To amplify DC Voltage wou will need a DC/DC boost converter, a circuit implemented with a power electronic switch. More on boost converters you can find at http://www.eecs.berkeley.edu/~tkg/EE100_42_F09/Documents/Laboratory/Lab5/lab5total.pdf

IMO the question is not about an electronic amplifer but sooner about an electromechanical one...

just FYI, an :amplifier" needs an INPUT, and an OUTPUT. In the case of the generator, where would the input be applied to ?

So, if we vary the excitation voltage as an input voltage and take the generated voltage as an output voltage, we can use the DC electromechanical generator as an amplifier...

What would the "amplification ratio" be ?

I don't know what the "amplification ratio" would be; I only suppose it can be more than one. Obviously, it depends on the number of the turns of the excitation and rotor windings, and also on the speed of the rotor...

I imagine the typical arrangement consists of two coupled DC machines - an electrical motor and a generator. The motor is externally supplied by an electrical power supply. But it is interesting that the generator can be driven by any mechanical motor; in this case, the power supply can be whatever - mechanical, thermal, chemical... even manpower:)

I have the feeling that such electromechanical amplifiers are (were) widely used in military installations. I don't know what the English name of this electromechanical amplifier is; I suppose the Bulgarian name is "електромашинен усилвател" :)

Now, here is an interesting concept: When I say, AMPLIFICATION RATIO, it doesn't have to be one-on-one, it could be a SMALL SIGNAL AMPLIFICATION RATIO, Let me elaborate:

A 1V could generate 2V at the output, 2 2V generates 4V. This is clearly an amplification of 2x (i.e., GAIN=2).

Instead, consider this example, a 1.01V generates 2.02 V at the output, but, 1.02V generates 2.04V, 1.03V generates 2.06V ... So, the DC bias at the output is 2.00V, and the input and output are related by Vo = 2.00 + 2*(Vi-1.00).

So, I still have an amplification of 2 ... In other words, (dVo/dv) / (dVi/dV) = 2.00

This could also create an amplification. This is identical to biasing a BJT transistor, and lettting the microphone input vary the output by 2x that input amount at the output. The AMPLIFICATION is 2x.

So, in military etc ... this concept can be used to employe an existing generator to do SMALL SIGNAL amplification ?

Interesting thoughts causing more questions... First, I wonder if this is the same sort of amplification as in the case of an electronic (e.g. a transistor) amplifier... It seems here we do not control the power supply energy by dissipating a part of it as the transistor does... Then how the supply energy is controlled?

Tolga, I agree with you that there is a need of a biasing... Just an idea - to bias by a permanent magnetic field...

Cyril, in military applications, do they really care about "energy consumption" ? The priority is something much more different ... I remember, vacuum tubes being considered "military-friendly" a while ago, since they were RADIATION RESISTANT ! So, here, is the priority really AMPLIFICATION WHILE STAYING ENERGY EFFICIENT ? or, more like AMPLIFICATION BY USING A DC GENERATOR ?

This "DC generator" is a "voltage-controlled voltage source" and somehow the input voltage (current) source controls (through the excitation coil) its voltage... and the question is, "How?"

IMO when the input quantity varies, the generator loading (of the input motor) varies as well; thus the generator absorbs precisely the needed power from the motor without any losses...

Generator converts mechanical energy to electrical while in amplifier input and out put both are electrical, by field control voltage can increased but it cannot consider as amplification.

Here the input (the excitation voltage) and the output (the generated voltage) are also electrical quantities...

Here is what Wikipedia says about such an arrangement:

https://en.wikipedia.org/wiki/Motor%E2%80%93generator

Bhupendra, your thought is good, in that, merging it with Cyril's point, you could actually have "tow different" inputs to a motor, MECHANICAL, and ELECTRICAL ... One with the excitation voltage, and one mechanical. The output could actually be two things too ... electrical and mechanical (i.e., motor is turning something) ... makes sense ?

And how do they operate?

well, here are links to related Wikipedia articles:

https://en.wikipedia.org/wiki/Amplidyne

https://en.wikipedia.org/wiki/Metadyne

yes

IGBT's

Just as we treat all the Phasors to rotate at Synchronous speed(Synchronous Frame of Reference),treating both the Excitation Voltage and Induced Emf of the D.C. Generator to be at the same Rotating Frame of Reference at the same speed,We can treat D.C. Generator also as an Amplifier.This also fits into the Amplifier Load Line Treatment,including Saturation.D.C. Voltage and Current also can be represented as Phasors,in a Frame rotating at a Variable Speeds instead of at Synchronous Speed.

P.S.

I had already told (though I deleted the message) earlier that rotating machines as amplifiers-Amplidyne and Metadyne were in use even prior to advent of Mercury arc amplifier and solid state devices. These machine were integral part of control system and it seems that basics of control system theory - comparison of two signals would have derived from rotating machine knowledge.

We have worked on those Machines long back in our Machines Lab. at JNTU College (Then Govt.College) of Engineering,Anantapur.

P.S.

Rakesh and Pisupati Sir,

I am talking about the mechanical amplification of the DC voltage within the machine, not by using any separate amplifier device along with the machine.

Read Performance & Design of Direct Current machine by Clayton & Hancock.

You can control the developed voltage at design level.