Broadening USB capabilities for reliable power delivery in portable devices-EEWORLD

Collect

1. Introduction

USB2.0 is a new PC interconnection protocol that makes the connection of external devices to computers more efficient and convenient. This interface is suitable for a variety of devices. It is not only fast, plug-in and ready, and supports hot dialing, but can also connect 127 devices at the same time, solving problems such as resource conflicts, interrupt requests and direct data channels, providing users with better performance. However, with the increase in the variety of digital portable devices and their core power supplies, it is necessary to further develop and expand the power supply function of the USB interface to ensure the normal and reliable operation of various portable devices. Why is this?

* The need for multiple core power supplies

As we all know, many digital portable devices (such as digital cameras, MP3 players and PDAs) today use DSP chips for digital signal processing. The DSP chip is powered by a single 5V power supply, but some are powered by 3.3V and 1.8V power supplies. The core voltage of the DSP chip is powered by 1.8V power supply, and the I/O is powered by 3.3V power supply. It is unrealistic to add power supply devices to these portable devices.

* Data exchange of portable devices

Because many portable devices, such as MP3 players, PDAs, etc., need to exchange data with PCs, if the battery can be charged using the same cable while exchanging data with PCs, it will greatly facilitate the use of the device. If the power supply function of the USB interface is combined with the battery charging function, a large number of devices can be freed from the constraints of power lines, such as mobile network cameras, which can work whether they are connected to a PC or not. In many cases, the clumsy AC adapters that are always used are no longer needed. It can be seen that the development of USB interface power supply (hereinafter referred to as USB power supply) should solve two major problems. The first is to obtain 5V, 3.3V, and 1.8V power from the USB port; the second is to charge the battery through the USB power supply. In

addition to the function of directly supplying power to the USB-connected devices, one of the most useful functions of the USB interface power supply is the ability to charge the battery. So where do these functions come from? To this end, we should first introduce the technical support for the power supply function of the USB interface (hereinafter referred to as USB power supply).

2. Develop technical support to expand the USB power supply function

The hub controller function and peripheral device function of USB are the technical support for the development of the USB power supply function.

So what are the hub controller function and peripheral device function of USB? Or what is the technical basis for having these two major functions?

2.1 USB Hub Function

(1) Let's first talk about what a hub is. There are three main building blocks in a USB network. First, there is a host in the personal computer. The function of the "host" is just as its name implies, it is mainly responsible for communication requests, receiving and inputting and outputting service information to the computer. Therefore, the host is responsible for handling most of the details of the network "aggregation". The other end of the network is called: "Device", which is a computer peripheral device; in addition, there may be a "hub" between the host and the device. These seemingly simple units can provide four (or seven) slots for devices to use, and only rely on a cable from the computer. The hub is responsible for introducing new "guest" devices to the host, and is also responsible for ensuring that all information between the host and the device is exchanged at the highest speed.

It can be seen that the function of the USB hub is to manage the connection/disconnection operations of the port, including hub configuration, detection of downstream port devices (regardless of whether the port is in suspend mode or resume mode), various states of the port, as well as bus failure and reconfiguration, power management and speed detection support.
The number of ports on the hub controller determines the number of downstream devices that can be managed. Products available to users include two-port type, two or three-port type, four-port type, one to five-port type and seven-port type.

(2) Classification of USB hubs.

* Self-powered USB hubs use local power to power downstream ports. However, the USB interface is allowed to absorb 100mA of current from its upstream port, which can be used to maintain the function of the USB interface while the rest of the hub is powered off. Self-powered USB hubs are required to limit and notify overcurrent conditions and must provide at least 500mA of current for each downstream port.

* A bus-driven USB hub draws all its power from the upstream connection and is required to provide 100mA to each downstream port. For configuration reasons, the USB specification limits the current drawn from the bus by a bus-driven USB hub to 100mA (or less) during power-up. Thereafter, a bus-driven USB hub is allowed to consume 500mA, providing 100mA to each downstream port, with the remainder for the hub itself. Because a self-powered hub may experience local power loss due to power loss or battery depletion, the bus controller can force itself to be recalculated as a bus-driven USB hub, requiring it to perform port power forwarding on all external ports.

2.2 USB Peripheral Functionality

A USB peripheral controller implements the USB connection between a peripheral and a host or hub. Unlike a hub or host, a USB peripheral does not support downstream functionality, but does have upstream-facing ports that must comply with the USB specification. USB peripherals can be divided into the following categories:

(1) Low-power, bus-driven USB functions, which draw all power from the upstream bus and limit the current consumption to 100mA;

(2) High-power bus-driven USB functions, which draw all power from the upstream bus and limit the current consumption during configuration to 100mA (maximum). After calculation, it can draw up to 500mA;

(3) Self-powered USB hub functions, which draw power from the power supply (rather than the upstream bus). It is allowed to draw up to 100mA from the upstream bus, but it is not required to do so.

There are two types of sockets for USB peripherals, both of which are smaller than the sockets for PCs and ordinary USB hosts. The "B series" and the smaller "Mini-B series" sockets are shown in Figure 1. The B series is powered by pins 1 (+5V) and pin 4 (GND), and the Mini-B series is powered by pins 1 (+5V) and pin 5 (GND).

Because the function of charging the battery of the developed USB power supply may be complex or simple, it depends on the requirements of the device connected to the USB interface. Factors affecting the design include not only the usual cost, size, weight, etc., but also the following important factors: when the load device with exhausted battery is plugged into the USB port, how quickly it is required to start full function operation; the time allowed for battery charging; the power distribution within the USB power limit; whether an AC adapter is required for charging. These problems and the corresponding solutions to be solved will be studied after discussing the power problem of USB. For this reason, the USB power problem should be further explained in technical support.

2.3 About USB power

All USB interface hosts, such as PCs and laptops, each USB jack can support a minimum current output of 500mA or drive 5 "unit loads". In USB terminology, "one unit load" is 100mA. Self-powered USB hubs can also drive 5 unit loads. Bus-driven USB hubs can only guarantee the driving of one unit load. According to the USB specification, the minimum available voltage at the peripheral end of the cable provided by the USB host or self-powered USB hub is 4.5V, and the minimum voltage provided by the bus-driven USB hub is 4.35V. There is only a small margin when charging Li+ batteries with a typical battery voltage requirement of 4.2V with these voltages, making the voltage drop of the charger extremely important.

It should be said that all devices connected to the USB port must consume no more than 100mA when started. Only after communicating with the host can the device determine whether it can use the full 500mA current.

There are two types of sockets for USB peripherals, both of which are smaller than the sockets for PCs and ordinary USB hosts. The "B series" and the smaller "Mini-B series" sockets are shown in Figure 1. The B series is powered by pins 1 (+5V) and pin 4 (GND), and the Mini-B series is powered by pins 1 (+5V) and pin 5 (GND).

Once connected to the host, all devices connected to the USB interface must first allow the host to identify themselves. This action is called enumeration. During the identification process, the host decides to accept or reject the power request of the USB device. If accepted, the current of the device can be increased from a maximum of 100mA to a maximum of 500mA.

3. Provide 3.3V and 5V power from the USB port

Since the USB port is a new type of hot-swappable interface, in addition to rich interface functions, it can also provide two types of power. First, the power provided by the low-power USB port is 4.4V-5.25V and the current is 100MA. Second, the power provided by the high-power USB port is 4.75V to 5.25V and the current is 500MA. Portable devices all use USB ports as interfaces, so the combination of USB ports and chips can generate 3.3V and 5V power. In order to meet the needs of DSP chips in portable devices, the design scheme is introduced.

(1) In addition to the communication channels D+ and D- (see the pins of the USB port on the left of Figure 2), the universal serial bus (USB) port can also provide power. When a portable device (such as a digital camera, MP3 player, and PDA) is powered by a battery and connected to a USB port for communication, the USB power supply can be used to charge the Li+ battery.

As shown in Figure 2, it is a circuit diagram that uses USB power to generate 3.3V and 5V power and charge Li+ batteries. Its IC1 chip MAX1811 is a charger for Li+ batteries. When using USB port power to charge the battery, for low-power USB ports, the SETI end of the IC1 chip should be pulled low and its charging current should be set to 100mA; for high-power USB ports, the SETI of the IC1 chip should be set high and the charging current should be set to 500MA. Similarly, when SELV is set high or low, the IC1 chip is configured to charge the Li+ battery to 4.2V or 4.1V, and the final charging voltage of IC1 reaches 0.5% accuracy. The CHG end allows the chip to light up the LED during charging.

(2) As shown in Figure 3, IC2 is a step-up DC-DC converter. The IC2 chip MAX1797 can boost the battery voltage Vbatt to 5V (Vout). And Vout can output 450MA current to the load.

Its low battery detection circuit and true shutdown capability will protect the Li+ battery from over-discharge. This "true" shutdown function reduces the battery to 2uA by disconnecting the battery and the output. The low battery voltage threshold (LB1 pin level) is set by the external resistor divider R3 and R4 between Vbatt and GND (the midpoint of the divider is connected to the LBI pin). Connect the low battery voltage output pin LBO to the shutdown pin SHDN. This causes IC2 to disconnect from the load under low battery voltage conditions. When
the low battery detection circuit disconnects the low voltage battery from the load, the internal resistance of the Li+ battery will make IC2 prone to oscillation. This is because when the voltage drop caused by the battery internal resistance disappears, the battery voltage will increase, causing IC2 to turn on again. For example, a Li+ battery with an internal resistance of 0.5Ω will produce a voltage drop of 250MV on its internal resistance when sourcing a current of 500MA. When the IC2 circuit disconnects the load, the battery current will drop to 0 and the battery voltage will increase by 250MV. To this end, by introducing hysteresis in the low battery detection circuit, the N-channel FETMOS tube at the LBO end will eliminate this oscillation.

The low battery threshold voltage of the circuit in Figure 3 is set to 2.9V. When Vbatt drops below 2.9V, LBO turns on (the level increases), pulls SHDN high, and the FETMOS tube turns on. When the FETMOS tube is turned on, resistors R5 (1.3MΩ) and R4 (249KΩ) form a parallel circuit, which increases the battery's turn-on voltage threshold (pin LB1) to 3.3V, thereby eliminating the oscillation.

(3) As shown in Figure 4, IC3 chip MAX1837 is a DC-DC step-down converter that can step down the 5V output to 3.3V and can output up to 250MA current to the load with an efficiency of over 90%.

From Figures 2, 3, and 4 above, it can be seen that when power is provided through the USB port, the MAX1811 MAX1797 MAX1837 chips can be used to generate 5V to 3.3V power for portable devices.

It should be noted that the USB power supply:

* Due to the voltage drop on the USB cable and connector, the USB device must be able to work up to 4.35V

* The USB device must ensure that its maximum operating current is less than 100MA until it is configured as high power through software.

4. Simple USB power and AC adapter charging configuration

4.1 Configuration scheme

For some of the most basic device loads, there is no need to use software overhead to manage and optimize the use of USB power. If the device load current is limited to less than 100mA, it can be driven by a host connected to the USB interface and a hub with its own power supply or a bus-driven hub. Based on this, this type of simple USB and AC adapter charging design can adopt the configuration of a basic charger plus a regulator as shown in Figure 5.

4.2 Solution Analysis

In the circuit shown in Figure 5, when does the device (system) load connect to the USB power supply or the AC adapter? When does the USB power supply and the AC adapter start charging the battery? At the same time, you need to ensure that the system load can always remain connected to the battery, which is solved by a simple linear regulator (IC2 MAX8881) that can provide a maximum current of 200mA in this example. If the system continues to consume such a large current, and the battery charging current through the USB is only 100mA, the battery will eventually discharge because the load current is greater than the charging current. In many small systems, the peak load current only occurs during part of the entire operation period. Therefore, as long as the average load current is less than the charging current, the battery will still be charged. When the AC adapter is connected, the maximum current of the charger (IC1) rises to 350mA. If the USB and AC adapter are connected at the same time, the AC adapter should be automatically given priority.

One feature that the USB specification requires the charger (IC1) to have (and, generally speaking, it is also beneficial to the charger) is that current is not allowed to flow back from the battery or another power source to the USB power input. In traditional chargers, this can be ensured by input diodes, but the difference between the USB minimum voltage of 4.35V (known from the USB power profile) and the voltage required for charging the Li+ battery (4.2V) is too small to make Schottky diodes suitable. Therefore, all return paths should be blocked inside IC1.

The circuit shown in Figure 5 is subject to some limitations in application and may not be suitable for some rechargeable USB devices. The most obvious limitation is the relatively low charging current. If the capacity of the Li+ battery is greater than a few hundred milliampere hours (mAH), the charging time will be very long. The second limitation is that the load of the basic charger (referring to the input of the linear regulator) is always connected to the battery (that is, the Li+ battery is connected to the IN terminal of the MAX8881 in Figure 5). In this way, if the battery is deeply discharged, the load device may not start working immediately when it is powered on. This is because there is a certain delay before the battery reaches the voltage required for the load device to work.

5. Improved technology: Improvements in charger charging current and peripheral circuits.

In more advanced systems, multiple improvements are required to the charger internal and peripheral circuits. These improvements may include: selectable charging current to match the current capability of the USB power supply or AC adapter or battery; load switching when the USB power supply is connected; and overvoltage protection.

5.1 Improved technical solution (see the circuit shown in Figure 6).

In the circuit shown in Figure 6, the above-mentioned improved functions are realized by using the voltage monitor (charging controller) inside the charger IC1 to drive the external MOSFET Q3 (FDN302).

5.2 Implementation of Improved Technology

MOSFETs Q1 (FDN302) and Q2 (FDN302) and diodes D1 and D2 bypass the battery and directly connect the available power (USB power input or DC power input-AC adapter conversion) to the load. When a power input (USB or DC power input) is valid, its monitoring output goes low and the corresponding MOSFET is turned on. When both inputs are valid, the DC input is used first. IC1 prevents both inputs from being used at the same time. Diodes D1 and D2 are used to block reverse current between the system load power supply path and the input. The internal circuit of the charger (consisting of the charging controller and the two MOSFETs it controls) can block the reverse current of the charging path (BATT).

MOSFET Q2 can also provide AC adapter overvoltage protection, with a protection voltage of up to 18V. The undervoltage/overvoltage monitor (at the DC end) only allows the battery to be charged when the AC adapter voltage is between 4V and 6.25V.

The last MOSFET Q3 is turned on when there is no valid external power supply (i.e. USB power input or DC power input) connected, and the battery is used to power the load device. When either the USB power supply or the DC power supply is connected, the "power on" (PON) output immediately turns off Q3 and disconnects the battery from the load device. In this way, when an external power supply is connected, the system can start working immediately even if the battery is deeply discharged or damaged.

5.3 Perfection and

Practicality Once a USB device is connected to the host, it communicates with the host to determine whether the load current can be increased. If allowed, the load current can increase from one unit load at the beginning to five unit loads. The 5 to 1 current range may be problematic for traditional chargers that are not designed specifically for USB. The problem is that the current accuracy of traditional chargers, although accurate enough at high currents, is affected by the offset of the current sensing circuit at low currents. The result may be that in order to ensure that the charging current does not exceed the 100mA limit at the low end (one unit load), the current must be set at a very low level, which makes it unusable. For example, for a 500mA current with an accuracy of 10%, the current sensing circuit will be very low. , to ensure that 500mA is not exceeded, the output can only be set to 450mA. This alone is acceptable, but to ensure that the charging current does not exceed 100mA at the low end, the average current can only be set to 50mA. The lowest value may be as low as 0mA, which is obviously unacceptable. If USB charging is required to be effective in both ranges, sufficient accuracy is required to provide the maximum charging current possible while not exceeding the USB limitations.

In some designs, due to system power requirements, it is not possible to use less than 500mA of USB budget power to power the load and charge the battery separately. However, using an AC adapter is not a problem.

* The emergence of a cost-effective solution

This requirement can be met with a cost-effective solution, that is, just simplify the circuit of Figure 6: remove Q1, D1 and the connection wires connected to the system load in the circuit of Figure 6, so that the USB power supply is not directly connected to the load, but only to the USB pin of MAX1874; the BATT pin of MAX1874 is connected to the system load through diode D (MBR0520L); charging and system operation still use USB power, but the system remains connected to the battery. The limitation of this design is the same as the circuit shown in Figure 5, that is, if the battery is deeply discharged when the USB is connected, the system will have to go through a certain delay before it can work normally. However, if a DC power supply is connected, the circuit of this solution can work in the same way as the circuit of Figure 6, and there is no need to wait regardless of the battery status. This is because the MOSFET Q (similar to MOSFET Q3 in Figure 6) connected to the "power" on pin (PON) of the MAX1874 is turned off, and the system load is switched from the battery to the DC input through the diode D (similar to D2 in Figure 6).

* Such a simplified, perfect and practical design solution appears, so that the USB power supply is not connected to the load, but the DC input is connected to the load. When the USB power supply is connected, the system is still powered by the battery. At the same time, the battery is charged.

6. NiMH battery charging

Although Li+ batteries can provide the best performance for most portable information terminals, nickel-metal hydride (NiMH) batteries still provide a viable option for the lowest cost design. When the load requirements are not too high, using nickel-metal hydride batteries is a good way to reduce costs. This requires the use of a DC-DC converter to increase the 1.3V battery voltage to a voltage that the device can use, typically 3.3V. Because any battery-powered device requires some type of regulator, and DC-DC is just a different type of regulator, not an additional addition.

The circuit shown in Figure 7 uses an unusual method to charge NiMH batteries. The “charger” is actually a current-limited DC-DC step-down converter (IC1) that charges the battery with a current of 300mA to 400mA. Although not a precise current source, its moderate current control accuracy still meets charging requirements and can maintain current control even if the battery is short-circuited. A great advantage of using DC-DC charging over common linear circuits is the ability to efficiently utilize limited USB power. When charging with a current of 400mA to — When charging a NiMH battery, the circuit draws only 150mA from the USB input. This leaves 350mA available for system use while charging. The load is switched from the battery to the USB by ORing the USB power source with the output of the boost converter (IC2 is a DC-DC boost converter) using a diode (Di). When the USB is disconnected, boost converter IC2 produces a 3.3V output. When the USB is connected, D1 pulls the output of the DC-DC booster (IC2) up to about 4.7V. When the IC2 output is pulled up, it automatically shuts down and draws power from the battery. The current drawn should not exceed 1µA. If the output change from 3.3V to 4.7V when connected to USB is not allowed, a linear regulator can be used in series with D1.

One limitation of this circuit is its reliance on system control to terminate charging. IC1 acts only as a current source and, if not limited, will overcharge the battery. R1 and R2 set the maximum output voltage of IC1 to 2V as a safety margin. The charging enable input can be used by the system to terminate charging of the battery. In addition, because the 150mA charger input current is greater than a unit load, it can also be used as a means to reduce the USB load current before enumeration if necessary.

7. Conclusion

The above introduction to the development and expansion of the power supply function of the USB interface and the introduction of the scheme of obtaining multiple power supplies (3.V and 5V power supplies) from the USB interface is actually to fully utilize the USB power supply function to analyze the charging characteristics of Li+ or nickel-metal hydride (NiMH) batteries of digital portable devices. This design technology can be complex or simple, depending on the power type and power of the USB interface and the combination of the USB device load and its battery performance. In other words, which application scheme can achieve the best cost-effectiveness depends on the actual situation of each portable device.

Reference address：Broadening USB capabilities for reliable power delivery in portable devices

Previous article：Dynamic energy consumption management scheme for embedded systems
Next article：Solar Traffic Light System Design

Popular Resources
Popular amplifiers