How to make adjustable voltage regulator using mosfet, voltage controller diy. ( dc to dc adjustable )
This circuit can provide 100V, 10A DC as maximum INPUT. It is adjustable from 0V to 100V DC. this is a very good Circuit. good luck to you.
Components used in this project:–
1. Irfp450 mosfet
2. 50K jar
3. 10A diode
4. 0.22-ohm Resistor
5. 10K
Today, MOSFETs are everywhere. It is a component manufactured in industrial quantities like the others so it has become affordable, like its bipolar precursor. They are easily obtained, a part costs about 10 cents and sometimes barely a penny by buying lots online: $10 for 1000 parts, shipping included, it is close to what is called "electronic dust because if you drop one, it's almost not worth bending down to pick it up.
1. Draw me a MOSFET...
A MOSFET is a component which makes it possible to switch current, ie to let pass more or less electrons between two electrodes according to a third. It can be seen as a switch (like a relay) or a variable resistor.
The maximum switching current depends on the size of the case and other characteristics, provided by the manufacturer. The MOSFET reference is chosen according to the type of application, and there is a very wide variety of models. We will focus on low power applications, with TO92 or SOT23 packages being the cheapest.
The most common references in TO92 housing (with lugs, which can be planted in a solderless plate) are BS170 for the N channel version (equivalent to the NPN), and BS250 for the P channel (the complement of the N channel) . The 2N7000 is another common reference, similar to the BS170, but with a different pinout.
The equivalents in SOT23 (3mm surface mount package) packages are BSS138 (N) and BSS84 (P). At our level, the 2N7002 is more or less equivalent to the BSS138.
Larger and more powerful MOSFETs are available, capable of withstanding hundreds of volts and/or hundreds of amperes. At the opposite extreme, transistors of the CMOS family (miniaturized version of MOSFETs) today constitute the bulk of current computer circuits and a chip can integrate billions of them. Due to their tiny size they can only switch a ridiculously low current and can't handle more than 2V, but in exchange they are super fast.
2. The three-legged beast
A discrete MOSFET transistor, as purchased individually, has three pins, like its bipolar cousin:
The source (equivalent to the emitter of a bipolar transistor) is usually connected directly to the power source.
The drain (equivalent to the collector) is usually connected to the circuit to be switched.
The gate, which is the base, controls the resistance between the source and the drain.
We could therefore almost replace a bipolar by a MOSFET. However, unlike bipolar, no current flows through the gate of a MOSFET*, as it is a type of insulated-gate transistor.
* contrary to what page 90 of Hackable n°7 suggests
3. Anatomy of a MOSFET
The term MOSFET stands for Metal Oxide Semiconductor Field Effect Transistor in English. Let's start by looking at it from the end.
Already this indicates that it is a transistor and it is actually a variable resistor. The word “transistor” is short for “Transfer Resistor” and a bipolar transistor behaves more like a current amplifier.
Field Effect means field effect. It is an electric field that controls the resistance of the component. Much like ancient vacuum tubes, MOSFETs are therefore controlled with a voltage, as opposed to bipolar transistors which are controlled by a current flowing through their base. They therefore have the advantage of consuming less energy and they can amplify with a much higher gain. They dissipate almost no energy when saturated, although it does take some energy to get there.
Metal Oxide Semiconductor represents the succession of layers that make up the transistor. A layer of metal (conductor) is deposited on a very thin layer of oxide (insulator) resting on a semiconductor (usually based on silica doped with impurities). In integrated circuits, the metal layer is now replaced by polycrystalline silicon, but the term Metal has remained.
The thinner the insulating layer, the more intense the electric field at the level of the semiconductor, and therefore the more sensitive the transistor will be and it will switch at a lower voltage. On the other hand, this makes it very sensitive to overvoltages, in particular to static electricity discharges. Insulation is fragile and it often happens that a MOSFET snaps for un yes or for a no!
This can manifest itself as an internal short circuit, for example, or an abnormal resistance. To reduce the chance of circuit destruction, make sure you have observed all possible ESD (ElectroStatic Discharge) precautions. For starters, stay grounded and avoid synthetic clothing and shoes. Their friction can generate incredible tensions, especially in dry weather!
4. Using a MOSFET as a Variable Resistor
Let's take a 2N7000 (Channel N) and connect it as shown in the following drawing.
The voltage between the source and the drain (denoted Vds) is fixed by a laboratory power supply at 10V.
The gate voltage (between the gate and the drain, denoted Vgs) is then varied (by means of another power supply) and the current which is output by the 10V power supply is measured. According to the manufacturer's documentation, we should obtain a curve like this:
If the gate voltage increases too much, the current will become too large, causing the component to heat up. Unlike a bipolar transistor, the resistance of a MOSFET also increases when the device is hot, which increases the dissipation even more and the process risks racing, causing the destruction of the device.
In fact, it is normal for the component to suffer, because this assembly connects it directly between the two terminals of a power supply and all the energy is absorbed by the component. This is called, in other words, a short circuit. Like what you should not stupidly copy the circuits from the datasheets.
5. Use a MOSFET as a switch
Although a MOSFET is able to absorb more or less power (as we will see later), it is preferable that this energy be transmitted to a load that consumes it: another circuit or another component such as a resistor, an LED, an inductor... In this case, the load is connected in series with the drain, as in the following diagram:
Switching with an N channel to 0V is called “Low Side”. N MOSFETs are more efficient and less expensive than P MOSFETs, but it is not always possible or reasonable to cut the 0V, it can create many problems for the rest of the circuit.
To switch the positive supply and keep the common ground, therefore in "High Side" assembly, you must use a P-channel MOSFET, the source of which will be connected to Vcc.
6. The Three MOSFET Regimes
What interests us then is the drain voltage (Vds), because the current is no longer critical, now that the load absorbs the voltage difference (therefore consumes the energy). As long as the MOSFET is much more conductive than the load, the voltage across it remains low, and since the power dissipated is the product of voltage and current, then it is not at risk.
Depending on the voltage between the gate and the source, the MOSFET will then be considered to be in a blocked, ohmic or saturated state. Voltages vary from model to model, even component to component due to manufacturing tolerances, and of course drain load characteristics. The transitions from one state to another are not very precise either, but we speak of these three operating regimes to simplify the calculations.
Blocked regime: at less than 2V, the MOSFET hardly lets any electrons through. The resistance between source and drain is almost infinite. This corresponds to an open switch and the load is no longer powered.
Ohmic or resistive regime: around a few volts, the resistance is inversely proportional to the gate voltage. The MOSFET then behaves like a resistor, adjustable thanks to this voltage. The transistor experiences both voltage and current and can therefore heat up.
Saturated mode: around about 10V, when the increase in the gate voltage no longer translates into a reduction in Vds, the transistor is saturated. The equivalent resistance can be ten or a hundred milliohms depending on the model, some can even reduce their resistance to the point of being considered as a normal conductor. This corresponds to a closed switch and the load is fully energized.
Obviously, the indicative values given above must be checked for each type of MOSFET you will use, some are optimized for very low operating voltages, others for very high currents...
7. Parasitic effects
We have seen that the MOSFET is a three legged beast. Under the hood, however, there is another conductive area in the semiconductor sandwich, which would cause problems if left to float. In most cases, this area is connected to one of the electrodes, which makes the component asymmetric. Without this internal connection, the drain and the source would be completely interchangeable (as on a unijunction FET such as the BF245).
This intrinsic, or parasitic, diode has advantages and disadvantages to keep in mind. It should be remembered that the current can pass in the other direction and disturb an assembly. On the other hand, it can be of service as we will see later.
For the rare cases where current needs to be blocked in both directions, I've never found a 4-pin MOSFET (except for too specific applications in high frequencies), but there is a trick of putting two MOSFETs in the lead - spade, so that their diodes cancel each other out.
7.1 Grid capacity
In electronics, two parallel electrodes separated by an insulator, this is called a capacitor. Due to its structure, a MOSFET contains such a device, between the gate and the semiconductor zone.
This has many effects, more or less desirable, the most important being a sort of memory effect: the MOSFET retains the on or off state if the gate is disconnected. Leakage current is usually negligible and the condition may persist for some time (see Fernand Reynaux's gun barrel sketch for more details).
It is this effect which is taken advantage of in dynamic memories: the gate is charged to a potential (high or low) and the state can be read by measuring the resistance. This type of transistor is extremely small and the leakage current not negligible, so you have to read and recharge the gate several thousand times per second (these are the famous refresh cycles). Most computers use this type of component (often assembled in strips) to constitute their working memory.
The memory effect can create a problem if the MOSFET is controlled by a 3-state output, such as a microcontroller input-output pin. Suppose for a moment that you are powering a circuit through an N MOSFET: its gate will “float” (and pick up any glitches passing by) as long as the MCU's control pin is not in a defined state. This happens at power-up (a few milliseconds before the state is configured by the program) or during a reset phase. Don't be surprised if your assembly is temporarily powered, or even if it stays on!
The solution is to set this default state by adding a small resistor of very high value between the gate and the source. Typically, 1MΩ is enough to shut off the MOSFET when its gate is no longer at a defined state.
With practice, this pull-up resistor becomes almost the counterpart of the essential gate resistor of a bipolar transistor, with the advantage that it is not necessary to calculate it, we simply put "about 1MΩ" .
7.2 Gate Driver for Power MOSFETs
The more current a MOSFET must pass, the larger the semiconductor and gate area. If we count the insulation, this forms a capacitor whose value can reach several nanofarads, and even more if we add the other parasitic effects. For example, the capacitance of the BS170 is around 30pF, but that of the SUM110 can reach 7nF.
For assemblies which switch at very high frequency, it is also necessary to take account of the Miller effect, which results for us in strong currents to charge the grid, during the transitions between the on state and the insulating state. Power transistors, in order to be as efficient as possible and dissipate the least amount of power, must be fully saturated, at voltages of 10V or more. Some can reduce their resistance to a few milliohms, but that requires pumping a lot of electrons to or from the grid. This turns into a strong instantaneous current, because the slower the transition, the longer the transistor will be in the resistive regime and dissipate energy. This is very important for switching power supplies, because the current must be chopped to several tens or hundreds of kilohertz!
Integrated circuits, called "MOSFET driver" or "power-MOSFET gate driver", have been designed to drive the gate with sufficient current peaks to compensate for the Miller effect and improve the switching time. Drivers in the form of integrated circuits are often very efficient and have many advantages, which may justify using them alone.
For example, if you need to control a medium power circuit (less than one amp), with very fast switching (lasting no more than a few tens of nanoseconds), but the operating voltage is higher than your control circuit , a MOSFET driver is ideal. It is a very interesting interface between a low voltage command and a medium voltage load.
8. Slow down switching to smooth transitions
We have already talked about the need to switch a MOSFET as fast as possible. This reduces the losses by decreasing the duration of the ohmic regime. However, extremely fast transitions have other effects, such as the generation of high frequency noise that must be filtered...
The situation is exacerbated when a coil is powered, for example the electromagnet of a relay. Electronically, it is modeled by a strong inductance: switching on is not a concern and the current increases slowly, but stopping creates a dangerous reverse voltage peak.
The tradition is to put a diode across the coil so that the reverse current recirculates in it. This "freewheeling diode" is a solution of the "cure" type instead of "prevent".
Another solution exists: put a capacitor in parallel with the coil, to absorb the peak and smooth the voltage. The capacitor must be sized in proportion to the coil, which increases costs and bulk.
When a coil is driven by a MOSFET, the MOSFET contains a diode that can act as a freewheeling diode, which injects the negative pulse to the power rail. Some power MOSFETs have an intrinsic diode optimized to pass high current, or even a Schottky diode in parallel.
But to prevent, instead of curing, we must think differently. The energy of the reverse current peak is also proportional to the rate of variation of the current. To reduce the peak, it is therefore necessary to slow down the variation. The large capacitor downstream can be replaced by a tiny capacitor upstream at the gate drive.
The slowing down increases the duration of the resistive regime, but it remains transient and therefore quite acceptable for a sufficiently well dimensioned transistor, if the switching is not too repetitive.
The duration of the transition is easy to calculate with the approximate formula t=R×C. In the example above, 1MΩ×100nF = 10^6×10^-7 = approximately 0.1s. These values are available in ultra-miniature and cost-effective SMT components. If the control MOSFET is "big", like the one on the left of the first photo, there is not even a need to add a capacitor in parallel, a very high resistor (10MΩ or more) may suffice.
9. Reverse polarity protection
When you design the power supply for your circuit, you may think of protecting it with a diode, so that no mishandling will damage it. Even with a polarizer, we are not immune to a power supply connected upside down, for example (it does not feel that the experience ...)
But now, the diode will add 0.7V drop to the power supply, so the latter must be sized accordingly. If your power supply is fixed at 5V, you will only have 4.3V available. And this fall is not just in tension: it dissipates power, for nothing.
Fortunately, there are low dropout diodes, in particular Schottky diodes, with lower loss: around 0.5V instead of 0.7, but this changes depending on the models, the reverse voltage supported and of course the fluent. It's a bit better, but it's still not a real solution.
A MOSFET can have almost zero resistance and can replace a diode. The unit cost is higher, but the voltage drop is negligible (if the right parameters are chosen). And the assembly is very simple!
MOSFET_14_Protection
Fig. 14: A MOSFET can also be used as a “super diode”.
This assembly is not very intuitive, but it is very clever. Indeed, it seems to work backwards compared to the assemblies that we have seen so far. The source is at a lower potential than the drain when the supply is plugged in correctly, current flows in the reverse direction of normal use for a P MOSFET.
For better understanding, remember that the drain and the source are interchangeable. The polarity is usually defined by the direction of the intrinsic diode, but it is not a requirement. Consider what happens if power is connected while the circuit is unloaded.
As the capacitor is empty, the voltage Vgs is zero and therefore the transistor is blocked. But if a positive power supply is supplied, the diode allows current to flow, which then charges the capacitor. When the voltage has risen enough, the MOSFET becomes fully on and the load can then draw as much current as it wants, with minimal loss.
On the contrary, if the input voltage is negative, the internal diode does not allow the current to pass. The transistor also does not turn on, because the voltage is in the wrong direction. The circuit is protected as long as the gate voltage does not exceed the limit supported by the component.
The assembly both in the “High Side” version (Channel P, on the illustration above) and in the “Low Side” version with an N channel. Pay attention to the direction of the drain and the source, by orienting the intrinsic diode in the normal direction of the current.
The other disadvantage, in addition to the price, is that it is also necessary to protect the MOSFET from the electrostatic discharges which it is possible to observe on the source, in particular when connecting a power supply which is not at the same ground than the circuit.
Finally, the circuit protects against inversions, but not against a voltage drop: the saturated transistor allows current to flow in the other direction and if the supply drops, it will also drop the voltage of the filter capacitor. More sophisticated electronics are needed to turn the MOSFET into a true ideal diode.
How to make adjustable voltage regulator using mosfet, voltage controller diy. ( dc to dc adjustable )
This circuit can provide 100V, 10A DC as maximum INPUT. It is adjustable from 0V to 100V DC. this is a very good Circuit. good luck to you.
Components used in this project:–
1. Irfp450 mosfet
2. 50K jar
3. 10A diode
4. 0.22-ohm Resistor
5. 10K
Today, MOSFETs are everywhere. It is a component manufactured in industrial quantities like the others so it has become affordable, like its bipolar precursor. They are easily obtained, a part costs about 10 cents and sometimes barely a penny by buying lots online: $10 for 1000 parts, shipping included, it is close to what is called "electronic dust because if you drop one, it's almost not worth bending down to pick it up.
1. Draw me a MOSFET...
A MOSFET is a component which makes it possible to switch current, ie to let pass more or less electrons between two electrodes according to a third. It can be seen as a switch (like a relay) or a variable resistor.
The maximum switching current depends on the size of the case and other characteristics, provided by the manufacturer. The MOSFET reference is chosen according to the type of application, and there is a very wide variety of models. We will focus on low power applications, with TO92 or SOT23 packages being the cheapest.
The most common references in TO92 housing (with lugs, which can be planted in a solderless plate) are BS170 for the N channel version (equivalent to the NPN), and BS250 for the P channel (the complement of the N channel) . The 2N7000 is another common reference, similar to the BS170, but with a different pinout.
The equivalents in SOT23 (3mm surface mount package) packages are BSS138 (N) and BSS84 (P). At our level, the 2N7002 is more or less equivalent to the BSS138.
Larger and more powerful MOSFETs are available, capable of withstanding hundreds of volts and/or hundreds of amperes. At the opposite extreme, transistors of the CMOS family (miniaturized version of MOSFETs) today constitute the bulk of current computer circuits and a chip can integrate billions of them. Due to their tiny size they can only switch a ridiculously low current and can't handle more than 2V, but in exchange they are super fast.
2. The three-legged beast
A discrete MOSFET transistor, as purchased individually, has three pins, like its bipolar cousin:
The source (equivalent to the emitter of a bipolar transistor) is usually connected directly to the power source.
The drain (equivalent to the collector) is usually connected to the circuit to be switched.
The gate, which is the base, controls the resistance between the source and the drain.
We could therefore almost replace a bipolar by a MOSFET. However, unlike bipolar, no current flows through the gate of a MOSFET*, as it is a type of insulated-gate transistor.
* contrary to what page 90 of Hackable n°7 suggests
3. Anatomy of a MOSFET
The term MOSFET stands for Metal Oxide Semiconductor Field Effect Transistor in English. Let's start by looking at it from the end.
Already this indicates that it is a transistor and it is actually a variable resistor. The word “transistor” is short for “Transfer Resistor” and a bipolar transistor behaves more like a current amplifier.
Field Effect means field effect. It is an electric field that controls the resistance of the component. Much like ancient vacuum tubes, MOSFETs are therefore controlled with a voltage, as opposed to bipolar transistors which are controlled by a current flowing through their base. They therefore have the advantage of consuming less energy and they can amplify with a much higher gain. They dissipate almost no energy when saturated, although it does take some energy to get there.
Metal Oxide Semiconductor represents the succession of layers that make up the transistor. A layer of metal (conductor) is deposited on a very thin layer of oxide (insulator) resting on a semiconductor (usually based on silica doped with impurities). In integrated circuits, the metal layer is now replaced by polycrystalline silicon, but the term Metal has remained.
The thinner the insulating layer, the more intense the electric field at the level of the semiconductor, and therefore the more sensitive the transistor will be and it will switch at a lower voltage. On the other hand, this makes it very sensitive to overvoltages, in particular to static electricity discharges. Insulation is fragile and it often happens that a MOSFET snaps for un yes or for a no!
This can manifest itself as an internal short circuit, for example, or an abnormal resistance. To reduce the chance of circuit destruction, make sure you have observed all possible ESD (ElectroStatic Discharge) precautions. For starters, stay grounded and avoid synthetic clothing and shoes. Their friction can generate incredible tensions, especially in dry weather!
4. Using a MOSFET as a Variable Resistor
Let's take a 2N7000 (Channel N) and connect it as shown in the following drawing.
The voltage between the source and the drain (denoted Vds) is fixed by a laboratory power supply at 10V.
The gate voltage (between the gate and the drain, denoted Vgs) is then varied (by means of another power supply) and the current which is output by the 10V power supply is measured. According to the manufacturer's documentation, we should obtain a curve like this:
If the gate voltage increases too much, the current will become too large, causing the component to heat up. Unlike a bipolar transistor, the resistance of a MOSFET also increases when the device is hot, which increases the dissipation even more and the process risks racing, causing the destruction of the device.
In fact, it is normal for the component to suffer, because this assembly connects it directly between the two terminals of a power supply and all the energy is absorbed by the component. This is called, in other words, a short circuit. Like what you should not stupidly copy the circuits from the datasheets.
5. Use a MOSFET as a switch
Although a MOSFET is able to absorb more or less power (as we will see later), it is preferable that this energy be transmitted to a load that consumes it: another circuit or another component such as a resistor, an LED, an inductor... In this case, the load is connected in series with the drain, as in the following diagram:
Switching with an N channel to 0V is called “Low Side”. N MOSFETs are more efficient and less expensive than P MOSFETs, but it is not always possible or reasonable to cut the 0V, it can create many problems for the rest of the circuit.
To switch the positive supply and keep the common ground, therefore in "High Side" assembly, you must use a P-channel MOSFET, the source of which will be connected to Vcc.
6. The Three MOSFET Regimes
What interests us then is the drain voltage (Vds), because the current is no longer critical, now that the load absorbs the voltage difference (therefore consumes the energy). As long as the MOSFET is much more conductive than the load, the voltage across it remains low, and since the power dissipated is the product of voltage and current, then it is not at risk.
Depending on the voltage between the gate and the source, the MOSFET will then be considered to be in a blocked, ohmic or saturated state. Voltages vary from model to model, even component to component due to manufacturing tolerances, and of course drain load characteristics. The transitions from one state to another are not very precise either, but we speak of these three operating regimes to simplify the calculations.
Blocked regime: at less than 2V, the MOSFET hardly lets any electrons through. The resistance between source and drain is almost infinite. This corresponds to an open switch and the load is no longer powered.
Ohmic or resistive regime: around a few volts, the resistance is inversely proportional to the gate voltage. The MOSFET then behaves like a resistor, adjustable thanks to this voltage. The transistor experiences both voltage and current and can therefore heat up.
Saturated mode: around about 10V, when the increase in the gate voltage no longer translates into a reduction in Vds, the transistor is saturated. The equivalent resistance can be ten or a hundred milliohms depending on the model, some can even reduce their resistance to the point of being considered as a normal conductor. This corresponds to a closed switch and the load is fully energized.
Obviously, the indicative values given above must be checked for each type of MOSFET you will use, some are optimized for very low operating voltages, others for very high currents...
7. Parasitic effects
We have seen that the MOSFET is a three legged beast. Under the hood, however, there is another conductive area in the semiconductor sandwich, which would cause problems if left to float. In most cases, this area is connected to one of the electrodes, which makes the component asymmetric. Without this internal connection, the drain and the source would be completely interchangeable (as on a unijunction FET such as the BF245).
This intrinsic, or parasitic, diode has advantages and disadvantages to keep in mind. It should be remembered that the current can pass in the other direction and disturb an assembly. On the other hand, it can be of service as we will see later.
For the rare cases where current needs to be blocked in both directions, I've never found a 4-pin MOSFET (except for too specific applications in high frequencies), but there is a trick of putting two MOSFETs in the lead - spade, so that their diodes cancel each other out.
7.1 Grid capacity
In electronics, two parallel electrodes separated by an insulator, this is called a capacitor. Due to its structure, a MOSFET contains such a device, between the gate and the semiconductor zone.
This has many effects, more or less desirable, the most important being a sort of memory effect: the MOSFET retains the on or off state if the gate is disconnected. Leakage current is usually negligible and the condition may persist for some time (see Fernand Reynaux's gun barrel sketch for more details).
It is this effect which is taken advantage of in dynamic memories: the gate is charged to a potential (high or low) and the state can be read by measuring the resistance. This type of transistor is extremely small and the leakage current not negligible, so you have to read and recharge the gate several thousand times per second (these are the famous refresh cycles). Most computers use this type of component (often assembled in strips) to constitute their working memory.
The memory effect can create a problem if the MOSFET is controlled by a 3-state output, such as a microcontroller input-output pin. Suppose for a moment that you are powering a circuit through an N MOSFET: its gate will “float” (and pick up any glitches passing by) as long as the MCU's control pin is not in a defined state. This happens at power-up (a few milliseconds before the state is configured by the program) or during a reset phase. Don't be surprised if your assembly is temporarily powered, or even if it stays on!
The solution is to set this default state by adding a small resistor of very high value between the gate and the source. Typically, 1MΩ is enough to shut off the MOSFET when its gate is no longer at a defined state.
With practice, this pull-up resistor becomes almost the counterpart of the essential gate resistor of a bipolar transistor, with the advantage that it is not necessary to calculate it, we simply put "about 1MΩ" .
7.2 Gate Driver for Power MOSFETs
The more current a MOSFET must pass, the larger the semiconductor and gate area. If we count the insulation, this forms a capacitor whose value can reach several nanofarads, and even more if we add the other parasitic effects. For example, the capacitance of the BS170 is around 30pF, but that of the SUM110 can reach 7nF.
For assemblies which switch at very high frequency, it is also necessary to take account of the Miller effect, which results for us in strong currents to charge the grid, during the transitions between the on state and the insulating state. Power transistors, in order to be as efficient as possible and dissipate the least amount of power, must be fully saturated, at voltages of 10V or more. Some can reduce their resistance to a few milliohms, but that requires pumping a lot of electrons to or from the grid. This turns into a strong instantaneous current, because the slower the transition, the longer the transistor will be in the resistive regime and dissipate energy. This is very important for switching power supplies, because the current must be chopped to several tens or hundreds of kilohertz!
Integrated circuits, called "MOSFET driver" or "power-MOSFET gate driver", have been designed to drive the gate with sufficient current peaks to compensate for the Miller effect and improve the switching time. Drivers in the form of integrated circuits are often very efficient and have many advantages, which may justify using them alone.
For example, if you need to control a medium power circuit (less than one amp), with very fast switching (lasting no more than a few tens of nanoseconds), but the operating voltage is higher than your control circuit , a MOSFET driver is ideal. It is a very interesting interface between a low voltage command and a medium voltage load.
8. Slow down switching to smooth transitions
We have already talked about the need to switch a MOSFET as fast as possible. This reduces the losses by decreasing the duration of the ohmic regime. However, extremely fast transitions have other effects, such as the generation of high frequency noise that must be filtered...
The situation is exacerbated when a coil is powered, for example the electromagnet of a relay. Electronically, it is modeled by a strong inductance: switching on is not a concern and the current increases slowly, but stopping creates a dangerous reverse voltage peak.
The tradition is to put a diode across the coil so that the reverse current recirculates in it. This "freewheeling diode" is a solution of the "cure" type instead of "prevent".
Another solution exists: put a capacitor in parallel with the coil, to absorb the peak and smooth the voltage. The capacitor must be sized in proportion to the coil, which increases costs and bulk.
When a coil is driven by a MOSFET, the MOSFET contains a diode that can act as a freewheeling diode, which injects the negative pulse to the power rail. Some power MOSFETs have an intrinsic diode optimized to pass high current, or even a Schottky diode in parallel.
But to prevent, instead of curing, we must think differently. The energy of the reverse current peak is also proportional to the rate of variation of the current. To reduce the peak, it is therefore necessary to slow down the variation. The large capacitor downstream can be replaced by a tiny capacitor upstream at the gate drive.
The slowing down increases the duration of the resistive regime, but it remains transient and therefore quite acceptable for a sufficiently well dimensioned transistor, if the switching is not too repetitive.
The duration of the transition is easy to calculate with the approximate formula t=R×C. In the example above, 1MΩ×100nF = 10^6×10^-7 = approximately 0.1s. These values are available in ultra-miniature and cost-effective SMT components. If the control MOSFET is "big", like the one on the left of the first photo, there is not even a need to add a capacitor in parallel, a very high resistor (10MΩ or more) may suffice.
9. Reverse polarity protection
When you design the power supply for your circuit, you may think of protecting it with a diode, so that no mishandling will damage it. Even with a polarizer, we are not immune to a power supply connected upside down, for example (it does not feel that the experience ...)
But now, the diode will add 0.7V drop to the power supply, so the latter must be sized accordingly. If your power supply is fixed at 5V, you will only have 4.3V available. And this fall is not just in tension: it dissipates power, for nothing.
Fortunately, there are low dropout diodes, in particular Schottky diodes, with lower loss: around 0.5V instead of 0.7, but this changes depending on the models, the reverse voltage supported and of course the fluent. It's a bit better, but it's still not a real solution.
A MOSFET can have almost zero resistance and can replace a diode. The unit cost is higher, but the voltage drop is negligible (if the right parameters are chosen). And the assembly is very simple!
MOSFET_14_Protection
Fig. 14: A MOSFET can also be used as a “super diode”.
This assembly is not very intuitive, but it is very clever. Indeed, it seems to work backwards compared to the assemblies that we have seen so far. The source is at a lower potential than the drain when the supply is plugged in correctly, current flows in the reverse direction of normal use for a P MOSFET.
For better understanding, remember that the drain and the source are interchangeable. The polarity is usually defined by the direction of the intrinsic diode, but it is not a requirement. Consider what happens if power is connected while the circuit is unloaded.
As the capacitor is empty, the voltage Vgs is zero and therefore the transistor is blocked. But if a positive power supply is supplied, the diode allows current to flow, which then charges the capacitor. When the voltage has risen enough, the MOSFET becomes fully on and the load can then draw as much current as it wants, with minimal loss.
On the contrary, if the input voltage is negative, the internal diode does not allow the current to pass. The transistor also does not turn on, because the voltage is in the wrong direction. The circuit is protected as long as the gate voltage does not exceed the limit supported by the component.
The assembly both in the “High Side” version (Channel P, on the illustration above) and in the “Low Side” version with an N channel. Pay attention to the direction of the drain and the source, by orienting the intrinsic diode in the normal direction of the current.
The other disadvantage, in addition to the price, is that it is also necessary to protect the MOSFET from the electrostatic discharges which it is possible to observe on the source, in particular when connecting a power supply which is not at the same ground than the circuit.
Finally, the circuit protects against inversions, but not against a voltage drop: the saturated transistor allows current to flow in the other direction and if the supply drops, it will also drop the voltage of the filter capacitor. More sophisticated electronics are needed to turn the MOSFET into a true ideal diode.
No comments:
Post a Comment