Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
Q1: Sure. a micro-controller or a Voltage controlled oscillator (VCO)with feedback.
Q2: As the frequency gets larger, everything gets smaller. transformers, capacitors etc. One of the reasons airplanes use 400 Hz. The parts are smaller than 50 or 60 Hz.
Why? Think about filters and the RC time constant. at 5 * RC the end value is at about 98%. Now let us add frequency, Say 60 Hz and 100 KHz. With a 100 KHz signal the RC network is charged every 100 KHz rather than every 60 Hz. Clearly, the 100 KHz would need less capacitance to maintain a DC value. Think RC discharge curve.
Q1: Yes, or a comparator and op amp to achieve PWM.
Q2: With smaller components, the derivatives get larger, so in a given time the amplitude excursions get higher and that contradicts obtaining a clean output eventually. With faster frequency, the larger derivatives dont have as much time to cause a large excursion in amplitude, so we get a cleaner output. We also obtain faster load step response as well as overall faster response in general.
In reply to Q2, KISS said that as the frequency gets larger, everything gets smaller such as transformers etc. Could you please tell how the size of a transformer is dependent on the frequency used? I'm sorry if I'm missing something obvious. Thanks.
In reply to Q1, MrAl said, "Yes, or a comparator and op amp to achieve PWM". Could you please show me how this circuit will be implemented?
I understand that in n-type material, electrons are majority carriers and the holes are minority carriers. Likewise, in p-type material, the holes are majority carriers and electrons are minority carriers. But how are the majority carrier and minority carrier devices differentiated? For instance, in a majority carrier device, majority carriers could either be electrons or holes.
It says here that a power semiconductor device is usually used in "commutation mode" (i.e., it is either on or off), and therefore has a design optimized for such usage; it should usually not be used in linear operation. What does it mean by "linear operation"? Perhaps, the linear operation refers to the condition where a semiconductor device keeps functioning in same state for prolonged period of time. Please help me. Thanks.
Higher-frequency designs benefit from thinner laminations but also need less core mass to handle the same amount of power. It is the weight savings possible with 400-Hz power that drives use on aircraft.
In reply to Q2, KISS said that as the frequency gets larger, everything gets smaller such as transformers etc. Could you please tell how the size of a transformer is dependent on the frequency used? I'm sorry if I'm missing something obvious. Thanks.
In reply to Q1, MrAl said, "Yes, or a comparator and op amp to achieve PWM". Could you please show me how this circuit will be implemented?
I understand that in n-type material, electrons are majority carriers and the holes are minority carriers. Likewise, in p-type material, the holes are majority carriers and electrons are minority carriers. But how are the majority carrier and minority carrier devices differentiated? For instance, in a majority carrier device, majority carriers could either be electrons or holes.
It says here that a power semiconductor device is usually used in "commutation mode" (i.e., it is either on or off), and therefore has a design optimized for such usage; it should usually not be used in linear operation. What does it mean by "linear operation"? Perhaps, the linear operation refers to the condition where a semiconductor device keeps functioning in same state for prolonged period of time. Please help me. Thanks.
For Q2 again, as you probably know, a core material can only take so much flux density for it's cross sectional area, so to keep the max flux density down with a given winding voltage the core has to have a large enough cross sectional area as the flux density goes down with core cross sectional area:
B=Kn/(A*Kd) (Kn and Kd being dependent on other things)
So the larger the cross sectional area, the less flux density, so by having a large enough area that helps prevent saturation of the core. But increasing the core area means making the core larger and thus the whole construction gets larger and takes up more room on the circuit board.
Luckily, the flux density also goes down with frequency:
B=Kn/(F*A*Kd) (Kn and Kd may be different here)
so the higher the frequency the lower the flux density. This enables us to increase the frequency instead of increasing the core area, which means the construction can handle the required primary voltage without increasing the size of the transformer.
Note the voltage is in the numerator:
B=Kn*E/(F*A*N*Kd) where Kn and Kd are constants different from above
so for a given voltage we need a certain core area and frequency, but increasing the frequency means we can decrease the core area which means we can decrease the total size of the transformer. The equation shown above is usually known as the "Transformer Equation" but there are other ones too.
For Q1 again, a circuit that might be used as the switch is shown in the diagram. Note the switch would be the transistor Q1 here, and the low pass filter is the inductor L1 and capacitor C2.
Also note that the comparator pull up resistors are not shown and they are usually required for the common comparators like the LM339. So for this circuit it is assumed that the output of those comparators go all the way up to the supply voltage when the output goes high.
Also very noteworthy is if we connect the output of X3 directly to the left hand side of R7 and remove X1 and X2, we end up with a purely linear regulator. So the triangle wave and X2 comparator simply chop up the linear drive signal in order to make a switcher out of a linear circuit. The common name for this kind of circuit is the "buck" switching regulator, or "step down" switching regulator.
Could someone please help me with the queries below?
I understand that in n-type material, electrons are majority carriers and the holes are minority carriers. Likewise, in p-type material, the holes are majority carriers and electrons are minority carriers. But how are the majority carrier and minority carrier devices differentiated? For instance, in a majority carrier device, majority carriers could either be electrons or holes.
It says here that a power semiconductor device is usually used in "commutation mode" (i.e., it is either on or off), and therefore has a design optimized for such usage; it should usually not be used in linear operation. What does it mean by "linear operation"? Perhaps, the linear operation refers to the condition where a semiconductor device keeps functioning in same state for prolonged period of time. Please help me. Thanks.
I understand that in n-type material, electrons are majority carriers and the holes are minority carriers. Likewise, in p-type material, the holes are majority carriers and electrons are minority carriers. But how are the majority carrier and minority carrier devices differentiated? For instance, in a majority carrier device, majority carriers could either be electrons or holes.
What you said is kinda circular. You defined n and you defined p and then you describe an entirely new concept called a "majority carrier device"/ The latter might be important in analysis, but n and p type usually rule.
I had two solid state physics courses and in my work I had to measure some of the semiconductor properties.
===
Would you use a "Digital Transistor" as an amplifier? What about an SCR? or a CMOS data selector (not a power device) or an IGBT?
I believe "linear operation" effectively means "amplify" and not the binary ON/OFF operation. It's like using a 100 W audio power amplifier to light a 100 W light bulb at 0 and 100%. Its just not an appropriate use.
When we look at "linear" and "non linear" and "digital" we really have to take a lot into account in order to get an idea what these terms refer to in the general field of electronics.
The definition of linear is "A straight line through the origin" which means a device is linear if the input-output characteristic follows a straight line AND it also passes through the origin. This means if we input a signal x1 and get an output y1, if we input x1 times a constant we get an output that is y1 times that same constant:
input x1 get output y1,
input A*x1 get output A*y1.
So for example if we input 2v and get 10v, if we later input 6v (we multiplied 2v by 3) we get an output that is 30v.
And also, if we input zero we get zero output.
That's the strictest definition, but it is often relaxed to include those devices with an input and/or output offset. So we might input 2v and get 10v,but when we input 6v we get 50v which came about through an offset of 1v:
(2v-1v)=1v times 10 equals 10v, and
(6v-1v)=5v times 10 equals 50v.
We often still call that linear, with the added stipulation that we will be using it with a certain range of inputs and outputs that follow that straight line characteristic.
But there is another interesting point about linear devices that makes them peculiar. That is that they have a rather wide input range and wide output range, very much unlike a digital device. This in itself is often referred to as "linear". Or at least if the device can operate effectively in that way it may be called linear even though it is not really linear as described above. In other words, it may be called linear if for a wide input range it has a wide output range, loosely speaking. If on the other hand it has a very strict output state that can only flip between two or three states, then it could be called digital.
An interesting example is the op amp. Most of us think of the op amp as a linear device. It is a linear device because (for a perfect op amp) if we input 1uv we might get 1v output, and for 2uv input we could get 2v output, which follows the definition of linear pretty closely. But that's not the way it is used especially since the gain can change quite a bit and is not guaranteed to be a certain set value. So we may never see that perfect linear operation using it like that. In fact, it looks like a digital device when we input 0v and get 0v but then input 1v and get 10v (the full supply voltage) as output. That's definitely digital. But when we use feedback we get a pretty nice linear operation. We input 1v and get 2v, and input 2v and get 4v with the resistors set to provide a gain of 2.
So the point is, with the raw op amp it acts more like a comparator then an amplifier for most of the input signal ranges we usually see in electronic circuits. It's only when we add the external components that we get good clean linear operation.
So if for a wide range of inputs we only get certain set outputs (usually 2 or 3 types of outputs only) that is considered digital, but for a wide range of inputs if we see a wide range of output levels then that would be considered a linear device even if it does not follow the strict definition of "linear".
The reason we might not want to use a digital device in the linear mode (it can be done however) is that it takes a lot of gain to get it to act as a linear device.
For a linear device we usually want to input a wide range of inputs and see a wide range of outputs, but for a digital device we might only have to change the input level a tiny amount to see a huge change in output.
For example, if we input 0.5v and get 1v output and then input 1v and get 2v output, that's linear. But if we input 0.5v and get 1v output and then input 0.6v input and get 10v (the full supply voltage) that might be unacceptable as a linear device because the output bangs all the way up to the supply voltage for a small change in input. This is true of a digital device, where we input a certain voltage and get a certain output, but then input a small change and see the output jump to a very different level, and the output levels are well defined as either one or the other.
So another way of looking at it is that a digital device has very set output levels while a linear device has a wide range of outputs. A digital device doesnt make a very good linear device because the levels are always banging between two different extremes.
It is true that we can 'linearize' a digital device, but it takes a lot of gain and so there are only certain cases where this is a good idea, and almost never a good idea in a power stage.
Finally, there is one more constraint that bugs us about linear vs digital. That is the power dissipation in the package. Some modern packages are made to work only in the 'digital' mode (either full on or full off) because in those two modes the power dissipation is minimal, and the package thermal characteristic is such that it can only handle the minimal power dissipated in those two modes alone. Trying to use it in the linear mode and still get the full current rating (or even near it) would cause the device to burn up simply because the package the die is in can not dissipate the heat fast enough to keep it cool. So the power dissipation factor is another constraint that means we simply can not use certain packages in linear mode no matter how bad we want to unless we severely limit their current levels which could be very much under their usual rating.
For a given cross section there comes a point when increasing the current any further doesn't increase the flux. I think of this like this. We can think there are very tiny magnets which generate this flux. After every half cycle of AC these tiny magnets get rotated. When current is increased, more and more tiny magnets get aligned in the same direction and contribute to the flux. But there comes a stage when almost all tiny magnets are aligned and increasing current won't do any good. This is saturation point.
I was thinking that how increasing the frequency lowers the flux. Let's try to understand this. Those tiny magnets need to rotate every half cycle of AC which means they will take certain amount of time for this rotation. Therefore, as frequency is increased, the number of tiny magnets which are able to rotate and contribute to the flux decreases. It means less flux with increasing frequency. Do I make any sense? Please let me know. Thanks.
You can use that analogy but the actual workings of the magnetism of a material is pretty complicated and i believe is still under investigation today. I think some materials are considered to have domain flips (small domains) and others are considered as having domain wall flexing (larger domains) as the basic mechanics behind the magnetic polarization.
There is at least one magnetic core model in use today that takes domain wall flexing and pinning as parameters, and that is used for example for power inductors.
The basis for all these is electron flipping, but i think that happens way too fast to consider as the sole reason for frequency response. Maybe it can be explained as that plus a domino effect where the domino effect means there will be a lot of flip times to sum per domain before the domain changes as much as it can given the present level of excitation (wall flex or other wall movement), but of course i cant be sure of this.
It's possible to flip very small domains in some materials in 1 picosecond using controlled laser light, and this might be incorporated into future computer memory similar to the rotating hard drives of today but instead of using a magnetic head it would use a laser head.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
