“This article discusses some of the power design challenges encountered in wearable IoT devices and proposes a design topology that addresses these challenges using commercially available components. Key design trade-offs are discussed throughout and some suggested mitigations are provided. The ultimate goal of this paper is to propose a robust power supply design topology that provides designers with an efficient solution that can work within the various constraints of wearable IoT devices.
“
By Noah Madinger, Colorado Electronic Product Design (CEPD)
Editor’s note: Designing wearable Internet of Things (IoT) products is often faced with a dilemma that underscores the need for a reliable, stable power system design. Specific considerations for this design include wearable product characteristics such as the product’s compact form factor, reliance on wireless communications, the need for efficient battery power management, and compliance challenges. This article will discuss how power system designs based on buck-boost switching regulators address and meet these design challenges. To this end, we also explore the operating specifications of some commercially available components. At the same time, the power requirements of LTE cellular transceiver modules, the performance specifications of buck-boost switching regulators, and the derating, ESR and capacitance of tantalum capacitors are all covered. Finally, the paper provides a power system topology and a use case to demonstrate the empirical performance of a buck-boost regulator for wearable IoT products.
introduction
The performance of cellular transceivers depends on the reliability and stability of the power rails. Therefore, design choices must be made to ensure adequate power headroom, proper ground plane dimensions, and sufficiently low ripple. These options are compounded when designs are squeezed into wearables, which not only need to be battery powered but also comply with regulatory standards.
This article discusses some of the power design challenges encountered in wearable IoT devices and proposes a design topology that addresses these challenges using commercially available components. Key design trade-offs are discussed throughout and some suggested mitigations are provided. The ultimate goal of this paper is to propose a robust power supply design topology that provides designers with an efficient solution that can work within the various constraints of wearable IoT devices.
Defining the Challenge: Reliability and Stability
In this article, reliability is defined as the ability of a power system to provide a voltage rail within the operating range of a radio transceiver (in this case, a cellular transceiver). This capability must also ensure that the source current meets the typical and peak current consumption expected in IoT products.
Stability is defined as the ability of the ripple present on the voltage rail to be within the specifications of the device. This ripple can be caused by the switching characteristics of the regulator, or it can be caused by a transient response to a sudden jump in current demand. Regardless of the reason, the responsiveness of a regulator is what determines its stability.
Power for Cellular Transceivers
Needless to say, cellular transceiver modules enable wireless connectivity between small and large devices to an unprecedented level of application. These devices are becoming more and more integrated, even incorporating on-board power regulators, temperature compensated oscillators and advanced co-processors. However, all of these devices still rely on key power supply parameters, namely reliability and stability.
The following product samples are intended to emphasize the latter point. Although these products are all commercially available and suitable as the basis for wearable IoT products, there are still power considerations. That is, without the correct power supply, none of these devices will perform at their optimum performance and capabilities.
u-blox
Table 1 lists the power supply parameters of the MPCI-L201-02S-00 cellular module.
Table 1: u-blox power supply parameters.
From this table, u-blox has some pretty strict requirements for powering this module.
The characteristics of a switching regulator connected to the VCC or 3.3 Vaux pins should meet the following prerequisites in order to meet the VCC or 3.3 Vaux requirements for this module:
・Power Capability: The switching regulator and its output circuit must be able to supply the VCC or 3.3 Vaux pin with a voltage value within the specified operating range, and must be able to use the maximum transmission power (TOBY-L2 or MPCI-L2 series data sheet specified) to provide maximum peak/pulse current consumption during burst transfer (Tx).
• Low output ripple: The switching regulator and its output circuit must be able to provide a “clean” (low noise) VCC or 3.3 Vaux voltage profile.
・ The voltage drop cannot exceed 400 mV.
In these requirements, two key aspects are emphasized: reliability and stability. Not only must the power rails be within the proper voltage range, the ripple must also be minimized. Interestingly, “ripple” is divided into two different types in this requirement specification: switching ripple and voltage drop. The first can be thought of as high frequency ripple, which is related to the switching of the regulator. The second is low frequency ripple, which may be caused by the power supply’s inability to respond quickly to high current loads. This can be related to the performance of the regulator; but it can also come from excessive resistance or inductance in the power path.
A voltage regulator used in a cellular dev kit design is likely to be sufficient, but a wall powered dev kit design is not suitable for battery powered wearable applications. Additionally, the shrinking physical footprint of a design, which is a must for wearable products, also affects parasitic resistance and inductance within the power path. Proper selection of regulators alone may not resolve this complication, so additional mitigations are required, especially when these parasitic characteristics threaten product compliance.
Digi
Table 2 lists the power supply parameters of the XBC-V1-UT-001 cellular module.
Table 2: Digi power supply parameters.
From this table, Digi has some fairly strict requirements for the power supply of this module, the specific analysis is as follows:
・ Power supply ripple should be less than 75 mV peak-to-peak.
• The power supply should be capable of delivering at least 1.5 A (5 W) at 3.3 V. Keep in mind that operating at lower voltages requires a power supply capable of delivering higher currents to reach the 5W power requirement.
・ Place enough bulk capacitor on the XBee VCC pin to maintain the voltage above the minimum specification during inrush current. During initial power-up and wake-up from sleep mode for cellular communications, the inrush current is approximately 2 A.
• Place a small high frequency ceramic capacitor close to the XBee cellular modem VCC pin to reduce high frequency noise.
• Use wide power traces or power planes to ensure peak current requirements can be met with minimal voltage drop. Colorado electronic Product Design recommends that the power supply and trace design be such that the voltage on the XBee VCC pin does not vary by more than 0.1 V between light loads (~0.5 W) and heavy loads (~3 W).
Similarly, for other cellular modules, the stability and reliability of the power rails are also key considerations. However, these instructions are more specific, they indicate the maximum ripple voltage, the expected inrush current, and provide some helpful hints on board layout.
Buck-Boost Power Topology C Provides Robust and Reliable Solution for Battery-Powered IoT Wearables
Challenges lie ahead. Design a power system that meets the following requirements:
• Provides power rails within the operating range of the selected module.
• Provide enough current to meet the average and peak current requirements of the module.
・Meet all of the above requirements, but make sure not to exceed the maximum ripple voltage and do not allow too much voltage drop in the power rails.
· Do all of this while being confined to a physical space suitable for wearable applications and trying to comply with regulatory standards relevant to the product’s use case.
As mentioned above, cellular modules have strict requirements on their power systems. All of this can be achieved within a limited physical space; however, higher-level considerations must be employed in order for the product to be successful. The topology in Figure 1 embodies the recommended approach.
Figure 1: Schematic diagram of an advanced buck-boost switching regulator. (Image credit: Colorado Electronic Product Design)
This topology outperforms some common design alternatives, which are also discussed below. Each aspect of this recommended topology and its respective design challenges are presented below, and how to address them.
Cell and battery pack internal resistance
The internal resistance of the battery pack will be higher than the resistance of the battery itself. This is due to the protective circuits, interconnects, fuses and other items used in battery packs for wearable applications. Table 3 lists the disassembled parts of common small lithium polymer battery packs used in mobile phones, and this model is also suitable for wearable IoT devices.
Table 3: Battery Pack Internal Resistance (Itemized). (Image credit: Battery University Group)
1) Connect the cellular module directly to the battery
Under typical current draw, this resistor will not produce a significant voltage drop; however, under peak loads, the battery voltage may drop by 0.13 VC 0.33 V (voltage values depend on the minimum and maximum current drawn by the cellular module shown). While this voltage drop may not drop the power rails below the module’s minimum operating value, it does create voltage drops and ripple that exceed the specifications of these cellular modules. Performance will be affected, therefore, it is not recommended to power the module directly from the battery.
2) Use larger capacitors
Another attempt to overcome this voltage drop is to add more local capacitance. However, this capacitor must supply sufficient current throughout the current draw and do so over the entire operating temperature range of the product. This in itself is a demanding requirement that is difficult to achieve for passive components.
The method becomes more complicated when considering the need for large capacitance. Based on the capacitor’s current formula,
Equation 1
This formula can be used to calculate the required capacitance for a given voltage, current and length of time,
Equation 2
Using a u-blox device as a reference, it can be seen that a high current pulse can remain active for 0.6 ms (4.615 ms / 8).
Figure 2: u-blox current consumption curve. (Image credit: u-blox)
So, how much capacitance is needed to supply 2 A in 0.6 ms to overcome the 0.26 V drop? Using the formula above, the calculated value is 4.62 mF (4.62 X 10-3 Farads). The largest ceramic capacitors are best because they typically have a low equivalent series resistance (ESR), around 680 μF, and are usually not surface mount components. Several of these capacitors must be placed in parallel, and voltage derating, temperature variations, and tolerances must be considered. Bulk tantalum capacitors are also available, but with these tantalum capacitors the ESR limits the amount of current that can be supplied. Again, several of these capacitors must be placed in parallel to account for the poor parasitic characteristics of these components.
In fact, having to use multiple capacitors consumes valuable board space that is already constrained in wearable products and can significantly increase bill of materials costs. Also, every time the battery or any other component in the power path is replaced, the capacitor must be redesigned. These limitations make capacitive solutions problematic when addressing this design consideration.
Buck-Boost Switching Regulators
The buck-boost regulator is at the heart of this power supply design topology. This section will introduce two commercially available buck-boost regulators. Both are suitable options for wearable IoT applications. However, before diving into these details, the following points will help explain the need for using this type of regulator.
1) Buck regulator alone is not enough
At this point, we discussed earlier that connecting the cellular module directly to the battery is not a good design choice. However, this section further shows that while using a buck regulator will be an improvement over connecting a battery directly, it is still not a design choice for most wearable IoT use cases. A boost is required, and why is explained below.
Figure 3: Discharge curves of lithium batteries (nominal 3.7 V) at discharge currents of 0.2 C, 0.5 C and 1 C. (Image credit: Innovative Battery Technology)
When the battery has 20% charge remaining, the battery voltage may be in the range of 2.8 V to 3.7 V. At this time, the under-voltage protection circuit may disconnect the battery when the voltage is lower than 3.0V. Based on this, assume that the “effective” voltage range is 3.7 VC 3.0 V for a battery with 20% capacity remaining. Combine this information with the fact that a buck regulator requires the input voltage to be greater than or equal to the output voltage, and the design challenges begin to emerge.
If VOUT is set to 3.3 V, and a buck regulator is used, the lowest usable battery voltage will be what the battery can sustain while the cellular module is pulling its peak current, as long as it is 3.3 V or greater.
Mathematically, efficiency is calculated as:
Equation 3
Rearrange the formula:
Equation 4
Assuming that the buck regulator is 90% efficient, if the design uses the u-blox module, the buck regulator must deliver 3.3 V * 2.5 A = 8.25 W. This means that the input power must be 8.25 W/0.9 = 9.2 W.
Apply formula
Equation 5
As can be seen, the battery must supply 2.49 A at its battery nominal value of 3.7 V. However, this is the current supplied to the regulator, which must first pass through the series resistance of the battery pack. Therefore, the actual battery voltage must be the sum of the voltage at the input of the regulator and the voltage drop across the series resistor of the battery pack: 3.7 V + (2.49 A * 0.13 ohms) = 4.02 V. Therefore, a voltage drop of 0.32 V is obtained across the series resistance of the battery pack.
This means that the lowest usable value for this battery should be 3.3 V + VSeries_Resistance = ~3.62 V. If the voltage of the battery pack falls below this value, the buck regulator’s input voltage will no longer be greater than or equal to the output voltage, and therefore, regulation will fail. This failure to regulate will cause the power rails of the cellular module to sag and will also violate the ripple voltage and sag requirements. Performance will be affected.
2) Other Considerations
In short, the boost portion of the buck-boost regulator allows the system to use the last remaining 20% of the battery pack’s capacity. With the buck-boost feature, as long as the battery can maintain power to the regulator, the module’s power rails are supported without premature shutdown while the battery still has charge remaining.
It’s worth noting that when using a buck-boost regulator, the last 20% of the battery will be drained faster than the previous 80%. This is because once the input voltage falls below the output voltage set point, the required input current increases. However, this current increase should be taken into account when choosing the maximum discharge current of the battery pack.
3) Product example C Renesas ISL91110
The diagram below shows the functionality of this part. The part features automatic switching from light to heavy duty operation. This effectively improves the efficiency over the full operating range of the output current.
Figure 4: Renesas ISL91110 Efficiency vs. VIN. (Image credit: Renesas)
Figure 5: Renesas ISL91110 0 A to 2 A load transient (VIN = 3.6 V, VOUT = 3.3 V). (Image credit: Renesas)
4) Product example C ON semiconductor FAN49103
The part also features automatic switching from light to heavy duty operation. Although this parameter is suitable for an output voltage set to 3.4 V (instead of 3.3 V), the part can still be used for this example application.
Figure 6: ON Semiconductor FAN49103 Efficiency vs. I Load (mA). (Image credit: ON Semiconductor)
Figure 7: ON Semiconductor FAN49103 0 A to 2 A load transient (VIN = 3.6 V, VOUT = 3.4 V). (Image credit: ON Semiconductor)
local capacitor
Local capacitors perform two important functions: provide local energy storage to meet sudden increases in load current, and filter out high frequency transients and ripple voltages that can be detrimental to performance.
The suggested capacitor placement in this design layout is critical. Capacitors should be placed in the recommended way to ensure the “cleanest” voltage rail is used to power the cellular module. This means that capacitors next to the cellular module must have the lowest ESR and ESL. In fact, their actual capacitance ratings can be in the picofarad range. C0G ceramic capacitors are recommended.
Now, while these small-capacity capacitors do high-frequency filtering well, they have almost no energy storage. To achieve this, place a larger tantalum capacitor in the hundreds of microfarads range farthest from the power pins of the cellular module. That doesn’t mean it’s far away; it’s just that it’s not placed as close as the aforementioned ceramic capacitors. Another important feature of such large capacitors is their low ESR at the fundamental frequency where current transients are expected. The recommended ESR value is 100 mΩ @ 100 KHz.
Figure 8 illustrates the recommended layout for the MPCI u-blox cellular module.
Figure 8: Recommended local capacitor placement scheme for the u-blox MPCI-L2 family. (Image credit: u-blox)
In Figure 8, C1 to C3 are low capacitance, low ESR, low ESL C0G capacitors. C4 C C5 are ceramic capacitors in the 0.1 C 10 μF range. Finally, C6 is a bulk tantalum capacitor with low ESR at the fundamental frequency of the transient load current.
It is extremely important to choose a rated voltage to mitigate derating. This is especially true for ceramic capacitors.
This section concludes with a description of several commercially available capacitors. Applicable parameters are provided.
1) KEMET
Part Number: T520D337M006ATE045
Capacitance: 330 μF
Tolerance: 20%
Rated voltage: 6.3 V
ESR @ 100 KHz: 45 mΩ
2) Panasonic Electronic components
Part number: 6TPF470MAH
Capacitance: 470 μF
Tolerance: 20%
Rated voltage: 6.3 V
ESR @ 100KHz: 10mΩ
Layout Design Considerations
While each selected component’s datasheet lists its specific layout recommendations, there are some general layout guidelines to achieve high efficiency and low noise performance.
1) Ground and power pours
Use polygon pours whenever possible. This is especially true for connections to the input voltage, output voltage, Inductor, and ground nodes. In short, don’t let copper layers sit idle, as these copper layers provide a low resistance and low inductance path for current flow, including any stray or switching currents. Figure 9 is a suggested top-level layout for Linear Technology’s LTC3113 buck-boost regulator, which nicely illustrates the preference for copper pouring.
Figure 9: Recommended top-level layout for the Linear Tech LTC3113. (Image credit: Linear Technology)
2) Absorption circuit
While every effort has been made to reduce parasitic resistance and inductance, this is a size-constrained wearable design. The ground and power planes are not as big as they should be. The configuration in this layout should allow placement of RC sink circuits. While these components initially do not require padding, leaving a padding area gives the designer flexibility in case this circuit is needed to reduce emissions.
These parasitic elements cause transient oscillations in the switching current (Figure 10).
Figure 10: Transient oscillations in the switching inductor current of a buck regulator. (Image credit: ROHM Semiconductor)
Now, as mentioned before, this may be an unavoidable problem because of space requirements. The snubber circuit shown in Figure 11 draws this stray energy to ground. If this is not done, these oscillations may push the designed emissions beyond acceptable limits for compliance. Snubber circuits are useful noise suppression tools for space-constrained regulators.
Figure 11: Recommended RC sink circuit location for a buck regulator. (Image credit: ROHM Semiconductor)
3) Ferrite beads
The last suggestion is to address any persistent high frequency noise that comes with the output power. Choose a high-current ferrite bead with proper attenuation at critical frequencies in series with the output of the buck-boost regulator. and should be placed between the output of the regulator and the bulk bypass capacitor.
Case Study C LTC3113 Powering u-blox SARA Module
The SARA module is a 3G cellular transceiver. Just like the aforementioned cellular module, it also draws high current in inrush due to the series resistance, causing the battery voltage to sag. The LTC3113 buck-boost switching regulator circuit in Figure 12 is designed to maintain a stable and reliable 3.3 V rail for this design.
Figure 12: LTC3113 buck-boost switching regulator circuit case study. (Image credit: Colorado Electronic Product Design)
This regulator design, combined with local bypass capacitors arranged as shown in Figure 12, produces a stable power rail at all operating currents drawn. The oscilloscope plot of Figure 13 captures the current drawn by the SARA (blue), the 3.3 V output rail from the buck-boost regulator (green), the input battery voltage, and any sag on that rail (purple) and the ripple voltage (orange) measured on the output power rail.
As can be seen, this high current spike does not cause sag or significant ripple on the regulated 3.3 V output rail. However, this does cause the input rail to sag.
Figure 13: LTC3113 buck-boost switching regulator circuit case study showing ~0.9 A module current consumed by SARA module (blue), 3.3 V output rail (green), battery input rail (purple) and 3.3 V output rail ripple (orange). (Image credit: Colorado Electronic Product Design)
Again, the stability and reliability of the output rail is consistent at solid-state 3.3 V with minimal ripple. However, the voltage sag on the battery input rail is about 0.32 V, which is beyond the specifications of the SARA module and the other modules mentioned in this article. Buck-boost regulators are able to accommodate these current spikes and maintain power rails suitable for the cellular module to operate under all its intended conditions.
Epilogue
Wearable IoT design presents a set of challenges for design engineers, and power systems are at the heart of many. Buck-boost regulator topologies directly address these challenges by providing stable and reliable power rails over the operating conditions of cellular modules. This is not to say that careful design work is not required. Rather, the topology will work if good design practices are followed. As wearable IoT designs become more compact, performance expectations have increased accordingly. Consider this robust topology for powering compact, high-performance wearable IoT designs.
Acknowledgements: Special thanks to the management and staff of Linear Tech/Analog Devices and CEPD (Colorado Electronic Product Design).
“This article discusses some of the power design challenges encountered in wearable IoT devices and proposes a design topology that addresses these challenges using commercially available components. Key design trade-offs are discussed throughout and some suggested mitigations are provided. The ultimate goal of this paper is to propose a robust power supply design topology that provides designers with an efficient solution that can work within the various constraints of wearable IoT devices.
“
By Noah Madinger, Colorado Electronic Product Design (CEPD)
Editor’s note: Designing wearable Internet of Things (IoT) products is often faced with a dilemma that underscores the need for a reliable, stable power system design. Specific considerations for this design include wearable product characteristics such as the product’s compact form factor, reliance on wireless communications, the need for efficient battery power management, and compliance challenges. This article will discuss how power system designs based on buck-boost switching regulators address and meet these design challenges. To this end, we also explore the operating specifications of some commercially available components. At the same time, the power requirements of LTE cellular transceiver modules, the performance specifications of buck-boost switching regulators, and the derating, ESR and capacitance of tantalum capacitors are all covered. Finally, the paper provides a power system topology and a use case to demonstrate the empirical performance of a buck-boost regulator for wearable IoT products.
introduction
The performance of cellular transceivers depends on the reliability and stability of the power rails. Therefore, design choices must be made to ensure adequate power headroom, proper ground plane dimensions, and sufficiently low ripple. These options are compounded when designs are squeezed into wearables, which not only need to be battery powered but also comply with regulatory standards.
This article discusses some of the power design challenges encountered in wearable IoT devices and proposes a design topology that addresses these challenges using commercially available components. Key design trade-offs are discussed throughout and some suggested mitigations are provided. The ultimate goal of this paper is to propose a robust power supply design topology that provides designers with an efficient solution that can work within the various constraints of wearable IoT devices.
Defining the Challenge: Reliability and Stability
In this article, reliability is defined as the ability of a power system to provide a voltage rail within the operating range of a radio transceiver (in this case, a cellular transceiver). This capability must also ensure that the source current meets the typical and peak current consumption expected in IoT products.
Stability is defined as the ability of the ripple present on the voltage rail to be within the specifications of the device. This ripple can be caused by the switching characteristics of the regulator, or it can be caused by a transient response to a sudden jump in current demand. Regardless of the reason, the responsiveness of a regulator is what determines its stability.
Power for Cellular Transceivers
Needless to say, cellular transceiver modules enable wireless connectivity between small and large devices to an unprecedented level of application. These devices are becoming more and more integrated, even incorporating on-board power regulators, temperature compensated oscillators and advanced co-processors. However, all of these devices still rely on key power supply parameters, namely reliability and stability.
The following product samples are intended to emphasize the latter point. Although these products are all commercially available and suitable as the basis for wearable IoT products, there are still power considerations. That is, without the correct power supply, none of these devices will perform at their optimum performance and capabilities.
u-blox
Table 1 lists the power supply parameters of the MPCI-L201-02S-00 cellular module.
Table 1: u-blox power supply parameters.
From this table, u-blox has some pretty strict requirements for powering this module.
The characteristics of a switching regulator connected to the VCC or 3.3 Vaux pins should meet the following prerequisites in order to meet the VCC or 3.3 Vaux requirements for this module:
・Power Capability: The switching regulator and its output circuit must be able to supply the VCC or 3.3 Vaux pin with a voltage value within the specified operating range, and must be able to use the maximum transmission power (TOBY-L2 or MPCI-L2 series data sheet specified) to provide maximum peak/pulse current consumption during burst transfer (Tx).
• Low output ripple: The switching regulator and its output circuit must be able to provide a “clean” (low noise) VCC or 3.3 Vaux voltage profile.
・ The voltage drop cannot exceed 400 mV.
In these requirements, two key aspects are emphasized: reliability and stability. Not only must the power rails be within the proper voltage range, the ripple must also be minimized. Interestingly, “ripple” is divided into two different types in this requirement specification: switching ripple and voltage drop. The first can be thought of as high frequency ripple, which is related to the switching of the regulator. The second is low frequency ripple, which may be caused by the power supply’s inability to respond quickly to high current loads. This can be related to the performance of the regulator; but it can also come from excessive resistance or inductance in the power path.
A voltage regulator used in a cellular dev kit design is likely to be sufficient, but a wall powered dev kit design is not suitable for battery powered wearable applications. Additionally, the shrinking physical footprint of a design, which is a must for wearable products, also affects parasitic resistance and inductance within the power path. Proper selection of regulators alone may not resolve this complication, so additional mitigations are required, especially when these parasitic characteristics threaten product compliance.
Digi
Table 2 lists the power supply parameters of the XBC-V1-UT-001 cellular module.
Table 2: Digi power supply parameters.
From this table, Digi has some fairly strict requirements for the power supply of this module, the specific analysis is as follows:
・ Power supply ripple should be less than 75 mV peak-to-peak.
• The power supply should be capable of delivering at least 1.5 A (5 W) at 3.3 V. Keep in mind that operating at lower voltages requires a power supply capable of delivering higher currents to reach the 5W power requirement.
・ Place enough bulk capacitor on the XBee VCC pin to maintain the voltage above the minimum specification during inrush current. During initial power-up and wake-up from sleep mode for cellular communications, the inrush current is approximately 2 A.
• Place a small high frequency ceramic capacitor close to the XBee cellular modem VCC pin to reduce high frequency noise.
• Use wide power traces or power planes to ensure peak current requirements can be met with minimal voltage drop. Colorado Electronic Product Design recommends that the power supply and trace design be such that the voltage on the XBee VCC pin does not vary by more than 0.1 V between light loads (~0.5 W) and heavy loads (~3 W).
Similarly, for other cellular modules, the stability and reliability of the power rails are also key considerations. However, these instructions are more specific, they indicate the maximum ripple voltage, the expected inrush current, and provide some helpful hints on board layout.
Buck-Boost Power Topology C Provides Robust and Reliable Solution for Battery-Powered IoT Wearables
Challenges lie ahead. Design a power system that meets the following requirements:
• Provides power rails within the operating range of the selected module.
• Provide enough current to meet the average and peak current requirements of the module.
・Meet all of the above requirements, but make sure not to exceed the maximum ripple voltage and do not allow too much voltage drop in the power rails.
· Do all of this while being confined to a physical space suitable for wearable applications and trying to comply with regulatory standards relevant to the product’s use case.
As mentioned above, cellular modules have strict requirements on their power systems. All of this can be achieved within a limited physical space; however, higher-level considerations must be employed in order for the product to be successful. The topology in Figure 1 embodies the recommended approach.
Figure 1: Schematic diagram of an advanced buck-boost switching regulator. (Image credit: Colorado Electronic Product Design)
This topology outperforms some common design alternatives, which are also discussed below. Each aspect of this recommended topology and its respective design challenges are presented below, and how to address them.
Cell and battery pack internal resistance
The internal resistance of the battery pack will be higher than the resistance of the battery itself. This is due to the protective circuits, interconnects, fuses and other items used in battery packs for wearable applications. Table 3 lists the disassembled parts of common small lithium polymer battery packs used in mobile phones, and this model is also suitable for wearable IoT devices.
Table 3: Battery Pack Internal Resistance (Itemized). (Image credit: Battery University Group)
1) Connect the cellular module directly to the battery
Under typical current draw, this resistor will not produce a significant voltage drop; however, under peak loads, the battery voltage may drop by 0.13 VC 0.33 V (voltage values depend on the minimum and maximum current drawn by the cellular module shown). While this voltage drop may not drop the power rails below the module’s minimum operating value, it does create voltage drops and ripple that exceed the specifications of these cellular modules. Performance will be affected, therefore, it is not recommended to power the module directly from the battery.
2) Use larger capacitors
Another attempt to overcome this voltage drop is to add more local capacitance. However, this capacitor must supply sufficient current throughout the current draw and do so over the entire operating temperature range of the product. This in itself is a demanding requirement that is difficult to achieve for passive components.
The method becomes more complicated when considering the need for large capacitance. Based on the capacitor’s current formula,
Equation 1
This formula can be used to calculate the required capacitance for a given voltage, current and length of time,
Equation 2
Using a u-blox device as a reference, it can be seen that a high current pulse can remain active for 0.6 ms (4.615 ms / 8).
Figure 2: u-blox current consumption curve. (Image credit: u-blox)
So, how much capacitance is needed to supply 2 A in 0.6 ms to overcome the 0.26 V drop? Using the formula above, the calculated value is 4.62 mF (4.62 X 10-3 Farads). The largest ceramic capacitors are best because they typically have a low equivalent series resistance (ESR), around 680 μF, and are usually not surface mount components. Several of these capacitors must be placed in parallel, and voltage derating, temperature variations, and tolerances must be considered. Bulk tantalum capacitors are also available, but with these tantalum capacitors the ESR limits the amount of current that can be supplied. Again, several of these capacitors must be placed in parallel to account for the poor parasitic characteristics of these components.
In fact, having to use multiple capacitors consumes valuable board space that is already constrained in wearable products and can significantly increase bill of materials costs. Also, every time the battery or any other component in the power path is replaced, the capacitor must be redesigned. These limitations make capacitive solutions problematic when addressing this design consideration.
Buck-Boost Switching Regulators
The buck-boost regulator is at the heart of this power supply design topology. This section will introduce two commercially available buck-boost regulators. Both are suitable options for wearable IoT applications. However, before diving into these details, the following points will help explain the need for using this type of regulator.
1) Buck regulator alone is not enough
At this point, we discussed earlier that connecting the cellular module directly to the battery is not a good design choice. However, this section further shows that while using a buck regulator will be an improvement over connecting a battery directly, it is still not a design choice for most wearable IoT use cases. A boost is required, and why is explained below.
Figure 3: Discharge curves of lithium batteries (nominal 3.7 V) at discharge currents of 0.2 C, 0.5 C and 1 C. (Image credit: Innovative Battery Technology)
When the battery has 20% charge remaining, the battery voltage may be in the range of 2.8 V to 3.7 V. At this time, the under-voltage protection circuit may disconnect the battery when the voltage is lower than 3.0V. Based on this, assume that the “effective” voltage range is 3.7 VC 3.0 V for a battery with 20% capacity remaining. Combine this information with the fact that a buck regulator requires the input voltage to be greater than or equal to the output voltage, and the design challenges begin to emerge.
If VOUT is set to 3.3 V, and a buck regulator is used, the lowest usable battery voltage will be what the battery can sustain while the cellular module is pulling its peak current, as long as it is 3.3 V or greater.
Mathematically, efficiency is calculated as:
Equation 3
Rearrange the formula:
Equation 4
Assuming that the buck regulator is 90% efficient, if the design uses the u-blox module, the buck regulator must deliver 3.3 V * 2.5 A = 8.25 W. This means that the input power must be 8.25 W/0.9 = 9.2 W.
Apply formula
Equation 5
As can be seen, the battery must supply 2.49 A at its battery nominal value of 3.7 V. However, this is the current supplied to the regulator, which must first pass through the series resistance of the battery pack. Therefore, the actual battery voltage must be the sum of the voltage at the input of the regulator and the voltage drop across the series resistor of the battery pack: 3.7 V + (2.49 A * 0.13 ohms) = 4.02 V. Therefore, a voltage drop of 0.32 V is obtained across the series resistance of the battery pack.
This means that the lowest usable value for this battery should be 3.3 V + VSeries_Resistance = ~3.62 V. If the voltage of the battery pack falls below this value, the buck regulator’s input voltage will no longer be greater than or equal to the output voltage, and therefore, regulation will fail. This failure to regulate will cause the power rails of the cellular module to sag and will also violate the ripple voltage and sag requirements. Performance will be affected.
2) Other Considerations
In short, the boost portion of the buck-boost regulator allows the system to use the last remaining 20% of the battery pack’s capacity. With the buck-boost feature, as long as the battery can maintain power to the regulator, the module’s power rails are supported without premature shutdown while the battery still has charge remaining.
It’s worth noting that when using a buck-boost regulator, the last 20% of the battery will be drained faster than the previous 80%. This is because once the input voltage falls below the output voltage set point, the required input current increases. However, this current increase should be taken into account when choosing the maximum discharge current of the battery pack.
3) Product example C Renesas ISL91110
The diagram below shows the functionality of this part. The part features automatic switching from light to heavy duty operation. This effectively improves the efficiency over the full operating range of the output current.
Figure 4: Renesas ISL91110 Efficiency vs. VIN. (Image credit: Renesas)
Figure 5: Renesas ISL91110 0 A to 2 A load transient (VIN = 3.6 V, VOUT = 3.3 V). (Image credit: Renesas)
4) Product example C ON semiconductor FAN49103
The part also features automatic switching from light to heavy duty operation. Although this parameter is suitable for an output voltage set to 3.4 V (instead of 3.3 V), the part can still be used for this example application.
Figure 6: ON Semiconductor FAN49103 Efficiency vs. I Load (mA). (Image credit: ON Semiconductor)
Figure 7: ON Semiconductor FAN49103 0 A to 2 A load transient (VIN = 3.6 V, VOUT = 3.4 V). (Image credit: ON Semiconductor)
local capacitor
Local capacitors perform two important functions: provide local energy storage to meet sudden increases in load current, and filter out high frequency transients and ripple voltages that can be detrimental to performance.
The suggested capacitor placement in this design layout is critical. Capacitors should be placed in the recommended way to ensure the “cleanest” voltage rail is used to power the cellular module. This means that capacitors next to the cellular module must have the lowest ESR and ESL. In fact, their actual capacitance ratings can be in the picofarad range. C0G ceramic capacitors are recommended.
Now, while these small-capacity capacitors do high-frequency filtering well, they have almost no energy storage. To achieve this, place a larger tantalum capacitor in the hundreds of microfarads range farthest from the power pins of the cellular module. That doesn’t mean it’s far away; it’s just that it’s not placed as close as the aforementioned ceramic capacitors. Another important feature of such large capacitors is their low ESR at the fundamental frequency where current transients are expected. The recommended ESR value is 100 mΩ @ 100 KHz.
Figure 8 illustrates the recommended layout for the MPCI u-blox cellular module.
Figure 8: Recommended local capacitor placement scheme for the u-blox MPCI-L2 family. (Image credit: u-blox)
In Figure 8, C1 to C3 are low capacitance, low ESR, low ESL C0G capacitors. C4 C C5 are ceramic capacitors in the 0.1 C 10 μF range. Finally, C6 is a bulk tantalum capacitor with low ESR at the fundamental frequency of the transient load current.
It is extremely important to choose a rated voltage to mitigate derating. This is especially true for ceramic capacitors.
This section concludes with a description of several commercially available capacitors. Applicable parameters are provided.
1) KEMET
Part Number: T520D337M006ATE045
Capacitance: 330 μF
Tolerance: 20%
Rated voltage: 6.3 V
ESR @ 100 KHz: 45 mΩ
2) Panasonic Electronic components
Part number: 6TPF470MAH
Capacitance: 470 μF
Tolerance: 20%
Rated voltage: 6.3 V
ESR @ 100KHz: 10mΩ
Layout Design Considerations
While each selected component’s datasheet lists its specific layout recommendations, there are some general layout guidelines to achieve high efficiency and low noise performance.
1) Ground and power pours
Use polygon pours whenever possible. This is especially true for connections to the input voltage, output voltage, Inductor, and ground nodes. In short, don’t let copper layers sit idle, as these copper layers provide a low resistance and low inductance path for current flow, including any stray or switching currents. Figure 9 is a suggested top-level layout for Linear Technology’s LTC3113 buck-boost regulator, which nicely illustrates the preference for copper pouring.
Figure 9: Recommended top-level layout for the Linear Tech LTC3113. (Image credit: Linear Technology)
2) Absorption circuit
While every effort has been made to reduce parasitic resistance and inductance, this is a size-constrained wearable design. The ground and power planes are not as big as they should be. The configuration in this layout should allow placement of RC sink circuits. While these components initially do not require padding, leaving a padding area gives the designer flexibility in case this circuit is needed to reduce emissions.
These parasitic elements cause transient oscillations in the switching current (Figure 10).
Figure 10: Transient oscillations in the switching inductor current of a buck regulator. (Image credit: ROHM Semiconductor)
Now, as mentioned before, this may be an unavoidable problem because of space requirements. The snubber circuit shown in Figure 11 draws this stray energy to ground. If this is not done, these oscillations may push the designed emissions beyond acceptable limits for compliance. Snubber circuits are useful noise suppression tools for space-constrained regulators.
Figure 11: Recommended RC sink circuit location for a buck regulator. (Image credit: ROHM Semiconductor)
3) Ferrite beads
The last suggestion is to address any persistent high frequency noise that comes with the output power. Choose a high-current ferrite bead with proper attenuation at critical frequencies in series with the output of the buck-boost regulator. and should be placed between the output of the regulator and the bulk bypass capacitor.
Case Study C LTC3113 Powering u-blox SARA Module
The SARA module is a 3G cellular transceiver. Just like the aforementioned cellular module, it also draws high current in inrush due to the series resistance, causing the battery voltage to sag. The LTC3113 buck-boost switching regulator circuit in Figure 12 is designed to maintain a stable and reliable 3.3 V rail for this design.
Figure 12: LTC3113 buck-boost switching regulator circuit case study. (Image credit: Colorado Electronic Product Design)
This regulator design, combined with local bypass capacitors arranged as shown in Figure 12, produces a stable power rail at all operating currents drawn. The oscilloscope plot of Figure 13 captures the current drawn by the SARA (blue), the 3.3 V output rail from the buck-boost regulator (green), the input battery voltage, and any sag on that rail (purple) and the ripple voltage (orange) measured on the output power rail.
As can be seen, this high current spike does not cause sag or significant ripple on the regulated 3.3 V output rail. However, this does cause the input rail to sag.
Figure 13: LTC3113 buck-boost switching regulator circuit case study showing ~0.9 A module current consumed by SARA module (blue), 3.3 V output rail (green), battery input rail (purple) and 3.3 V output rail ripple (orange). (Image credit: Colorado Electronic Product Design)
Again, the stability and reliability of the output rail is consistent at solid-state 3.3 V with minimal ripple. However, the voltage sag on the battery input rail is about 0.32 V, which is beyond the specifications of the SARA module and the other modules mentioned in this article. Buck-boost regulators are able to accommodate these current spikes and maintain power rails suitable for the cellular module to operate under all its intended conditions.
Epilogue
Wearable IoT design presents a set of challenges for design engineers, and power systems are at the heart of many. Buck-boost regulator topologies directly address these challenges by providing stable and reliable power rails over the operating conditions of cellular modules. This is not to say that careful design work is not required. Rather, the topology will work if good design practices are followed. As wearable IoT designs become more compact, performance expectations have increased accordingly. Consider this robust topology for powering compact, high-performance wearable IoT designs.
Acknowledgements: Special thanks to the management and staff of Linear Tech/Analog Devices and CEPD (Colorado Electronic Product Design).
The Links: LQ104V7DS01 G190EG02VG