few queries on transistor

[PG1995]

Hi

Please help with the folloing queries. Thank you.

1: Which kind of transistor is used modern computer processors? BJT?

2: Why is Intel more popular (at least as far as I know) than AMD? Does it have to do with maketing only, or, are there technological differences also?

3: A transistor in itself a simple type of amplifier.

 

[betwixt]

1. Definaitely not BJT. Modern processors use MOS devices, either NMPS, PMOS or CMOS.

2. Intel invented the microprocessor so they had a head start in the market, they also produced other types of ICs (ROM, RAM etc) many used as support for their processors. AMD came into the market quite late and and used Intel's established technology (some would argure they stole it) to produce microprocessors without many of the development costs. Since then AMD have done more development work themselves and their products are similar in performance to Intel's. My personal opinion is that Intel make 'work horse' devices, which give good all round performance whereas AMD make ones which are more tailored to performance in particular applications. So marketing does come into it, AMD can boast better performance in specific circumstances, Intel can boast better all round performance but may lag in specific applications.

3. Not necessarily, transistors have many applications, amplification is just one of them. They can be used as logic switches (as in microprocessors) as buffers to give isoaltion without gain and in some instances they can be used as rectifiers. In all of these a transistor still needs other components to support it, little can be done with a transistor alone.

Brian.

 

[PG1995]

Thanks a lot, Brian. It's very kind of you.

1. Definaitely not BJT. Modern processors use MOS devices, either NMPS, PMOS or CMOS.

A transistors has different types and two of them are: BJT, FET. I haven't read much about the transistors but I take it that the basic 'operation' of the all these types is same. For example, both CD and DVD have the same operation of storing information in digital form, the difference lies in their in their internal structure. The structure of DVD gives it edge over CD. So, what would you say about the internal structure of different types of transistor and why one type is considered more efficient over others? Please remember that I'm a beginner, therefore a simple and short would do. Thanks.

3. Not necessarily, transistors have many applications, amplification is just one of them. They can be used as logic switches (as in microprocessors) as buffers to give isoaltion without gain and in some instances they can be used as rectifiers. In all of these a transistor still needs other components to support it, little can be done with a transistor alone.

Brian.

Thanks for telling me this. Even there are many kinds of amplifications: inverting amplification, differential amplification, etc. Correct?

Thank you for the guidance and your time.

Best wishes
PG

 

[betwixt]

The situation with transistors is a little different to CD and DVD. With optical disks the difference is basically a scaling factor, the laser wavelength is different and therefore the surface impressions are different sizes and so it the number of them you can fit in the area of a disk. BJT and MOSFET are completely different constructions although both use doped semiconductor materials. Bipolar transistors (BJTs) inject current into the base junction to control the current betwen the emitter and collector pins, MOSFETs use charge displacement to 'strangle' the flow between source and drain pins. From a practical pont of view, you have to provide base current to operate BJTs but only a voltage to operate MOSFETs. It is rather more complicated than I can explain here but I'm sure there will be many references on the Internet about the specifics of the two types.

There isn't really an issue over efficiency, both are equally efficient at their own jobs but in some situations one type is better suited than the other. It's rather like comparing two methods of transport, a bicycle and a bus. Using a bus to carry one person isn't efficient but neither is trying to carry a crowd on a bicycle! Both are efficient if used in the manner intended. For microprocessors, BJT is not used because their junctions usually are physically larger and the requirement to supply base current limits the total current the device could draw. Even tiny base currents when there are millions of transistors becomes a huge heat and power problem. MOSFETs can be made smaller and consume less power so they are better suited to dense constructions. On the other hand, BJT is generally better at higher power, low density applications.

Amplifers come in all shapes and sizes, their common factor is that they take a signal and enlarge it in some way. Inverting amplifiers also turn the signal upside down so positive going sections becme negative going and vice versa. Differential amplifiers have two inputs and amplify the difference in signal between them. There are other types as well but they are beyond explanation in a s imple message like this.

Brian.

 

[PG1995]

Many, many thanks, Brian. Your reply was very helpful. Please keep your replies simple and straightforward so that I can understand them. Thank you.

The situation with transistors is a little different to CD and DVD. With optical disks the difference is basically a scaling factor, the laser wavelength is different and therefore the surface impressions are different sizes and so it the number of them you can fit in the area of a disk. BJT and MOSFET are completely different constructions although both use doped semiconductor materials. Bipolar transistors (BJTs) inject current into the base junction to control the current betwen the emitter and collector pins, MOSFETs use charge displacement to 'strangle' the flow between source and drain pins. From a practical pont of view, you have to provide base current to operate BJTs but only a voltage to operate MOSFETs. It is rather more complicated than I can explain here but I'm sure there will be many references on the Internet about the specifics of the two types.

But they both achieve the objective using different way.

There isn't really an issue over efficiency, both are equally efficient at their own jobs but in some situations one type is better suited than the other. It's rather like comparing two methods of transport, a bicycle and a bus. Using a bus to carry one person isn't efficient but neither is trying to carry a crowd on a bicycle! Both are efficient if used in the manner intended. For microprocessors, BJT is not used because their junctions usually are physically larger and the requirement to supply base current limits the total current the device could draw. Even tiny base currents when there are millions of transistors becomes a huge heat and power problem. MOSFETs can be made smaller and consume less power so they are better suited to dense constructions. On the other hand, BJT is generally better at higher power, low density applications.

"...junctions usually are physically larger..." - Would you please tell me what you mean by "junctions" here? Are you taking about the junction between n-type and p-type?

"...base current limits the total current the device could draw..." - Are you saying that using BJTs in a certain device would require a lot of current to switch on the transistors? Would you please elaborate your statement a little? Thanks.

"There isn't really an issue over efficiency..." - I'm sure what you say is correct. But I remember someone saying (or, something close to it that BJT takes more time to switch on than a FET. What was that someone trying to say? Can you make anything out of it?

Amplifers come in all shapes and sizes, their common factor is that they take a signal and enlarge it in some way. Inverting amplifiers also turn the signal upside down so positive going sections becme negative going and vice versa. Differential amplifiers have two inputs and amplify the difference in signal between them. There are other types as well but they are beyond explanation in a s imple message like this.

Brian.

Can a BJT work with AC? I don't think so. In a PNP BJT current flows flows from collector to emitter, so during +ve half cycle of AC the transistor will conduct and during -ve half cycle wouldn't. On the other hand I think an amplifier, such as op-amp, can also work with -ve half cycle would mean that amplifier can work with AC. Please correct me.

Thank you very much for all the help and please excuse me if my queries sound somewhat stupid!

Best wishes
PG

 

[betwixt]

"...junctions usually are physically larger..." - Would you please tell me what you mean by "junctions" here? Are you taking about the junction between n-type and p-type?

Yes, transistors have at least two P-N junctions. The physical construction of a MOS device makes it easier to fabricate on silicon and also smaller.

"...base current limits the total current the device could draw..." - Are you saying that using BJTs in a certain device would require a lot of current to switch on the transistors? Would you please elaborate your statement a little? Thanks.

A BJT is switched on by providing a current into it's base pin (the middle of PNP or NPN), the current has to come from somehwere. In small systems the current is provided by surrounding components, commonly resistors but these are difficult to fabricate in silicon. In large systems like microprocessors, even though the base current may be very small, when you take into accouint that there may be several million transistors, the combined currents of all of them could be very large. There is also a heating factor to consider, every time current flows a little heat is produced and in a large system this could be many Watts. A MOSFET is voltage driven, no current flows into the gate pin so much of the current and power loss is eliminated. There is still some current but only as the voltage changes because there is a small capacitance between the pins of a MOSFET (BJT as well) and it has to charge and discharge as the voltage across it varies. In comparison with BJT base current it is very small though.

"There isn't really an issue over efficiency..." - I'm sure what you say is correct. But I remember someone saying (or, something close to it that BJT takes more time to switch on than a FET. What was that someone trying to say? Can you make anything out of it?

Not true, although the size of the device plays a part in how fast it can switch. I have some MOSFETs here that can only switch at low speed (less than 1MHz) and I have some BJT transistors rated at 25GHz, some 25,000 times faster. The speed of MOSFETs is largely decided by how large the gate junction is and therefore how much capacitance it presents as a load to the circuit driving it. If you look at circuits of switch mode power supplies for example, you often see BJTs being used to drive MOSFETs to ensure a fast rise and fall of gate voltage.

Can a BJT work with AC? I don't think so. In a PNP BJT current flows flows from collector to emitter, so during +ve half cycle of AC the transistor will conduct and during -ve half cycle wouldn't. On the other hand I think an amplifier, such as op-amp, can also work with -ve half cycle would mean that amplifier can work with AC. Please correct me.

A bipolar junction can only conduct in one direction. There are some special devices such as triacs which conduct when the voltage across them is reversed but they contain more than one PN junction arranged in opposite directions to achieve this. Some MOS devices are symetrical inside which means you can swap the source and drain conections, they are used as controlled resistances but most have some built in protection between the gate and source which prevents their use in that kind of way. Amplifiers can not normally "produce" a negative output. When their output voltage swings above and below 0V it is because they have both positive and negative power supplies to them. The output is essentially referenced to 0V by the components around them but can change in both directions by drawing curent from one supply line or the other. They can not be powered from AC.

Thank you very much for all the help and please excuse me if my queries sound somewhat stupid!

The only stupid questions are the ones never asked. I can't guarantee my answers aren't stupid though :smile:

Brian

 

[PG1995]

A bipolar junction can only conduct in one direction. There are some special devices such as triacs which conduct when the voltage across them is reversed but they contain more than one PN junction arranged in opposite directions to achieve this. Some MOS devices are symetrical inside which means you can swap the source and drain conections, they are used as controlled resistances but most have some built in protection between the gate and source which prevents their use in that kind of way. Amplifiers can not normally "produce" a negative output. When their output voltage swings above and below 0V it is because they have both positive and negative power supplies to them. The output is essentially referenced to 0V by the components around them but can change in both directions by drawing curent from one supply line or the other. They can not be powered from AC.

Thank you very much, Brian. What I appreciate about you is that you have the ability to make things simple for people like me.

Please have a look on this linked page where you can also see my question: https://img545.imageshack.us/img545/7082/acamplifier.jpg

Thanks a lot for your help and time.

Best wishes
PG

 

[betwixt]

Yes, your image is correct as long as you have a sufficiently high DC voltage present to allow the collector to reach peak voltage. Remember that the tranisistor can not make voltage appear out of nowhere, it can only reduce the collector voltage by conducting through the load component, it can't make more voltage than the supply line to it.

When the base is more positive than the emitter (assuming NPN transistor, otherwise it has to be more negative) the transistor conducts and the collector voltage drops. When the voltage on the base is more negative, the transistor does not conduct and the collector stays at supply voltage. Note that the signal is inverted, more voltage in means less voltage out.

To amplify a full AC signal so it appears at the collector, the transistor has to be held in partial conduction, in other words with some current flowing all the time. The input signal then makes the conduction less or more but not enough to stop conducting altogether or reach saturation. Saturation is the state where the transistor is as conducting as it can manage and an increase in base current cannot make the collector current go any higher. For example, to get most amplification without clipping the top or bottom of the output waveform, if the supply was 20V, you would set the base current so the collector sat at 10V with no signal present. This would allow it to swing 10V in both directions in sympathy with the input signal.

Brian.

 

[engr_irfan]

Good questions by PG1995 and great replys by betwixt