how to make a unit gain amplifier ???

[abishek007]

unit gain

Hi All

i am working on my masters project on Phase lock loop . I have to make a unit gain amplifier to include it in my charge pump design . i was wondering how to
make it with few complications . also can i use 2 inverters connected in series ???
I will appreciate for your replies

 

[FvM]

unit gain amplifier

What do you mean with inverter in this regard? A digital gate from two complementary MOSFET? Cause it has a gain much higher than unity, I don't see how this shall work. A transistor in common drain circuit as voltage follower, possibly with a current source as active load, would be more suitable,

 

[abishek007]

how to make an amplifier

Yeah i was talking about the complementary Mosfets . i main aim is to keep 2
different nodes at the same voltage . books suggest to use unit gain amp but i was thinking since i have 2 keep 2 nodes at same voltage i was thinking that i can use 2 inverters in series ( so that there is no voltage inversion ) can i still do that ?

 

[FvM]

make an amplifier

i was thinking that i can use 2 inverters in series ( so that there is no voltage inversion ) can i still do that ?

No inversion in fact, but a high voltage gain. It doesn't work.

 

[abishek007]

how to make a low gain amplifier

so i need to use the voltage follower circuit
thanks

 

[zhouyao2003]

how to make amplifier

Voltage follower is different to 2 stage inverter!!!

 

[antonio_eda]

make amplifier

I hope the author solved the problem of a voltage follower... it is not so difficult

I am looking for another follower: I want it to have a tunable gain as well as a tunable output resistance. And it have to be as simple as possible. Is it possible to design such a circuit or not?

If it is impossible to make a simple and fully tunable follower, how to make a simplest follower with strictly constant output resistance?

 

[LvW]

how to build gain-amplifiers

I hope the author solved the problem of a voltage follower... it is not so difficult
I am looking for another follower: I want it to have a tunable gain as well as a tunable output resistance. And it have to be as simple as possible. Is it possible to design such a circuit or not?
If it is impossible to make a simple and fully tunable follower, how to make a simplest follower with strictly constant output resistance?

Your description sounds a bit contradictory:
- You are looking for a "follower with tunable gain". Normally, the term "follower" is used only for a unity gain amplifier.

- What do you mean with "strictly constant output resistance" ? Completely independent on frequency ?

 

[antonio_eda]

how can i make amplifier

Hmm... I need the gain to be controlled from about 0.90 to 0.999V/V. Maybe a term "follower" is not as good as the unity gain amplifier. It is possible to convert some circuits with opamps into functionally the same circuits, but with close-to-unity gain amplifiers...

I do not know if the second problem is more difficult. I wanted a constant output resistance --- only the real part of the impedance have to be constant. In each simple MOSFET configuration the output resistance depends on the operation point. Even the CS stage with "diode" load... and I am dreaming about the OP independent output resistance.

 

[FvM]

how to build a follower amplifier

I need the gain to be controlled from about 0.90 to 0.999V/V

What's the meaning of controlled in this case? Controlled by what?

I wanted a constant output resistance --- only the real part of the impedance have to be constant.

The specification is rather vague without telling the frequency range of interest, it may be kHz or Ghz as well.

Generally real output impedance amplifiers are designed by adding a series resistor to a low output impedance buffer. Depending on the frequency range and the required precision and linearity, either open loop (voltage follower) or closed loop (unity gain OP buffer) amplifiers are used. This is the usual method to drive e.g. impedance matched analog signal cables in wide band instrumentation.

Closed loop buffers have basically an inductive output impedance (magnitude increasing with frequency) below it's bandwidth limit.

 

[antonio_eda]

how to make voltage amplifier

Thanks for your reply. The gain have to be controlled by any external signal, for example a voltage or a current. I have tried to use some simple MOSFET stages, but there is a problem of the output resistance. When the gain is changing then the output resistance is moving too. The output resistance is not independent on the value of the gain --- for example in a simple CS stage.

The output resistance have to be about 2-200kOhm. Ideally, independent on the gain or at least constant --- in the range of 100MHz.

 

[FvM]

I can't understand this specification. 200 k real impedance @ 100 MHz involves a fF range output capacitance. This can't be achieved even on a chip level. Off chip, real impedances above a few 100 ohms aren't feasible to my opinion.

 

[antonio_eda]

200 k real impedance @ 100 MHz involves a fF range output capacitance.

What does it mean exactly?
As far as I know the impedance is a complex value, and the resistance is its real part independent on frequency...

 

[FvM]

When you say 200k output impedance, I assume, that you mean, it should be almost real in the frequency range of interest. I simply mentioned, that this is impossible to my opinion.

If my assumption is wrong, you may want to specify a complex impedance you want to achieve.

P.S.: The complex impedance equivalent circuit involves a series circuit of a capacitor and a resistor. To take an additional pin capacitance into consideration, it's rather meaningful to use an admittance parallel equivalent circuit. Unfortunately, 200k in parallel to 1 pF @ 100 MHz turns into an impedance of 1 pF + 13 ohms. No chance to raise the impedance real part above this level.

 

[antonio_eda]

Our discussion is too complicted, while the problem of the output resistance of the amplifier is away.

Forget about the frequency range and about the output impedance. Let's take into consideration only a DC and small-signal conditions.

And now imagine the amplifier with a tunable gain. You have the input and the output and one control input for changing the gain. A simple CS stage with a current mirror as a load may be an example.

When you change the value of the load, the gain changes as well as the output small-signal resistance. And I need them to be independent as far as possible

 

[FvM]

It seems to me, that I'm far from understanding your intentions with the cicrcuit.
I'm used to think about electronic circuits in terms of real world applications. I don't see it yet in this case.

 

[antonio_eda]

Real world is important, for me too. But try to stay at the possibly simplest level of the circuit theory. Let it be just a theoretical simplification.

If you have any suggestion about an Mosfet amplifier, which gain can be tuned without any impact on its output resistance, I'll be happy.

 

[LvW]

Real world is important, for me too. But try to stay at the possibly simplest level of the circuit theory. Let it be just a theoretical simplification.
If you have any suggestion about an Mosfet amplifier, which gain can be tuned without any impact on its output resistance, I'll be happy.

I am pretty sure that the only way is to have at least two stages: one for the gain (adjustable) and the second one for providing a constant output resistance.

 

[FvM]

A variable gain amplifier with a sufficient low impedance load resistor can be considered as an approximation
of the requested design. It's also a traditional solution for wideband VGA and multipliers, e.g. with Rload = 50 ohm.