Good morning so today we will start millimetre

wave activity devices Now different applications they need different types of requirements

For example one satellite let us say geo stationary satellite it is 36000 kilometres above earth

surface so obviously it will need very high power to communicate Now mobile phone it needs

maybe milliwatts that should be sufficient but for mobile base stations it is a few watts

so different applications have different types of requirements Now for any millimetre wave

system we need millimetre wave sources and for any amplifier design we also need different

types of amplifiers power amplifiers low noise amplifiers oscillators switches so all involved

active devices So we will see different types of popular active devices how they operate

what are their characteristics so let us start with the basic chart So in general microwave devices they can be

divided into 2 parts solid state devices and microwave tube based devices Again solid state

devices they can be categorised according to their working principles into 4 major categories

transistor field effect transistor transferred electron devices and avalanche transit time

devices So under transistor we have bipolar junction transistor HBT tunnel diode et cetera

Under field effect transistor we have JFET MOSFET MESFET HEMT and different types of

memories and CCDs Under transferred electron devices we have gunned diode LSA diode Indium

Phosphide diode then (diff) different other types of drives and under avalanche transit

time devices we have READ diodes IMPATT diode TRAPATT diode BARITT diode so they are categorised

according to their working principles So among all different types of active devices

popular are HEMT (High Electron Mobility Transistor) Gunn diode and the IMPATT diode So in todays

discussion we are going to discuss about these popular active devices and under now microwave

tubes we have 2 different types of tube linear beam tubes or sometimes we call simply O type

tubes and Crossed field tube or sometimes also called M type tube So we are not going

to discuss about the tubes so mainly we will concentrate on active devices solid state

devices So here is one chart which relates the average

power versus frequency so if we look at this chart where the this comparison uhh we are

doing among different types of tubes and different types of solid state devices are being considered

here So if we look at this chart for any given device if we increase frequency maximum available

power it decreases And not only that the high frequency parts and high power together these

are mainly dominated by tubes but the problem with tubes is that they are expensive and

bulky so we try to avoid them in low cost uhh applications And now looking at the different

solid state devices we have again solid state devices based in gallium arsenide technology

gallium nitride technology and uhh also different examples are HEMT BJT so again for these solid

state devices also power handling (capa) capacity it decreases with increasing frequency And if you look at this plot not only that

so gallium nitride usually it provides high power and gallium arsenide based devices they

are used at much higher frequencies So the stop curve it belongs to Gyrotron but the

problem with Gyrotron is its size and cost For example Gyrotron is shown here it occupies

almost one building so you can compare the size one with 1 human being Now Microwave power requirements so in this

chart we see that different applications they have different types of power requirement

and they correspond to different frequency bands So right now at millimetre wave frequencies

mainly we have defence and space locations so consumer applications is very much limited

but with the incoming 5G Wireless application it might be at millimetre wave frequencies

may be at 60 gigahertz so it is expected that millimetre wave activity devices will be used

for this type of communication Now for solid state devices gallium nitride based device

it can provide more power compared to other solid state devices and this is the theoretical

limit for gallium nitride devices and the 2nd line it shows the theoretical limit for

gallium arsenide devices So we will see which parameter determines high frequency operation

and uhh its maximum power capacity Now here are some more examples of microwave

and millimetre wave power uhh sources So in most of these applications till now we have

many different types of tubes they are being continuously used but the disadvantage of

these tubes they are of large size they are bulky usually they operate at fixed frequency

so frequency tune ability it becomes a problem with them And not only that they are usually

narrowband and it has also spurious spectrum with that so for example our in our today

use we use microwave oven usually it uses a magnetron to generate microwave power at

frequency 2 point 45 gigahertz Now in this plot I am showing you microwave spectrum of

1 magnetron so you can see the pic it is here around 16 point 5 gigahertz but it is associated

with many sidebands So we have many spurious bands so the quality of wave form spectrum

it is very poor and another disadvantage is its high cost So then going back to solid state device so

2 points become very important one is power handling and another one is high frequency

operation Now if we compare different types of solid state devices which are already available

in market so here is the chart So in the 1st figure we are showing the output power versus

frequency so if we increase frequency maximum output power from the device it decreases

and again we can see that gallium nitride based HEMPT it provides maximum power whereas

gallium arsenide based PHEMPT it provides high frequency operation We also have indium

phosphide HEMPT it also goes beyond 100 gigahertz Now another important parameter is power density

considering component miniaturisation so how much power is available per square millimetre

versus frequency Again gallium nitride based devices are the

winner but their frequencies are much limited to let us say 30 to 40 gigahertz So if we

want to use them at higher millimetre wave frequencies then again we have to go for gallium

arsenide based devices but gallium arsenide based devices it has again one problem compared

to silicon based devices that it consumes much power so we will discuss this point later

in detail So some of the popular devices their applications

and frequency limitations so for example IMPATT device they are typically used below 300 gigahertz

so it can cover the whole millimetre wave spectrum starting from 30 to 300 gigahertz

This so this frequency whatever is mentioned here uhh it is it represents the popular applications

so but there are examples where IMPATT diode has been used above 300 gigahertz and physical

substrate used for IMPATT diode fabrication are silicon gallium arsenide indium phosphide

and they are popular for transmitter amplifiers So in high power amplifiers where we need

power amplifier we can use IMPATT diode but they are not much popular as a source millimetre

wave source because they are associated with phase noise we will see later Next is Gunn diode typically used below 180

gigahertz substrate gallium arsenide and indium phosphide It is a very popular millimetre

wave source which we use in laboratory experiments in the universities so it is popular in local

oscillators and also it can be used it transmitter amplifiers Next FET and HEMT typically used

below 140 gigahertz again based on gallium arsenide and indium phosphide and they are

widely used in different types of amplifiers oscillators switches mixers mixers phase shifters

so different types of applications Next is p i n diode typical frequency below

100 gigahertz and the materials used for fabrication silicon gallium arsenide p i n diode is mainly

popular in switching applications and at millimetre wave frequencies sometime they are also used

as variable resistor Next Varacter and it can give variable capacitor so where we need

any tunable component we can use varactor and we can electronically tune the capacitance

of a varactor diode Typical applications multipliers tuning phase shifters and different types

of modulators Now how to choose the material for any given

solid state devices There are several parameters when we go for millimetre wave frequency applications

the frequency is so high that the signal is changing very fast So device it should be

narrow enough so so that carriers from left side to right side it takes minimum time typically

less than the time period of the given signal so we characterise it by transit time So then

the 1st conclusion is the device size should be small whatever carriers we are using here

so carrier velocity should be very high Now carrier velocity inside the substrate material

it depends on many parameter it depends on electric field it depends on mobility electron

mobility and hole mobility so electron mobility and hole utility it is a it depends on the

(sub) type of substrate So depending upon our application requirement we can choose

a proper substrate for fabricating the solid state devices So here are some examples gallium arsenide

substrates they are used because of its high mobility Silicon substrate the fabrication

procedure uhh it is very low cost and also high yield so for consumer market silicon

substrate is very popular Gallium nitride substrate it is mainly used for high power

applications but their high frequency application is limited typically they are used below 30

or 40 gigahertz Now we see the characteristics of some popular semiconductor materials so

we are comparing the band gap energy and mobility at room temperature 300 Kelvin So if we look at the silicon it has a band

gap of 1 point 12 electron volt at room temperature whereas for gallium arsenide it is 1 point

43 so the band gap value is higher than means its power handling capacity will be higher

Now look at the mobility values so if I look at the electron mobility silicon electron

mobility is 1600 centimetres square per Volt second whereas for gallium arsenide 8500 So

obviously for gallium arsenide based devices electron mobility will be much higher under

a given electric field and we can increase (())(15:24) of the device we can go for high

frequency operations with gallium arsenide based devices So some important parameters how we characterise

any solid state device so for example output power or what is the maximum power available

from the device power density what is the maximum frequency of operation then Power

added efficiency so these are the mostly well used parameters So output power P max it is

proportional to V max multiplied by I max where V max it represents the breakdown voltage

and I max it depends on how fast we can remove heat from the device also gate width and length

because the resistance depends on it Next is Power density power density is equal to

V max current density so V max is the breakdown voltage and current density it is limited

by the band gap and thermal conductivity Next high frequency operation F max it is

proportional to V s by L where V s it is the saturated carriers velocity and L is the gate

length for a given let us say gate and P max it is proportional to 1 by F square so if

we increase the frequency it is expected that the power will decrease very fast so what

we seen in the previous plot Next is Power added efficiency so here what we do let us

say any given solid state device we are using in amplifier applications so we will be having

RF good power and then the amplified RF output power And to amplify the signal we have to

apply some energy to the device and this is being done by DC source Now how much power it will absorb from the

DC source and what would be the efficiency of the device we call it the power added efficiency

and it is defined as 100 multiplied by output RF power minus input RF power divided by P

DC total total DC power what it consumes So it depends on many parameters such as wave

shape what is the impedance of the device then what is the leakage current and power

gain of the device Now let us start with the 1st device Bipolar

Junction Transistor so bipolar junction transistor also we used in basic electronics lab It has

very similar principle and only thing is that (in) whenever we use for low frequency applications

let us say at megahertz frequency we do not use different capacitances offered by PN junctions

We have two PN junctions in BJP but for high frequency applications we have to consider

all these capacitances and the device also is associated with some inductance because

of its leads we have to also consider the effect of inductance so high frequency model

it will be little different than what we use at lower frequencies So let us look at one BJT so this is a typical

BJT used at lower millimetre wave frequencies so you can see the base emitter and collector

so the bottom most portion it is called the sub collector and actual collector it sits

over the sub collector and on top of this collector we have a thin layer of base Gain

of the transistor it depends on base width thinner is the base we will have higher gain

so look at this plot then (act) this base lead it is connected to this thin layer and

over it we have emitter and collector connection is coming from right hand side So for high

frequency operation to we have to decrease the transit time and if we need to decrease

the transit time we have to reduce the base width Now looking back to basic operation of BJT

I will skip the very basic things only I am going to discuss the limitations which arise

at higher frequencies so let us consider one NPN type BJT Usually an NPN type BJT is popular

at higher frequencies since it involves mostly electron and electron mobility is higher compared

to whole mobility so emitter it will eject electron and for normal operation we will

forward bias the base emitter PN junction and we will keep the base collector junction

in reverse bias condition So due to the applied electric field then electrons drifts into

base where we have recombination with the holes available in base and then rest of the

part we have 1st diffusion and then drift inside the collector Now for high frequency

operation we have to decrease the transit time so that means we have to decrease the

base width But if we will decrease base width the problem

we will be facing that (base ret) based resistance will become very high so we have to avoid

this problem so how we can avoid them this (resita) high resistance problem One solution

is that we can increase doping inside base but if we increase doping inside base than

the hole which drifts into electron from base that part will increase so the reverse situation

current will increase so then we just cannot keep on increasing doping concentration inside

the base so that is why conventional BJP it is not used at millimetre wave frequencies

We have some other version which can take care of this diverse saturation current and

this modification is called Hetero junction bipolar transistor or HBT that comes just

after this So HBT modelling the 1st model this is uhh

it shows low signal low frequency model you can see here we are just considering r Pi

the current roles g m V be and output resistance R 0 and we are not considering any capacitor

this is the low frequency model Now if I go back to device we have a forward biased PN

junction for this base emitter junction so forward biased PN junction it is associated

with some capacitance we also have reverse biased base collector junction it is again

associated with some capacitor and usually reverse biased capacitance is smaller than

the forward biased capacitor so we have to consider all these capacitors their typical

values of the order of Pico farad at low frequency that is why we avoid this but at millimetre

wave or at microwave frequencies we cannot avoid their effect So then the modified circuit how it looks

here shown here In addition to the previous one we have introduced two more capacitors

C be it represents the base emitter forward biased PN junction capacitor and C bc it represents

the base collector uhh reverse biased PN junction capacitor and this is the internal model of

BJP In addition to this we have packaging effect we have external leads so we have to

add more resisters capacitors and inductors for that will take into account the effect

of packaging effect of the leads Next high frequency limitations so high frequency

limitations it depends on many parameters and the condition sometimes we call it junction

condition so it depends on the following parameters saturated grip velocity So carriers inside

the substrate it has some velocity under given electric field if we keep on increasing electric

field value then this carrier velocity will increase but finally it will reach a saturation

which after which it cannot increase for silicon or germanium or this type of materials semiconductor

materials We call that velocity as the saturated velocity V s but there are another categories

of material usually group 3 group 5 semiconductor conductor materials like gallium arsenide

So for this type of materials after a half highest velocity V s if we keep on increasing

electric field then V s decreases So we will consider the maximum velocity of

carriers V s and then next is the dielectric breakdown dielectric breakdown it depends

on applied electric field and it is the property of that given dielectric so let us called

the break down electric field is E m then maximum current it is also limited by the

base width So considering all that effect if we plot current gain for a BJT it becomes

a function of frequency So here in this graph we are showing the plot of HBT for current

gain or sometimes we call it Beta So typically you see at lower frequencies it is fairly

constant but at higher frequency decreases and at a frequency current gain it becomes

one or unity we call that frequency f T of the device so it mainly depends on the capacitance

value Now let us see some parameters which determine

its maximum frequency of operation so the 1st one is voltage frequency limitation So

here V m is the maximum allowable applied voltage device it is given by E m multiplied

by L minimum so L minimum is distance between emitter and collector and V m multiplied by

f T where f T this is 1 by twice Pi Tau transit time cut off frequency it is equal to E m

V s by twice Pi So E m that is the maximum allowable electric field and V s that is this

saturated trip velocity E m V s by twice Pi then we see that it is related to the maximum

allowable applied voltage multiplied by f T so if we increase frequency then maximum

allowable applied voltage it decreases We can also express this quantity E m V s by

twice Pi in terms of current frequency in terms of power frequency so simply we have

to replace the voltage by the corresponding expression of current and corresponding operation

for power so it is shown in next slide So current frequency limitation here I m into

X c multiplied by f T equal to E m V s by twice Pi where I m is the maximum current

of the device and X c this is the reactive impedance of the device so it depends on f

T and mainly based to collector junction capacitance then the power frequency limitation so square

root of P m into X c multiplied by f T equal to E m V s by twice Pi We also can define

power gain frequency limitation so G m V th V m whole square root multiplied by f T this

is equal to E m V s by twice Pi where G m this is the maximum available power gain and

V th this is the thermal voltage so it depends on room temperature so what we see then gain

of any BJT it becomes a function of frequency At higher frequencies we will aspect lower

gain and not only that it is also a function of temperature So we will take a break and

then we will move to next topic HBT thank you