Fullscreen HTML5Med Res (??MB)HTML4

Circuit Surgery Regular clinic by Ian Bell Ideal diode integrated circuits L ast month, inspired by a question on EEWeb forum, we looked at diode power circuits and the ‘ideal diode’ circuits which can be used in place of diodes in these applications. We mainly covered the basic principles. This month, we follow on by exploring in more depth some of the ICs which are used to implement ideal diode power circuits. Diode power circuits To recap, the diode power circuits which we were discussing last month are used for reverse-power protection and diode ORing of power supplies – see Fig.1 and Fig.2 respectively. Reverse-power protection uses the fact that diodes can only conduct in one direction to prevent circuits from being damaged by reversal of the supply voltage. This easily occurs due to user installation error in battery-powered systems. Another possible reversepower scenario is transient reverse voltage due to inductive load switching on the same supply (a common problem in automotive electronics). Diode-OR circuits allow multiple supplies to be connected to the same circuit. If only one supply is connected then its associated diode will conduct, and that circuit will behave in the same way as with a single reverse protection diode. If both supplies are present, with one sufficiently higher than the other, then the higher supply will provide all the power to the load. If both supplies have very similar voltages, then both supplies can deliver power to the load (load sharing). VF IF Supply VSupply VCir Ideal diode controllers As discussed last month, a MOSFET can be used in place of a diode for reverse protection, with the advantage of lower voltage drop and power dissipation. However, MOSFETS may conduct in the reverse direction in certain situations, which can cause significant problems in some applications. Many circuits require reverse-current protection as well as reverse-supply-voltage protection, and we will discuss this in more detail later. The diode-OR circuit can also make use of MOSFETs, but this requires appropriate gate control to switch the individual MOSFETS correctly (or balance operation in load sharing). With appropriate control, a MOSFET can behave in a manner close to an ideal diode in a reverse-voltage protection or diode ORing circuit. By ‘ideal diode’ we mean, low voltage-drop and low power dissipation compared with using standard diodes; reverse-current protection; and accurate switching between conducting and non-conducting states. The circuitry providing this functionality is called an ‘ideal diode controller’ and numerous ICs are available which provide this Supply2 + D1 Conducting diodes have a forward voltage drop (VF) and diode current (IF), therefore they dissipate power VFIF, which can be a problem in terms of both energy loss and heat management. The diode voltage drop can also reduce effective battery life. These problems can be reduced if Schottky diodes are used instead of standard silicon diodes, due to their lower voltage drop at comparable currents, but this does not eliminate the problem. Supply1 D2 VSupply2 44 The LTC4358 IC is a ‘5A Ideal Diode’ device developed by Linear Technology (now Analogue Devices). It is called an ‘ideal diode’ rather than an ‘ideal diode controller’ because it has an internal MOSFET. The word controller refers to control of an external MOSFET in the terminology of these devices. The LTC4357 is a similar device which uses an external MOSFET and is described as a ‘Positive High Voltage Ideal Diode Controller’. Both devices are aimed at diode-OR applications for redundant power supplies and can regulate load sharing. The LTC4358 provides reversecurrent protection, but does not provide reverse-voltage input protection. We will look at the LTC4358 in detail and use some simplified simulations to help explain its operation. A block diagram of the LTC4358 is shown in Fig.4. The In, Out and Gnd pins correspond with the generic ideal diode controller in Fig.3. The Drain pin is connected to the internal MOSFET drain and corresponds with the MOSFET drain in Fig.3. This must be connected to the Out pin. The drain is connected to an exposed pad on the base of the IC package (see Fig.5) and is used to conduct heat to PCB copper, which acts as a heat sink. The Gate pin corresponds with Supply + VSupply1 VCir Circuit – Fig.1. Circuit with reverse-voltage protection diode. LTC4358 ideal diode M1 D1 Circuit – facility. They vary in their specific target applications, features and capabilities. A generic ideal diode circuit is shown in Fig.3, this is equivalent to the circuit in Fig.1. The diode symbol in Fig.3 represents the body diode of MOSFET M1, not a diode component. Fig.2. Diode-OR power source circuit. In Gate Out Ideal diode controller Control Gnd Status + Circuit – Fig.3. Circuit demonstrating ideal diode controller concept. Practical Electronics | August | 2023 Drain Gate In Out Charge pump + + – – 25mV Gate Amp VDD + – FPG Comp In + 25mV – Gnd Fig.4. LTC4358 block diagram. (Based on Analog Devices/LT datasheet) Fig.6. LTspice schematic of Diode-OR circuit using two LTC4358 ideal diode ICs. Based on the example provided with LTspice. constantly switching between these two arrangements the the Gate pin in Fig.4, but as the MOSFET is internal this pin second capacitor’s charge is ‘pumped up’ to provide (and is not normally connected to anything externally. However, maintain) twice the supply voltage. As long as any current it could be used to monitor the gate control. The VDD pin is taken from the second capacitor is small compared to the the device’s power supply (which is not shown in Fig.3); it is available average ‘pump up’ current it can be used as a power typically connected directly to the Out pin. supply for parts of the circuit that need it. The LTC4358’s block diagram in Fig.4 shows two subcircuits involved in the control of the MOSFET via its gate voltage. During normal operation, when the MOSET is conducting, LTC4358 operation the Gate Amplifier is used to regulate the voltage across the Before looking at the internal subcircuits of the FLTC4358 in MOSFET to be 25mV where possible. The FPD Comparator more detail we will simulate its normal operation as LTspice (Fast Pull Down Comparator) is used to rapidly switch off includes a model of the devices as part of the current download. the MOSFET in situations where reverse current would Fig.6 shows an LTspice schematic of a basic diode-OR circuit otherwise flow through the MOSFET. (This provides the using the LTC4358. This is similar to the test fixture example reverse-current protection mentioned above). We will look at provided with LTspice. The simulation ramps the voltages simplified simulations of the Gate Amplifier and FPD Comparator subsystems separately to help explain operation of the device. To switch on the MOSFET, the voltage at its gate must be positive with respect to its source, so, given that its source and drain are at, or close to, the source supply voltage when the MOSFET is conducting, the gate voltage, and hence the output of the Gate Amplifier, must be higher than the supply voltage. This means that the Gate Amplifier requires a higher supply voltage than the source supply. This is achieved in the LTC4358 (and many similar devices) by using a charge pump circuit (as shown on Fig.4). Charge pumps work by charging and switching capacitors. For example, if a capacitor is charged up to the supply TOP VIEW voltage and then IN 1 16 IN switched so that what was the ground end IN 2 15 IN is now connected to IN 3 14 IN the positive supply, IN 4 13 IN 17 then the other side DRAIN NC 5 12 IN of the capacitor GATE 6 11 NC (previously connected to the supply) will be NC 7 10 OUT at twice the supply GND 8 9 VDD voltage relative to ground. This voltage Fig.5. LTC4358 pin can be used to charge configuration for 16-pin another capacitor to TSSOP package. (Based on (towards) twice the Analog Devices/LT datasheet) supply voltage. By Fig.7. Simulation results from the circuit in Fig.6. Practical Electronics | August | 2023 45 Fig.8. LTspice circuit to model operation of LTC4358 gate amplifier. on the two inputs between 0 and 12V in opposite directions. The results are shown in Fig.7, where it can be seen that the output voltage (middle plot) is equal to the higher of the two input voltages (upper plot). The current taken from the two inputs is shown in the lower plot – current is taken from the input with the highest voltage and varies in proportion to the supplied voltage because the load is a fixed resistance. The simulation takes the highest input down to 6V, which is below the minimum specified operating voltage of 9V, but still shows correct operation. Real devices (and/or their simulation models) may operate at lower than specified voltages – however, the manufacturer will not guarantee correct operation or full performance outside operating ranges specified on the datasheet. Gate amplifier Now we look at the operation of the Gate Amplifier in conjunction with the MOSFET in more detail. As discussed last month, MOSFETS used in ideal diode circuits are connected so that the body diode (see Fig.4) conducts when the power source input voltage is first connected This provides the initial output voltage and supply for the LTC4358. Once the supply is established, the Gate Amplifier will regulate the voltage across the MOSFET to be 25mV, so for a single device (not necessarily in a diode-OR circuit) the output voltage will be 25mV below the input voltage as long as the input voltage is within the operating range of the device, which is 9V to 26.5V. Fig.9. Selecting an operating point current to display. 46 Fig.8 shows an LTspice schematic of the LTC4358 Gate Amplifier subsystem based on the block diagram in Fig.4. The circuit does not attempt to accurately represent the full details of the LTC4358’s internal circuitry (which we do not know) but is sufficient to demonstrate principles of operation. The op amp is modelled using LTspice’s UniversalOpAmp2, which is a generic, semi-ideal model. The MOSFET is an IRF2204, which is a high-current, power MOSFET, but which was not chosen for any very specific reason – this is a simplified model, not a circuit design. Many n-channel MOSFETS will work here but will show different gate voltages depending on their characteristics and are not likely to exactly match the LTC4358’s MOSFET (for which we do not have a model). Resistor R1 represents the load – the circuit that is being supplied via the LTC4358. The charge pump is simply modelled as a voltage source (V3) at about twice the source supply (V1) voltage. We need an operating point simulation (.op command) to obtain the voltages (and currents) in the circuit. LTspice produces these results as a text file in a pop-up window. For the circuit in Fig.8 we obtain: V(gate): V(chargepump): V(in): V(inv): V(out): Id(M1): Ig(M1): Is(M1): I(R1): I(V3): I(V2): I(V1): Ix(u1:1): Ix(u1:2): Ix(u1:3): Ix(u1:4): Ix(u1:5): 14.0804 20 10 10 9.975 -0.9975 -6.24878e-009 0.9975 0.9975 -0.000491457 1.00026e-008 -0.997009 -1e-008 -1e-008 0.000491457 -0.000491428 2.61943e-016 voltage voltage voltage voltage voltage device_current device_current device_current device_current device_current device_current device_current subckt_current subckt_current subckt_current subckt_current subckt_current LTspice operating point labels It is often more convenient to show the operating point values of interest on the schematic (as in Fig.8). To do this, run the operating point simulation, close the pop-up window and then right-click on a wire on the schematic and choose Place .op Data Label from the menu. When a simulation is run again after this the values on the schematic will change to ‘???’ and the pop-up window will appear as before. After closing the window, the numerical values will be displayed on the schematic. As can be seen in Fig.8, the text might take up a fair amount of space and some adjustment of a schematic layout may be required to obtain a clean display of the values. The above instructions apply to voltages, but displaying operating point currents on the schematic is more tricky and non-obvious. Here’s how you do it. Right-click on the schematic background and from the menu select: Draft > .op Data Label. Click on the schematic and place the ‘???’ text where you want the data displayed Right-click on the ‘???’ text. In the dialog window that appears (see Fig.9) delete the $ character from the ‘expression’ box and then click on the data item from the list that you would like to display. In this case Id(M1), the drain current for MOSFET M1, is selected. Unlike voltages written next to wires it is not necessarily obvious what the numerical value represents when looking at the schematic, so adding comment text (‘Aa’ button on the menu bar) next to the value may be a good idea (as has been done for the MOSFET current in Fig.8). The current value for Id(M1) is reported as negative because it is defined as the current into the drain, but it is actually flowing out of the drain. Practical Electronics | August | 2023 Gate amplifier simulation Fig.10. LTspice simulation of load resistance sweep for LTC4358 gate amplifier subsystem model. Fig.11. Results from simulation of circuit in Fig.10 Practical Electronics | August | 2023 Looking at the results in Fig.8, we see that the source supply (VIn from V1) is 10V. The charge pump supply is at 20V (VChargePump from V3). Thus, the Gate Amplifier (U1) operates on a 10V supply – the difference between VIn on its negative supply and VChargePump on its positive supply. The Gate Amplifier is wired as a control loop – the negative feedback from the LTC4358’s output (node Out) to its inverting input (node Inv) tries to maintain zero volts between the amplifier’s inputs. The noninverting input is at VIn (10V), so the inverting input voltage (Vinv) will also be at this voltage, or very close to it when the Gate Amplifier is actively controlling the circuit (as seen on Fig.8). Controlling Vinv to be equal to VIn means that Vout will be at VIn – 25mV by virtue of the 25mV source (V2). This is the same as saying the voltage across the MOSFET (source to drain) is controlled to be 25mV. The value of Vout is equal to R1IL, where IL is the load current through load (R1). Current into the Gate Amplifier input is very small, so we can assume IL is equal to the current through the MOSFET. Thus, the output voltage is controlled by the MOSFET current, which is in turn controlled by its gate-source voltage (VGS). The Gate Amplifier sets VGS on the MOSFET to whatever value is required for a 25mV drop between source and drain, in this case it is about 14.1V (with respect to ground), so VGS = 4.1V. This value would be different if a different MOSFET was used, and will change under different conditions. The specific value of VGS does not matter as long as it is within the range that the Gate Amplifier can output and does not exceed the MOSFET’s operating range. We can simulate the behaviour of the LTC4358 Gate Amplifier under varying load conditions using a parametric sweep. The specific values seen here do not necessarily match the actual device – this is a simplified, illustrative model with different component parameters. The setup for this simulation is shown in Fig.10. The schematic is the same as Fig.8, except that we are not showing the current and voltage values and have replaced the fixed 10Ω value of R1 with the parameter {Rload}. The simulation varies R1 from 0.2Ω to 4Ω in 0.1Ω steps, as defined using the .step command. Relatively low Rload values are used to highlight the limit of regulating the MOSFET voltage drop. The results are plotted as circuit operating point voltages and currents against load resistance, as shown in Fig.11. For relatively high load resistance values (on the right-hand side of Fig.11, for values above about 1.2Ω), the circuit is able to regulate the voltage across the MOSFET. The top plane shows both the fixed 10V input voltage, and the output voltage, which is also more or less constant for larger load resistance values. The second pane shows the difference between the input and output 47 25mV. The maximum value of VGS is limited by the power supply to the Gate Amplifier, which is 10V relative to VIn in this example. For load resistances below 1.2Ω the gate voltage would have to exceed 10V, which it is unable to do, so the regulation is lost. The load current (bottom plane) varies with load resistance and is equal to Vout/Rload as would be expected (from Ohm’s law). FPD comparator Fig.12. LTspice circuit to model operation of LTC4358 FPD comparator – normal operation scenario. Fig.13. LTspice circuit to model operation of LTC4358 FPD comparator – reversecurrent scenario. Fig.14. LTspice circuit to model operation of LTC4358 FPD comparator. Potential reverse-current detection used to switch off MOSFET. voltages, which for larger resistance values is constant at 25mV, as expected from the discussion above. Over the load range for which the MOSFET drop is regulated (above about 48 1.2Ω) the gate voltage varies with varying load resistance. As the load resistance decreases, the MOSFET VGS, which is shown in the third pane, has to increase to maintain the drain-source voltage at Fig.12 shows an LTspice schematic of the LTC4358 FPD Comparator subsystem based on the block diagram in Fig.4. This version of the schematic does not have the output of the comparator connected to the MOSFET so that we can observe the MOSFET reverse-current situation which the comparator is used to overcome. To simplify the simulation, we have replaced the Gate Amplifier with a fixed gate-source voltage of 4V provided by V6. In order to observe a reverse-current situation, we need another voltage source at the output. In a real circuit this could come from another LTC4358 in a diode-OR circuit or a capacitor charged to the output voltage. Here we use a voltage source (V5) and diode (D1) as a simple model for another LTC4358 and its source supply. The situation shown in Fig.12 is similar to that in Fig.8. The V1 source is providing the power to the load resistor. The additional supply (V5) is effectively disconnected because the diode (D1) is not conducting (only microamps flowing through it ). The FPD Comparator is set up to switch to a logic 1 output at a voltage 25mV above V In. This will not occur when the LTC4358’s internal MOSFET is conducting (the voltage on V out is 25mV below VIn under these conditions). Therefore, in the scenario in Fig.12 the output of the FPD Comparator is at logic zero (close to ground voltage) and (if it was connected) it would not affect the operation of the device. Fig.13 shows the same circuit as Fig.12 with the source supply (V1) reduced to 9V. This means that the additional supply (V5) is providing current to the load (R1) via diode D1. The 4V VGS is still applied to the MOSFET, so it still conducts, but current flows in the opposite direction (note the change in sign of the M1 drain current compared with Fig.12). A current of about 1A is flowing from the additional supply to the source supply of the LTC4358 via Simulation files Most, but not every month, LTSpice is used to support descriptions and analysis in Circuit Surgery. The examples and files are available for download from the PE website. Practical Electronics | August | 2023 In VOut 12V 5A Drain Out LTC4358 Gnd VDD R1 100Ω CLoad 2.6µA SHDN Shutdown C1 100nF MMBD1205 the MOSFET. This is the reverse current condition that must be avoided to provide correct ideal diode behaviour. In this scenario the output voltage (Vout) is above the input (VIn) by about 31mV. This is greater than the 25mV-above-VIn threshold set for the comparator, so its output is a logic one (equal to its positive supply voltage). Reverse current protection The above two scenarios show that the comparator output indicates the direction of current flow through the MOSFET. Its output can therefore be used to switch off the MOSFET under reverse-current conditions. This is achieved using an NPN transistor connected between the gate and source of the MOSFET (see Fig.4). When the comparator output is logic 1 the NPN transistor switched on, effectively shorting the gate of the MOSFET to its source and reducing the gate-source voltage below threshold. This switches the MOSFET off and prevents reverse current flow. Fig.14 shows the same circuit as in Fig.12 and Fig.13, but with the NPN transistor (Q1) added. The voltage sources are configured in the same way as shown in Fig.13 – the reversecurrent scenario. The switching off of the MOSFET by the FPD comparator via Q1 can be seen – the MOSFET’s drain-source current is effectively zero (nanoamps). The current from the additional supply via D1 is reduced as it is now only providing the load current in R1. The current in Q1 shown in Fig.14 is not meant to be representative of the actual LTC4358 circuit – it is mainly determined by the source resistance value of V6, which prevents ideal-source behaviour resulting in totally unrealistic currents. The above simulations show the principles of operation of the LTC4358 in regulating the MOSFET voltage drop and preventing reverse currents. In a real circuit the reverse current condition is caused by the drop in the source supply voltage to the LTC4358. The additional supply (V5 and D1 in the above examples) will be due to either capacitors connected to the LTC4358 output or other supplies in a diode-OR arrangement. The drop in source supply may occur very rapidly. The LTC4358’s datasheet states that the ideal diode is turned off within 500ns of a reversal, preventing the reverse current rising to a damaging level and minimising any disturbance on the output. Reverse input protection As was noted earlier, the LTC4358 does not provide reverseinput protection. The datasheet suggests the circuit shown in Practical Electronics | August | 2023 Negative Comp 30mV + – Gate Amp – Fig.16. (right) The LTC4358 includes negativeinput detection and an external shutdown control pin. (Based on Analog Devices/LT datasheet) Charge pump f = 500kHz + + Fig.15. (above) Providing reverse-input protection for the LTC4358. (Based on Analog Devices/LT datasheet) – –1.7V + In Out Gate – VIn 12V Source In + 30mV – VSS Fig.15 to provide protection. The RC circuit (R1 and C1) hold up the supply to the LTC4358 during supply disturbances, which helps it deliver fast switch-off. Other devices similar to the LTC4358 do provide more reverse protection, for example the LTC4359 has a comparator to detect reverseinput conditions and switch off the external MOSFET (see block diagram Fig.16). The external MOSFET is not shown in Fig.16, the MOSFETs in the block diagram are used in the same way as the NPN transistor in Fig.4 to switch of the external power MOSFET. The LTC4359 also features an external pin which can be used to switch off the device. ESR Electronic Components Ltd All of our stock is RoHS compliant and CE approved. Visit our well stocked shop for all of your requirements or order on-line. We can help and advise with your enquiry, from design to construction. 3D Printing • Cable • CCTV • Connectors • Components • Enclosures • Fans • Fuses • Hardware • Lamps • LED’s • Leads • Loudspeakers • Panel Meters • PCB Production • Power Supplies • Relays • Resistors • Semiconductors • Soldering Irons • Switches • Test Equipment • Transformers and so much more… Monday to Friday 08:30 - 17.00, Saturday 08:30 - 15:30 Station Road Cullercoats North Shields Tyne & Wear NE30 4PQ Tel: 0191 2514363 sales<at>esr.co.uk www.esr.co.uk 49