout?
Look on web for a tutorial about H parameters and their use for calculating common emitter features (gain, impedances and so on.
Alternatively look for a tutorial about Giacoletto Equivalent Circuit, which usually introduce to the internal "machinery" of an active device (mainly bipolar devices but, with some change in parameter, this kind of circuit may be useful for any three terminal active device, bipolar, FET, vacuum triode...)
Hi
Piercarlo
lumanauw said:
Look on web for a tutorial about H parameters and their use for calculating common emitter features (gain, impedances and so on.
Alternatively look for a tutorial about Giacoletto Equivalent Circuit, which usually introduce to the internal "machinery" of an active device (mainly bipolar devices but, with some change in parameter, this kind of circuit may be useful for any three terminal active device, bipolar, FET, vacuum triode...)
Hi
Piercarlo
Regarding the wish for not attenuating the input signal, ofcourse you should go for the highest practically possible Zin.
BUT: If you want to ensure perfect transmission of the signal from the source - to the amp , you'd better look into how the professionals do it.
A low Zin (600ohms) gives you some advantages. Of course you need the output of the preamp to be able to cope with it, and you might experience some attenuation, especially if the output impedance of the preamp is correctly matched (not nescesarily 0 ohms). But noise pickup, cable length, and cable quality is now much smaller issues.
If you optimise your outputs and inputs for perfect transmission, you would not be able to hear any difference between standard supplied RCA cables and expensive 'snake-oil' cables.
Are the results of your experiments pointing in this direction Fotios?
BUT: If you want to ensure perfect transmission of the signal from the source - to the amp , you'd better look into how the professionals do it.
A low Zin (600ohms) gives you some advantages. Of course you need the output of the preamp to be able to cope with it, and you might experience some attenuation, especially if the output impedance of the preamp is correctly matched (not nescesarily 0 ohms). But noise pickup, cable length, and cable quality is now much smaller issues.
If you optimise your outputs and inputs for perfect transmission, you would not be able to hear any difference between standard supplied RCA cables and expensive 'snake-oil' cables.
Are the results of your experiments pointing in this direction Fotios?
Johnson or Niquist noise and the Zin
Every amplifier has an output, even with no signal input. An output of this type caused from the noise which reduces the performance of whole amplifier. Especially the noise developed at the input where the signal is smallest from the other stages of amplifier seems to be the main cause for the total noise developed in a multistage amplification circuit. Any unforeseen current that throws from one part of the circuit to another can be referred as noise. These currents caused usually from the distributed inductance or capacitance on the different stages of the amplifier which allows an unwanted coupling between them. An example is the power supply variations which coupled to the input stage if there is not careful shielding. Also a careless design of PCB such as corners on the copper rails of the board can cause unwanted capacitance.
An unwanted inductance can developed due to without reason length and meanders of the copper rails on the board. Certainly these can avoided with a careful design of PCB. But there are other types of noise caused from the passive and active elements of the circuit. These noises are fundamental and owed to the physical structure of the elements. In a resistor, due to thermal energy that contains, there is a continuous motion of electrons. And this motion fluctuates with time developing a small current across the resistor. Of course, if not applied an external voltage at the terminals of this, current finally average to zero. But the variations around this average of current produce unforeseen noise. If the thermal energy increases in the resistor, as a consequence there is an increase in the noise produced. This thermal noise is known as Johnson noise and some times as Niquist noise and described by the equation in the attached file.
k is the Boltzman’s constant, T is the ambient temperature, R is the resistor value and Äf is the frequency bandwidth. Thus careful amplifier design call for minimizing R and limiting the bandwidth Äf only to the needed range. As you understand, this is of big importance so for the choice of amplifier Zin what as for the determination of the closed loop gain. The higher Zin make the amplifier an easy load to driven from whichever preamplifier. From the other hand, some limitation of Zin needed to avoid the increase of Johnson noise in the input. In order to we stop here the theoretical analysis and we pass in more practical subjects, I will report the conclusions from my experiments. An intermediate value of Zin at 50KÙ satisfies also the two requirements. From my trials, even preamplifiers with very low output level (in other words with relatively high Zout) can easily drive such load. From the other hand, the Johnson noise touches just the acceptable limits and this happens in frequencies above 100 KHz.A Zin about 100KÙ while it appears as more ideal, practically does not offer nothing more as load, while on the contrary the Johnson noise is increased. In amplifiers with tubes without fail it will pass unnoticed. In solid state amplifier however, it will decrease the S/N ratio in level that will be heared. A low pass filter around 100 KHz also needed as front end circuit before the differential amplifier input. And a high pass filter around 10Hz needed in the feedback loop. Thus the frequency bandwidth limited from 10 to 100000 Hz. In my amplifier, I have made one important but also controversial intervention as it will be proved by the expected replies – as happened moreover and in my thread “ Zobel network corrections ” before –. I removed the RC low pass filter of 100 KHz from the input, between the collectors of the long tailed pair. And i changed it from parallel to series. I had an idea for this, and it was proved that it functioned well in practice. Curious is that it operates better with a corner frequency of 250 KHz. Without i am sure, i have the impression that it is related with the two resistors, shunt and series, at the bases of long tailed pair. In any case it serves his aim.
Every amplifier has an output, even with no signal input. An output of this type caused from the noise which reduces the performance of whole amplifier. Especially the noise developed at the input where the signal is smallest from the other stages of amplifier seems to be the main cause for the total noise developed in a multistage amplification circuit. Any unforeseen current that throws from one part of the circuit to another can be referred as noise. These currents caused usually from the distributed inductance or capacitance on the different stages of the amplifier which allows an unwanted coupling between them. An example is the power supply variations which coupled to the input stage if there is not careful shielding. Also a careless design of PCB such as corners on the copper rails of the board can cause unwanted capacitance.
An unwanted inductance can developed due to without reason length and meanders of the copper rails on the board. Certainly these can avoided with a careful design of PCB. But there are other types of noise caused from the passive and active elements of the circuit. These noises are fundamental and owed to the physical structure of the elements. In a resistor, due to thermal energy that contains, there is a continuous motion of electrons. And this motion fluctuates with time developing a small current across the resistor. Of course, if not applied an external voltage at the terminals of this, current finally average to zero. But the variations around this average of current produce unforeseen noise. If the thermal energy increases in the resistor, as a consequence there is an increase in the noise produced. This thermal noise is known as Johnson noise and some times as Niquist noise and described by the equation in the attached file.
k is the Boltzman’s constant, T is the ambient temperature, R is the resistor value and Äf is the frequency bandwidth. Thus careful amplifier design call for minimizing R and limiting the bandwidth Äf only to the needed range. As you understand, this is of big importance so for the choice of amplifier Zin what as for the determination of the closed loop gain. The higher Zin make the amplifier an easy load to driven from whichever preamplifier. From the other hand, some limitation of Zin needed to avoid the increase of Johnson noise in the input. In order to we stop here the theoretical analysis and we pass in more practical subjects, I will report the conclusions from my experiments. An intermediate value of Zin at 50KÙ satisfies also the two requirements. From my trials, even preamplifiers with very low output level (in other words with relatively high Zout) can easily drive such load. From the other hand, the Johnson noise touches just the acceptable limits and this happens in frequencies above 100 KHz.A Zin about 100KÙ while it appears as more ideal, practically does not offer nothing more as load, while on the contrary the Johnson noise is increased. In amplifiers with tubes without fail it will pass unnoticed. In solid state amplifier however, it will decrease the S/N ratio in level that will be heared. A low pass filter around 100 KHz also needed as front end circuit before the differential amplifier input. And a high pass filter around 10Hz needed in the feedback loop. Thus the frequency bandwidth limited from 10 to 100000 Hz. In my amplifier, I have made one important but also controversial intervention as it will be proved by the expected replies – as happened moreover and in my thread “ Zobel network corrections ” before –. I removed the RC low pass filter of 100 KHz from the input, between the collectors of the long tailed pair. And i changed it from parallel to series. I had an idea for this, and it was proved that it functioned well in practice. Curious is that it operates better with a corner frequency of 250 KHz. Without i am sure, i have the impression that it is related with the two resistors, shunt and series, at the bases of long tailed pair. In any case it serves his aim.
Attachments
Hi fotios,
Yes, in an idealized sense. Practically speaking, most consumer products are not designed well enough to take advantage and will not come close to an ideal noise performance. Tubes have other issues again. They also have some advantages.
What it all comes down to is a series of tradeoffs in equipment design. You are in a position to optimize the parameters you deem most important - that's good. I would only question someone who blindly optimizes one parameter without considering the others.
You can design an amplifier with a 1 Meg input impedance that is driven with a 1 ohm source impedance. This will "short" your johnson noise to ground. Let's not get into cables please!
-Chris
Yes, in an idealized sense. Practically speaking, most consumer products are not designed well enough to take advantage and will not come close to an ideal noise performance. Tubes have other issues again. They also have some advantages.
What it all comes down to is a series of tradeoffs in equipment design. You are in a position to optimize the parameters you deem most important - that's good. I would only question someone who blindly optimizes one parameter without considering the others.
You can design an amplifier with a 1 Meg input impedance that is driven with a 1 ohm source impedance. This will "short" your johnson noise to ground. Let's not get into cables please!
-Chris
lumanauw said:Hi, Piercarlo,
ThanksYou got URL that I can read on "output impedance"? I've googled, but still not found it.
Back to the original question. DartZeel (based on the papers on his website) seems uses no Rg resistor on its input.
If you can't find anything on web, the topic is just taught in introductory courses of analog electronic. Pheraps finding a college textbook on analog electronic may be more useful than googling. IMHO.
For your quesition: Rg resistor, evident of hidden exist always in all practical circuits.
H
Piercarlo
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.