domenica 19 dicembre 2010

16/12/2010

The Gilbert cell.
In this lecture we saw how to use an IC to exploit the property of commutant mixer and how to build the input passaband amplifier.

The realisation of our receiver is based on an IC (NE602) that is an implementation of the Gilbert cell. This is an ideal circuit with three input (two signal plus a periodic signal-the shape doesn't matter!) and two output (one the opposite of the other).
Studying the properties of the equation that represent the Gilbert cell we can distinguish two term that depends on the hyperbolic tangent. For the first one we have to ensure that the difference of the input signals is the linear region, on the other hand, for the second term we want to works on the constant region (+1, -1).
After saw the Gilbert cell we study what we need to add to the NE602 to make it works in a circuit (in theory we need an oscillator but in reality it's half implemented in the IC and therefore we only need a few capacitance and the quartz crystal).

We are at the end of this course and we are expert in realising passband filter. I remember that we need the passaband filter here at the input of the receiver because we want a little bit of amplify, we realize the matching condition and we avoid interference.

We want to the laboratory to test the first part of the receiver, the passaband filter and the mixer, we can observe through PICOSCOPE that we have the desired signal at around 7 Mhz (the distance from 10 Khz is because of the little deviation of the oscillator in the emitter and the receiver) .

Now lack only the last part, we have our signal down in frequency (7 Khz) and we want to turn on a led and sound a speaker, the processing in low frequency its very easy and can be both make electronically with circuit (amplifier plus envelope detector) or via the sound card of a PC.

mercoledì 15 dicembre 2010

14/12/2010

Receptor de una radiobaliza.
We have started the last part of the programme of DR, after seeing the design of a radio beacon transmitter, we moved to the design of a receiver for the 27 Mhz signal of the radio beacon.

First of all we rewied some aspect of the noise, power of the signal in input at the receiver and signal to noise ratio.
Supposing that we need a amplitude demodulator based on an envelope detector based on a diode we can determine the value K of total gain of our receiver.

We remembered from the past consideration that making a passband filter at 27 Mhz it's very hard and expensive, a very clever (and old) solution it's called "heterodyne receiver".
The fundamental idea is to insert a mixer as a first stage of the receiver, at the output we find the same information signal but shifted around a lower frequency for example 10 Khz. Hence elaborate this signal and build a amplify passaband filter in this frequency is incredible more easier and cheaper. Remark that we have to pay attention to the image frequencies (insert a passband filter before the mixer).

We understand soon that we need new information about how to design a mixer. We always see the schematic block in the book but we don't know how to realize one!
It's important to observe that in reality we don't need a complete mixer but we need a circuit that is able to multiply the incoming signal by a sinusoidal one.

Josè Maria show us an alternative mixer: the switching mixer. The basic idea is to exploit the property of the Fourier series of a square wave (+1, -1) that show as a first term a cosine, a square wave for us is simple switch Vin and -Vin.
In the last part of the lecture we saw how to built such a switching mixer in low frequency with the use of OpAmp and how is it implemented in high frequency commercial device.

giovedì 2 dicembre 2010

02/12/2010

On-Off: the last step for the "radiobaliza".
In the first part of the today lecture we made a little brief about what we made up to know: the Pierce oscillator implemented with a bipolar transistor, the tank circuit and the cheap quartz oscillator.

In the last class we stopped talking about how can we interrupt our transmission to implemented a very simple modulation: On-Off.
The first idea that we would have is to switch on and off the voltage source, this is not working fine because of the presence of the condensator electrolytic that have a slow discharge period.
The other possible solution (the one that we decide to implement) is placing the transistor out of the active region, the most easy way is to insert a controlled generator (on-off) with a value of Vcc in series with the R_e, resistance of the emitter. In this way we can alter periodically the polarization network and therefore stop the transmission.

If we want to turn on and off the voltage source we need a component that generate a square wave, because this is a very well-know and common problem there is a IC that makes this works, its called 555. To control the period of on and off we only have to add two resistance and a condesator.

We went to the laboratory, we built the circuit, we added and antenna and we were surprising how our signal is well received by Josè Maria with a commercial radio receiver.

In the last minute of the lesson we talked about how connecting in a boat (and also in a air-plane) the antenna: it's a good engineering problem!

The next step??? We want to build up a receiver for this frequency and look how all the chain tx - rx works fine!

martedì 30 novembre 2010

30/11/2010

Oscillator with the transistor bipolar.
In the previous lecture we saw how to use a logical CMOS port to obtain an inversor, this tecnic was is called Pierce oscillator thanks to his inventor.

This morning we understood how to realize the oscillator using a bipolar transistor.
The final scheme is very simple because there is very few component, but is not to so easy to understand in depth it, in fact we had to make a lot of preparatory lessons.
We polarized the transistor in the active region simply using two resistor for the base and one for emitter. We understand that with another resistance Rc in the collector we can take the signal in output of the transistor. But what happen if instead of this resistance Rc we insert a "tank" circuit (a parallel of L and C). In DC mode the voltage at the collector would be Vcc and the transistor remain in active region.
It important to note that the range of the output in AC mode would be from 0 to 2 x Vcc, that is around Vcc.
In what frequency we have to put the tank circuit? It's important to note that is not the frenquency of the total oscillator! We need a frequncy below that, because we remembered that the tank circuit is equivalent to a parallel of a C and a R above the oscillating frequency, that is what we need.
Finally we note that we can use a quarz of the cheap one not tuned only for the oscillator frequency that we need, but cut for low frequency, we are sure that with the tank circuit the quartz crystal works in the tone that we need.

After the theory we moved to the laboratory and we can build up our oscillator, was great! Incredible, so few component but all well working.
An important comment is that the tank circuit can vary with temperature weather umidity hit but this doesn't affect the work of the oscillator: now this circuit can perfectly fit in a boat!!!

In the last part of the lecture we start to understand that in some way we have to add a little information in our signal, the most easy way is an audible tone like On-Off.

giovedì 25 novembre 2010

25/11/2010

The strength of the stones.
This morning we finished our considerations about the Voltage Controlled Oscillator understanding how we can realize a circuit that can be connected to a PC that is able to controlled the voltage of our varicap diode. This circuit is based on a microcontroller connected to the pc, it generates a digital voltage with a duty cicle tau then there is low pass filter that isolate only the continuos voltage, after that there is an separating stage and at the end the rest of out circuit of the oscillator.

The purpose of our study in the last lecture is to buil an oscillator at 27 Mhz that is used for build a radio beacon for emergency. We have to observe that the frequency of our oscillator have to be very stable and unalterable. However we saw at the laboratory that only a light movement in the inductor is transformed in a change of the oscillating frequency: we have to find a solution.
We want our frequency to be stable and fix like a stone, also to the change of temperature and shot the stone remain unalterated.
Here is the time to introduce the quartz crystals. They have two important properties: the first is the unalterablility and the other is the piezoelectric.
The unalterability is links to the shape and to the cut of the material: a stone.
The piezoelectric is connected to the dimension of the slice of the quartz.

As always when we want to understand a new component the first thing that we made is going in the laboratory (virtually) and make measure on it. If we characterize the quartz crystal we find that for low frequencies it behaves like a condensator, then for a precise fix frequency there is the presence of a short circuit and if we increase a little bit the frequency there is a very strange behavior. In a very small range of frequency from the short is transformed in an open circuit and the behavior is like an inductor that takes in that range all the possible value.

After this discover our question is how to build a circuit with Rs, Ls and Cs that act like this quartz crystal. We analyze quickly a circuit proposed by Jose Maria and we discover that the crystal behavior is like a seris of a little R a little C an high L, everything in parallel of a C (that is the parasitic capacitance of the quartz crystal).

What we can do with this quartz crystal? We understand that its vibrates at only one precise frequency (the one of the short) therefore if we inserted it in our wire of feedback from the output to the input we obtain that our oscillator only works if the frequency is perfectly tune with the one of the crystal (in reality it works in a little range around it).

Our aim is to build from this crystal an oscillator that works always. We don' t want to periodically calibrate it or something similar, therefore we study more in depth this quartz of crystal, a little before we say that is able of be an inductor L of whatever frequency that we want, therefore anywhere in a circuit we have an inductor of value L1 at a frequency we can replace it with a qurt crystal of that frequency.

Later on we study another possibility of make an oscillator that is with an inverter amplifyer and a passaband filter. If the amplifyer is invertent we need to restablish the right phase adding a 180° more of desphase and to do it we need at least three reactive elements.
For the implementation the professor showed us a very clever way that is with the use of a logic port in CMOS tecnology (note that the port is exploited it in the region prohibited for the digital circuit!).
Another way of implementing the inverting amplifyer is by a bipolar transistor; moreover we can substitute the L element of the passband filter with the quartz crystal.
At the end we obtain a very simple realization of an oscillator sinusoidal.

We need to comment that we use a cheap quartz oscillator for the 27 Mhz that is realize expoiting high armonics (the third for precision. High armonic is higher armonic oscillation of the slice of quartz). If we don't want to see the low armonics we have to implement in the circuit also a tank circuit to suppress it.

martedì 23 novembre 2010

23/11/2010

VCO and spectrum analyzer.

At the end of the last lecture we comment about the measurament of the power of our oscillator that we did at the laboratory.

We understand that our sinusoid is not perfect and therefore we have to mesure the power also of the armonics. A well-build oscillator is an oscillator that have a very low power at the superior armonics.

But what kind of instrument we need to see the spectrum of our oscillator? Obviously we need a spectrum analyzer.

To understand profundly how a spectrum analyzer works Josè Maria had shown us a step-by-step path about how can we build such an instrument from only the basic notion that we knoe about passaband filter voltimeter and envelope detector. We need a passaband tunable filter to window the input signal, we saw that we could have an electronic control to sweep all the frequency.

In reality making a very well shaped filter that can be tunable is very difficult and the BW is very important baceause is directly related to the resolution of the spectrum analyzer. What we can do is a very good filter (in BW and in costant gain) at a fixed precise frequency.

Anyway exist a solution to this problem: we use a fixed frequency very good passaband filter and we use a mixer between the input signal and a signal from an oscillator controlled in voltage. With some passage of math can be demostrated that is the same of having a tunable passaband filter.

We went to the laboratory to test the use of the spectrum analyzer, this spectrum analyzer is made of two stage including an oscilloscope. We test this spectrum analyzer with a sinusoid from a generator of % Mhz and 0,466 Volts, we could saw the spectrum and recognise the armonics and the noise from the FM radio (in the band around 100 Mhz).

At the end we understand how we can control the frequency of our oscillator using a variable resistor and a diode varicap (is s diode polarized in inverse, it works like a variable capacitance) and so varying the voltage. We need it because we want that our oscillator produce a very good sinusoid at a precise frequency and therefore in some way we have to build up a system that can works alone without the need of an human hand that change the capacity of a variable capacitor (like in out old oscillator circuit).

venerdì 19 novembre 2010

18/11/2010

Using transistor for the amplifier of the oscillator.
It' s clear that if we want to create high frequency sinusoids we have to works with a transistor instead of an OP amp.
The structure of the oscillator that we want to study is made of an amplifier made with a transistor and with an amplify of 1, a resistance of input high and a low impedance at the output, in other words this is a voltage buffer (esp: seguidor de tension). The band pass filter have an amplification higher that 1 in the peak of resoncance.

First of all we study that circuit without the connection input-output, so we find the condition of oscillation and the frequency of the peak (that depend also on the value of a variable capacitor).
In a second step we study the design for the voltage buffer, hence we found the equation for the polaritation of the transistor (obviously we used the little signal model for the transistor).
Finally we dimensioned all the value of the rsistance inductance and capacitors to obtain an oscillator fot 27 Mhz.

After the theory part of the lecture we went to the laboratory to build the circuit and to test it. We simulate the load of the antenna with a resistence and we used an autotrasformer that Jose Maria given us. We have to measure the power that is trasmitted to the resistance (the "antenna") we have to observed that we don' t obtain a perfect sin wave and therefore we understand that the power that we measured is not only of the component in the 27Mhz.

The quality of an oscillator stay in how much are attenuated the armonics of the principal frequency f0.

mercoledì 17 novembre 2010

16/11/2010

The new project: diseño de una radiobaliza - Emisor -.
During the lectures of any electronic or telecom courses we speak of sinusoidal signal, we used source of sinusoidal signal and so on... but where they come from? How in reality we can produce this kind of signal? The basic idea is the sinusoidal oscillator.

If we think how to obtain a sinusoidal signal (a precise "tone" in frequency) we can tought to the parallel of L and C. But it' s important to understand that this is only an ideal circuit. We saw in the practice that we can' t have a real component called L, but we have an inductance with her parasitic resistence R. Hence the circuit is not again an oscillator, because of the presence of the R we obtain a fading sinusoid.

We need a new idea and we start to study a different structure: an amplificator of gain k and a filter H. If in input we have a sinusoid signal from a generator, what we obtain in general at the output? And what are the condiction in the amplifier and the filter to obtain that the output is a replica of the input? If we answer to these question we conclude that the product of k and the amplitude of the filter at the input frequency must be equal to 1 and the change in phased 0.
What happen if we connect the input with the output? Because they are the same (under the condiction writed above) nothing happens in the sense that the results doesn' t change.
Therefore we have removed the generetor and we find that the circuit admits for solution a sinusoidal signal: we found the oscillator sinusiodal!

For example using an OPamp (because we need small amplification factor) we can realize the amplifier and we can realize in practice the filter with a resitance a capacitor and an inductor. We have to pay attention that the conditions are met (1 in total amplification and 0 phase shift for a particular frequency f0 that depends on the value of L and C).

There is two important remarks. First, what happens if the amplify is not perfectly 1? The answer is that the sinusiod can make increasing oscillations or decreasing one. Second, if we don't have the generator (obviusly because we are trying to making it!) how can this circuit start to works? The secret is that at any temperature the amplifier itself make some termal noise, this noise is a constant in all the frequency and therefore the frequency f0 is excited and if we have a passaband filer we obtained the desired sinusoid.
We have to note that the product of the amplify factor k and the value of the amplitude of the filter must be little higher than 1.

Where we have to take our sinusoidal signal producted from our oscillator? There is two possibilities, one is at the output of the filter and the other is at the output of the amplifier. The first chose is positive if we look at the shape of the sinusoid but we have to observe that is an high impedence node and this is not good because anything that we connect modify the oscillator circuit. The other possibility is to take the signal at the output of the amplifier, this is a low impedance node but we find that the shape is not a perfect sinusiod because if the amplitude is more than 1/H(f0) we find a crescent sinusoid that is cutting because of the saturation limit of the amplier.

In conclusion we obtain an oscillator if we have an amplifier and a passaband filter appropriately designed, the output of the filter is connected to the input of the amplifeir. Under certain conditions this circuit start itself to produce sinusoid signal at the frequency of the peak of the passaband filter, we obtain purer sinusoid if the Q quality factor of the peak is higher.

giovedì 11 novembre 2010

11/11/2010

Transformer or not transformer.

In the MW receiver we used a transformer to don't degrade the bandwidth of our antenna-tuner inserting the rest of the circuit. It's obvious that we have always a receiver antenna and we have to connect it to the receiver circuit so this is a general problem, therefore we understand that the transformer could resolve this problem in a very elegant way.
It' s important to observe that there is a price that we have to pay: first of all the high price for the materials (copper and ferrite) and second we lose in amplify factor (i repeat we preserve the bandwidth.)
We also see a "reverse" application of the transformer that is connecting a voltage source in series with a resistance at the secondary stage of a transormer, in the primary we have for example an high value of resistance, we find that we have "adapted the line" that is another very important and common problem building receiver in RF.

There is another clever solution to the problem of not degrade the Q factor of a resonant circuit: the use of two capacitance in a very particular way.
To study this solution we used two particular transformation that sound strange (but that are possible under certain frequency condition): the transformation parallel to series and the vice versa.

We have to understand that now we have the instrument to study complex circuits in an easy way, the observation is that we don' t use the articulated expression of the transfer function (with denominator not of the second grade).

Finally we see how to connect two different stage to obtain a wider passband filter (costant in a wide bandwidth) and further the concept of autotransformer that can simplify the realization of a transformer.

mercoledì 10 novembre 2010

09/11/2010

HF filter and trasformer.
After seeing that our MW receiver correctly works (was very surprising and exciting!) we start to re-study the circuit that we used in the MW receiver to obtain a more general knowledge about HF filters.

In class we study the basic circuit: inductance, capacitance, low pass, high pass. After that we revise the concept of Q (quality factor) and we try to extend the knoweldge about the antenna-tuner-band pass filter to other circuit: we used the Thevenin equivalent.
Then we study the "dipole RLC" and a very simple but very useful ciruit called "tank", that is a resonant circuit with a resistance Rs in series with the L.
The osservation that we made after a few step of easy math is that for a particular frequency (the resonance one) and for a quality factor Q>5 we can replace the Rs with a resistance in parallel called Rp: this technics is very very useful when we want to understand the trend of different circuits without making a lot of math.

Jose Maria showed us two possible application of the tank, the first is converting a simply inverting amplifier in a band pass filter, and the second is to obtain a tuned amplifier with a transistor.

Changing topic, we started talking about how to make inductances, the first way is around a toroidal ferrite (we understand also why we use a ferromagnetic material and not simply iron!), the second is around a kernel of ferrite (like our antenna in the MW receeiver) and the third is without ferrite component by only rolled in free air (the value of L is lower but we have less parasite resistance and capacitance). N.B: Nagaoka obtained a sperimental equation to describe this kind of coils.

The natural flow of thought moves from inductors to the magnetic coupling and transformer. The professor advised us not to run in error and to understand very well the difference between the ideal-theory concept of transformer and a realistic way of made them: the different is very deep and we have to take it into account.

martedì 2 novembre 2010

28/10/2010 and 02/11/2010

The last stage of our radio receiver.
We had stopped the last time writing about the positive regeneration obtained inserting an another solenoid in the same ferrite kernel of the primary, we obtain an higher amplify and a better selectivity.

I think that to understand better the last necessary parts of our radio receiver i have to make a summary of it:
- first of all we have to build a separating stage;
- second the envelope detector;
- then an audio amplifier;
- finally the transformer fnecessary if we want to connect some speakers.

The need for the separating stage.
We saw in the last lecture how to polarized the transistor to obtain an output voltage in DC of 4,5 V. It's clear that at the output of the transistor we have to insert other circuit for our radio development, but what happen if we directly connect the rest of the circuit in the output terminal of the transisor, obviously we change the value of the resistence and capacitance used for the polaritation.
Therefore we have to build up "something" that separate the rest of the circuit that we will connect to the transistor, in technical language we said that we need to transform our low impedence terminal in an high impedence one.
To make this separating stage we used a common Op-Amp. Correctly Josè Maria spend some time during the lesson to explain to us the carachteristics of these Op-Amp, the most interesting thing is that we have to pay a very high attention reading the specific and understand that we are working in high frequency and therefore the characterist are very very different from what we remember in our last electonic circuit theory.
Finally we understand that we used this Op-Amp in our separating stege for his carachteristic of high input impedance and not for amplify our signal.

Obtaining the original information: the envelope detector.
We have to remember that the original information is in the envelope of the signal that we receive (AM modulation) therefore now that our signal is amplified we need to restore this information. To do it there is a very simple way, we only need a diode a resistence an a capacitance.
First of all we revised the way how to obtained a continuos corrent from a sinusoid, therefore we understand better how this work also in the laboratory (trying different diode -Si or Ge-and different value of resitance).
We have to note that in a sense we need a circuit that follow the trend of the voltage (the envelope of our signal) and therefore we understand that there are some upper (to avoid diagonal distorsion) and lower limits (for the portant) in the value of the resitance.
We went to the laboratory another time and we remain surprise how good is the envelope detector used to recover a voice signal.
(Note that this receiver and this envelope detector works fine with no fast transient music).

The last step: the audio amplifier and the transformer.
In the last part of the today lesson we saw the audio amplifier need to connect the speaker. We have to understand that now we are at audio frequency (20hz - 4khz) therefore we use an active amplify TL081.
We saw the limitation of this amplify in term of current in output and after a few analysis we arrive at the conclusion that we can't connect directly at the output terminal of the amplify the commerical speakers that have a typical value of impedence of 8 Ohm.
Josè Maria showed us a very clever (but commercially expensive) solution to connect the speaker, this idea is based on the transformer, but we have to observed that we need a very high inductance value therefore it can't disturb our circuit, to obtained this high impedance we have to make a lot of coils (with the same number ratio).

martedì 26 ottobre 2010

26/10/2010

We spent the first our of this morning lesson in the laboratory. We have have to find the curve of amplify of our circuit.
We know from what we study in the theory that we would have bad value in high frequency (more than 1 Mhz) because of the parassite capacitance of the transistor. In the laboratory we found that kind of decrasing curve.
In the previous lecture we saw that we can improve the trend inserting an inductor L to compensate the effect of the parassite capacitance in high frequency, therefore we inserted that inductor and we could find better performance.

In the second part of the lecture Josè Maria explained to us the concept of positive regeneration. The basic idea is that we want to obtain an higher amplify. This idea is patended by Armstrong in 1915 and we have to understand that like in the principle of the electric age the engineers want to obtain the maximum with the minimum, in this case we want the maximum amplify with the minimum number of transistor (at the principle vacuum tube).
The way to obtain this positive regeneration is based on the idea that after we receive the signal we can emitting it with an antenna and then we re-received it.
There are two fundamentals problems: the first one is that we have to re-receive the signal in a constructively way (adding in phase) and the second problem is the saturation.
In practice to build up this "emitting antenna" we have only to add a little solenoid in the same ferrite kernel that we used for the antenna-tuner (paying attention to the phase), and a controlled resistence.
Controlling the resistence we can modify both the sensibily and the selectivity of our circuit (improving both at the same time) and we can avoid the problem of saturation.

giovedì 21 ottobre 2010

21/10/2010

Trying the transistor.
Today we start the lesson seeing that if we put our information voltage signal in series with the batteries we don' t obtain the gain that we desidered. Therefore in this montage our transistor "only" act a controlled switch.
Josè Maria said to us that we have to put our signal in parallel with the source power hence from base to ground and we have to add a capacitor that doesn't change the circuit in the DC mode.
Using the incremental model and the fasorial representation we arrived to calcute the amplify in this case: in principle we can choose the parameter to obtain the amplify that we want.
But there is a problem, the high we want the amplify the high result the impedence at the input of our amplifier, and remembering the previous lecture this is exactly what we don't want because we need to preserve the sensitivity of the antenna-tuner stage!
The problem of this high input resistance is well know in literature with the name of Miller effect. Fortunately there is a very clever solution to delete the undesidered Miller effect and hence we to replace the Rb (resistence in the base terminal) with a series of two Rb/2 resistence and in the middle a capacitor.

Jose' Maria showed us two other improvement for our circuit. The first is to insert a inductor L to compensate the effect of the parassite capacitance in high frequency, the second improvement is to insert in the emitter terminal a little resistence Re that doesn't afflict the polaritation circuit but that is seen multiplied for Beta in input (is fine for the sensitivity!).

We spent the last thirthy minutes of the lecture in the laboratory. We built the polaritation circuit, we turn on the power and we measured the real parameters of the circuit. We found a Collector voltage of Voq=4.58 V (theorically is 4.5 V) and a Base voltage of Vbq=0.61 V (teorically 0.6 V), with this mesured value we can calculate the various current and also the amplify factor that is A=169,99 that is in line with the theory.

mercoledì 20 ottobre 2010

19/10/2010

Putting transistor in the active region.
After seeing another time how the transistor works we want to stress the fact that the nobel idea in the transistor is that little variations in the input imply big variations in the resistence in the output because there is a strong sensitivity in the voltage Base-Emitter.
Transistor comes from the words transfer + resistor that means that we can transfer the same current from an high resistence to a resistence that we can choose.

If we want to obtain the ransistor effect we have to put the transistor in the active region which means that we need a positive voltage both for the Collector-Base and for Base-Emitter.

If we want to check if the transistor in a circuit is in active region we have to check if at leat one of these current are positive: emitter, collector or base.

In the class we saw some proposals about how put a transistor in the active zone, in the simpliest cases we found some problem about the instability of the circuit due to temperature, the dependence from the value Beta of different transistor or the need for 2 batteries to aliment the circuit.
The last example that we have seen is made with only one batteries, with four resistance and it doesn't depend from the different value Beta of transistor (Note that also if the model of the transistor is the same we can find very different value of the parameter Beta).

In the last part of the lecture we developed an incremental model for study circuit with transistors and we start talking about the possibility of insert our voltage signal in parallel with the voltage Base-Emitter.

lunedì 18 ottobre 2010

14/10/2010

The second stage.
The last lecture we saw how to make a transformer in our circuit to see a lower input impedance of the HF amplifier. Now we have to build this amplifier!
Beacuse we are in high frequency we can't use an operational amplifier and we have to remember and to exploit the propierties of the transistor.
Josè Maria explain us that to better understand the proprierties of the transistor we have to revise the feautures of the diode.
First of all we understand that a diode is made from a semiconductor (silicon), more precisely there are two different region doped P or N joint togheter.
After we have to observe that whenever we have a diode in a circuit, at that time the circuit is no longer linear, and so we saw the curve voltage vs corrent.
In practice we want to work with a linear circuit and therefore we saw how to approssimate a diode with a linear circuit.: we have to distinguish two different stage, one when the voltage is upper than the thereshold voltage and the second when the voltage is lower.
After that is necessary to see what happen when the voltage is moving a little bit from the fixed point used for the approximation. We saw that we have to image that there is a voltage generator and a resistence.
In summary if we have to study a circuit with diodes we have to proceed in two step: in the first step we find the operating point of the diode and in the second step we study what happen with little variation (we also exploit the proprierty of superposition).

After studied the diode we started to see how works a transistor. A transistor is made by 3 regions (normally N-P-N), we observed that the P-region is very very thin and so we can't approssimate the operation of the transistor like simply 2 diodes. The transitor has three terminals: Emitter, Collector and Base.
We saw that a little variations in the voltage of B-E correspond in an high variations in the current C-E, therefore if we insert a resistance we obtain an high variation in its terminal. Therefore we found the amplifiy effect of the diode!

venerdì 8 ottobre 2010

07/10/2010


Selectivity vs. sensibility.
We saw that we need more amplification to achieve the 0.3 V. So we have to insert an amplificator HF, but there is a problem, beacause of this amplifier the bandwidth of our antenna-tuner-band pass filter widens, and the amplitude of the peak decreased.
There is two important parameter that we have to consider, the sensibility of our system and the selectivity, there first is connected with the capability of detect poor power signal and the second is directly linked with the ability of the filter of cutting out the other station near the station that we want to hear. Is easier to hear that a radio system detect more than one station at the same time that see that the radio hasn' t a fine sensibility.
So we want first of all to preserve the bandwith of our filter, the most simple way to reduce the effect of the resistence of the amplificator HF is to implement a transformer so we can see an higher resistence. (Note that we lose in sensibility.)

Next we have to make this trasformer real in our circuit. To do it we roll (10 times) another wire above our primary coil, and we want to obtain in theory a parameter for the transormar equal to n=60/10=6. So we went to the laboratory and after made it we make some measure and we find that n=6.06. The resistence of the amplificator in HF will be seen multiplied by n.

martedì 5 ottobre 2010

05/10/2010

Measuring our coil.
Our aim of today is to measure the performance and some values of the coil that we made the last time.
The first things that we observeb is that we can measur the values of inductance L and the parasite resistance Rs from some value that we can see in the oscilloscope, we can calculate L from the value of the frequency of the peak and then with this value and the value of the max amplitude we can calculate Rs.
Like Heisenberg said in his uncertainty principle we can't measure some values without modifing that values, only beecause we are trying to mesure with an instrument. So our target is to minimize the effect of our instrument.
Jorgè Maria showed us a low capacitance probe that is able to "put" only 12 pF o capacity in our mesuring set-up. The price that we have to pay is that the amplitude is scaled by a 10 factor.

Now we were ready to go to the laboratory and test our coil and calculate L and Rs. We made this experiment for two position of the coil respect the kernel of ferrite so we can have an idea of the range of the values. Unfortunately me and my companion can't make the second mesure with the coil in a side of the ferrite because our coil doesn't slide well in the ferrite. The value that we obtain with the coil in the middle is L=310 uH and Rs=22 Ohm.

What happen? Why Rs is so hight? Tthe problem is that at high frequency (in our case around 1 Mhz) the skin effect is high and the copper becomes not a so good conductor.
With an Rs so high the performance of our antenna-filter-amplificator is worse. And so we have less amplification and a more large bandwidth.

Finally we observe that if we want to add an other amplificator in HF we have to minimize the effect of the resistence that is seen at the terminals otherwise we have, another time, a worsening in the performance.

giovedì 30 settembre 2010

30/09/2010

Let's do it.
We said that we have to built a pass band filter that also act as a tuner. If it's possible we want also that the filter is able to amplificate our signal so we can respect the condition on the volt that have to be more than 0.3 so that the envolope detector can works.
José Maria has shown us a very clever solution based on the series of a resistor an inductor and a capacitor. Formally this is a low pass filter but if we choose parameters properly that filter show a peak in the frequency that we want... and so we have the pass band filter and also the amplificator!
I was very impressed when we saw that a simple antenna (coil and ferrite) in parallel with a capacitor (variable) act perfectly as the scheme that we saw for the passband filter, so in the end we have our antenna that is the same of having an antenna, a passband filter, a tuner and an amplificator. Incredible.
In the second part of the lecture we study how we have to make the coil since we have our variable capacitor (we have to buy it). We found that we have to make 60 rotations around the ferrite.
After that we went to the laboratory and we start to make it, in the first time seemed simple but if you want to make it well you have to do a little effort. Now i' m curios to measure the performance of it.

mercoledì 29 settembre 2010

28/09/2010

Simple and inexpensive.
We saw that if we want to recover the original sound signal we have to use a moltiplicator and then a low pass filter to eliminate the high frequency components. But how much difficulty is need to build up a moltiplicator in the receiver that is with the same parameters of that one in the transmitter?
Now think that we want the radio be a mass media? First of all we have to find a way of divide the frequency to different station. This is not a problem beacause at each station we assign a different portant frequency, we only to pay attention to separate enough the different channels. Secondly, we have to produce very cheap receiver because we want to sell a lot, and so people how to pay very few money to buy it.
Our target is to eliminate the moltiplicator at the receiver (is the most difficult, and so expensive, component to make). We find the solution, in doing that we have to complicate the transmitter! Now at the receiver we only need to replace the moltiplicator with an envelope detector: this is the amplitude modulation (AM).
At the end we observed that the envelope detector need more than 0.3 Volts to work and we have to build a pass band filter that works also as a tuner.
After seeing the theory we went to the laboratory to mesure an unknow AM signal, we use PicoScope (a software that works like an oscilloscope). Only seeing at the wave form and mesuring some amplitude and distance (in frequency) we have discover what was the frequency of the portant, what was the frequency of the tone and what was the parameter m (modulation index) that is the ratio between the max amplitude of the envelope signal and the costant volts added to obtain a positive signal.
Here there is two capture from PicoScope


giovedì 23 settembre 2010

23/09/2010

bending the earth!
If the sender and the receiver are too far what happens? We find a problem: the earth is a sphere and the propagation of the electromagnetic wave is straight. Fortunately there are some natural features that gives us an hand: the effect of the surface waves and the refraction by the ionosphere. So we can transmit in HF also at more than 300 kms.
The first thing that we want to transmit is the voice, after understood how the voice is working (at what frequencies) we arrived at the conclusion that we need to "shift" our information (the voice) up in the spectrum of frequencies, to do this we need to build a moltiplicator and to understood how to recover the original voice information.

martedì 21 settembre 2010

21/09/2010

Trying to send and receive.
Now that we have the antenna (dipole) we can try to send and receive some messages. The first thing that we can do is to think at a code (Morse code) where each letter of the alphabet is coded into points and lines. When we want to send a point we close for a short time a switch and we transmit a sinusoid, when we want to send a line we close the switch for a longer time. At the receiver we put an oscilloscope connected to the receving dipole and when we see a short sinusoid we "decode" it like a point and when we see a longer sinusoid we relate it to a line.
There is a problem, in the pauses between the points and lines we don't see the oscilloscope in the zero state, why? Because we see all the electromagnetic noise that belongs from the human industrial activities and from the nature. If we are too far from the transmitter, that noise can cover our desidered signal. We can' t think to amplify it because we amplify also the noise. We observed that there is more noise in the low frequencies due to the fact of real impulse. We defined the ratio between the signal and the noise (S/N) and also the minimum detectable signal (related to the minimum power that i need in reception if i want to obtain some information).
If we are too far from the transmitter we can think to change the antenna with one that is more directive (eg. Yagi-Uda or parabolic).
In the end we saw the definition of dB and dBm and we applied it at the formula of received and trasmitted power. We also observed these definition (dB and dBm) in real datasheets of some commercial antennas.

venerdì 17 settembre 2010

14/09/2010 and 16/09/2010

The first two lectures.
I remember when i was a child that i was always questioning me about things around me . I grew up and i started to disassemble and riassemble everything that was mechanical or electronic like old computers or watch. Then i decided to enter at university. Unfortunately at the end of the most of the courses the questions are more than the answers, i always thought that the cause was the total lack of practical laboratories in my university in Italy.
The course of DISEÑO DE RADIO RECEPTORES is started exactly how i dreamed. Professor José Maria drew in the blackboard a typical block diagram of a communication system (one of those i saw many times in the books and i wrote many times in my notes), i always asked to myself how in the real world (with circuits, resistors, wires, transistors, ecc...) are made each block of those diagrams. This is the same question that José Maria asked to us. Very well, i hope that in this course i will able to answer to old (and new) questions.
The first chapter is REINVENTADO LA RADIO, we have to imagine that we are only experts of circuits theory, and we want to build up a radio (to communicate).
First of all we need to generate electromagnetic waves, so we have to build an antenna (the dipole is the most simple). The lenght of the dipole is very important and it's directly related at the wavelength (or at the frequency of interest). It's important to know that the same antenna used in transmission can be used in reception to capture the desired signal.
Build this dipole antenna can be very difficult in the real world for low frequencies and so we studied others types of antenna: monopole or ferrite coil.
We saw the laws of the radiating and receiving powers: they are not only formulas but each term has a exact physical meaning connected for example to the volume of a sphere, to the gain of an antenna (if is not isotropic) and to the distance from the transmitter to the receiver.