Texas Instruments | Reducing Power Loss and Overheating During Faults with eFuses | Application notes | Texas Instruments Reducing Power Loss and Overheating During Faults with eFuses Application notes

Texas Instruments Reducing Power Loss and Overheating During Faults with eFuses Application notes
Application Report
SLVAED9 – June 2019
Reducing Power Loss and Overheating During Faults with
eFuses
Rakesh Panguloori
1
Introduction
Several components inside electronic equipment have an inherent maximum current capacity, which when
exceeded, overheats (possibly to an unrecoverable state). Some of the common components are:
• Semiconductor devices
• Passive filters
• Printed circuit board (PCB) traces
• Connectors
• Interfacing cables
This makes current limiting an essential requirement to protect components in almost all systems. Current
limiting maintains transient overload currents within the maximum current rating limits. Protecting the
power path against faults is crucial but after the fault is cleared, it is also important for the system to
recover quickly without manual intervention. This minimizes maintenance costs by reducing visits of
technician on-site. All of these drive the fault response of the protection devices as one of the key system
level requirements.
• Fuses, also known as a mechanical fuse or melting fuse, are traditionally used as protection devices to
isolate overload or short-circuit faults from the main system. While fuses are an inexpensive solution,
they need to be physically replaced every time they melt, because they become open and nonoperational when a fault occurs. This increases system downtime and maintenance costs.
• Positive temperature coefficient (PTC) resistors eliminate the need for human intervention by providing
resettable overcurrent protection. The PTC resistor auto-resets the system from a circuit trip as the
temperature cools. However, the ON-resistance of PTC increases after every fault, which raises
concerns about achieving repeatable performance over time.
• An electronic fuse or eFuse acts as a “self-healing” device. When it “breaks the path”, it automatically
turns back on, and attempts to restart the circuit and provides the same robust performance after each
overcurrent event.
Figure 1 shows fault response summary of fuse, PTC, and eFuse devices.
Fault Response
Open and
Non-operational
Self-resets but performance
degrades after each reset
Repeatable performance
after each reset
Manual
Intervention
Must
Moderate
Very Low
Figure 1. Summary of Fault Response for Fuse, Resettable Fuse, and eFuse
SLVAED9 – June 2019
Submit Documentation Feedback
Reducing Power Loss and Overheating During Faults with eFuses
Copyright © 2019, Texas Instruments Incorporated
1
Introduction
www.ti.com
An eFuse is an active circuit protection device with an integrated FET used to limit currents, and keep
voltages at safe levels during fault conditions. For an overload event, the eFuse does the current limit and
serves the load until it reaches thermal shutdown. For additional information on how the eFuse protects
different types of faults, refer to the Basics of eFuses Application Report. An eFuse provides the flexibility
to select fault response depending on the type of eFuse. A latch-off version remains in the OFF position
and needs power recycle to turn ON again. Figure 2 shows the overload response of the latched
TPS259540 device, where the device provides a limited current of 4 A to the output before going into
thermal shutdown. A reset at the enable pin is needed to bring the device back into operation. This is
different to the latch-off functionality, where an eFuse with the auto-retry function continues to power cycle
itself indefinitely until the fault is cleared. As shown in Figure 3, the TPS259541 (auto-retry variant)
devices attempts to power cycle after a retry delay of 93 ms and resumes normal operation as soon as the
overload fault is cleared.
Output falls to zero due
to thermal shutdown
Device reset at EN pin
eFuse limits current to
4A for an overload event
Figure 2. Latch-off Version of an eFuse (TPS259540) Waits for Power Cycle or EN Cycle
Device auto-retries for
every 93ms
Output recovers after
overload fault is cleared
eFuse limits current to
4A for an overload event
Figure 3. Auto-retry Version of an eFuse (TPS259541) Continues to Power the Cycle Until the Fault is
Cleared
2
Reducing Power Loss and Overheating During Faults with eFuses
Copyright © 2019, Texas Instruments Incorporated
SLVAED9 – June 2019
Submit Documentation Feedback
Introduction
www.ti.com
The auto-retry function is generally preferred in many of the applications to avoid permanent system
shutdown for any temporary faults or transients. This allows systems in remote or inaccessible
installations to try and recover from faults without need for technician intervention. Another advantage with
the auto-retry function is its ability to support unknown loads in a hiccup manner to make a successful
start-up. For example, a main board can have several output slots with daughter cards of unknown
capacitances. An improper start-up design pushes the device into current limit mode and then into thermal
shutdown. However, the auto-retry function can charge the unknown load capacitance in multiple attempts
to successfully start up the system, as shown in Figure 4.
Output builds up for
every retry event
Starts in current limit (4A) due to
large unknown capacitive load
RILM = 487 Ω
Device auto-retries for
every 93ms
CdVdt = 3.3 nF
COUT = 2.2 mF
Figure 4. The TPS259541 Device Starts an Unknown Load in Current Limit Mode
SLVAED9 – June 2019
Submit Documentation Feedback
Reducing Power Loss and Overheating During Faults with eFuses
Copyright © 2019, Texas Instruments Incorporated
3
Challenges with Indefinite Auto-Retries
2
www.ti.com
Challenges with Indefinite Auto-Retries
The auto-retry function helps to improve the system uptime without the need of manual intervention, but
there is a concern in a scenario where the fault is real and persistent. This scenario brings several
system-level challenges as discussed in the following sections.
2.1
Higher Operating Temperature
As shown in Figure 3, under a persistent fault, the junction temperature of an auto-retry eFuse hits thermal
shutdown threshold (for example, it is 157°C for the TPS259541) in every retry cycle and shuts OFF.
Then, the device waits to cool down for thermal hysteresis (typically 5°C to 10°C) and a retry delay interval
before attempting next power cycle. Though the auto-retry delay helps to cool down the device, it may not
be sufficient to bring the operating temperature to a normal level. Figure 5 shows a thermal image of the
TPS2595EVM board during an overload event. The average case temperature of the device reaches
~121°C, and as a result, the average junction temperature of the device reaches close to the maximum
recommended value, thus raising concern on the long-term reliability under a persistent fault. This also
increases the operating temperature of the PCB and nearby components. Any adjacent power-dissipating
components could enhance the problem due to a mutual heating effect and can potentially lead to
overheating. Therefore, a limited number of auto-retries or longer retry delay would be needed to avoid
this situation.
Figure 5. Case Temperature of the TPS259541 eFuse on the TPS2595EVM Under a Sustained Fault
Condition
2.2
Power Loss
Under sustained fault condition, even though the fault current is limited, it simply gets shunted to ground
without being used for any specific work. For example, in case of overload fault with 2 A current limit
setting, the device limits the output current to 2 A. The peak power drawn from the input source is 12 V ×
2 A = 24 W (refer to Channel-M1 in Figure 6), which causes an average power loss of 24 W × 5.25 ms /
94.54 ms = 1.33 W. Latching-off the device after a limited number of retries or a longer retry delay time
minimizes the unnecessary power loss.
4
Reducing Power Loss and Overheating During Faults with eFuses
Copyright © 2019, Texas Instruments Incorporated
SLVAED9 – June 2019
Submit Documentation Feedback
Challenges with Indefinite Auto-Retries
www.ti.com
eFuse limits current to 2A
for an overload event
Power drawn from the
source (24W Peak)
Power dissipated across
the device (20W Peak)
RILM = 1020 Ω
ROUT = 1 Ω
CH-M1 = VIN × I-IN
CH-M2 = (VIN – VOUT)
× I-IN
Figure 6. Power Dissipation Profile of the TPS259541 Under Sustained Overload Fault Condition
SLVAED9 – June 2019
Submit Documentation Feedback
Reducing Power Loss and Overheating During Faults with eFuses
Copyright © 2019, Texas Instruments Incorporated
5
Techniques to Configure Fault Response
3
www.ti.com
Techniques to Configure Fault Response
In compact, enclosed systems, where excessive power dissipation and higher operating temperatures are
not tolerable, a limited number of auto-retries or longer retry delay would be needed to ensure proper
system functionality. There are several techniques to configure the fault response (that is the number of
retries and retry delay), which includes the following:
• Discrete components around eFuse
• Using a system microcontroller
• Sophisticated devices that have a digital interface
This section discusses the implementation challenges of these approaches and explains how the
integrated auto-retry scheme of the TPS25982 smart eFuse simplifies the design.
3.1
Using RC Network
In this implementation (Figure 7), the FLTb and EN pins of the eFuse are connected together and a RC
network is used. For a fault, the eFuse asserts the FLTb pin which in turn pulls down the EN pin voltage.
As soon as the EN pin voltage drops below the turnoff threshold ENL, the eFuse gets turned off and
releases the FLTb pin OPEN. This allows the external RRETRY and CRETRY network to charge from the
supply voltage VIN and restart the eFuse once the EN pin voltage reaches the turnon threshold voltage,
ENR. The time constant of the RC circuit and the EN pin voltage hysteresis determines the retry delay
time. Figure 8 shows the test waveform realizing retry delay of 570 ms. This approach provides a simple
solution to configure the time between retries but does not limit the number of retries. So, the unnecessary
power loss under sustained fault condition cannot be avoided.
VIN
IN
RRETRY
OUT
VOUT
Load
eFuse
EN
FLT
CRETRY
Figure 7. RC Network Delay at the EN Pin Sets the Retry Delay Time
6
Reducing Power Loss and Overheating During Faults with eFuses
Copyright © 2019, Texas Instruments Incorporated
SLVAED9 – June 2019
Submit Documentation Feedback
Techniques to Configure Fault Response
www.ti.com
ENR
ENL
Device auto-retries for
every 570ms
eFuse limits current to
2A for an overload event
RRETRY = 1 MΩ
CRETRY = 1 µF
Figure 8. Example Test Waveform using TPS259541 to Realize Retry Delay of 570 ms
3.2
Using System Microcontroller
Figure 9 shows a block diagram of a typical implementation using a system microcontroller. In this
approach, the system prefers to use the latch-off version where the eFuse latches off for a fault and stays
off until told to turn back on by the system controller. After the eFuse asserts the FLTb pin, the system can
decide on the number of retries and the retry delay interval. For the test case shown in Figure 10, the
eFuse is power cycled at the EN pin for four times with 200-ms delay. As the fault still exists, the device is
latched-off, indicating the need for technician intervention. This approach is feasible provided the system
uses a microcontroller and GPIO pins are available to realize this function, otherwise it adds cost to the
solution.
VIN
IN
OUT
VFLT
eFuse
EN
VOUT
Load
RFLT
FLT
GPIO
GPIO
System µC
Figure 9. Using a System Microcontroller to Control Fault Response of the eFuse
SLVAED9 – June 2019
Submit Documentation Feedback
Reducing Power Loss and Overheating During Faults with eFuses
Copyright © 2019, Texas Instruments Incorporated
7
Techniques to Configure Fault Response
www.ti.com
200ms
Power cycled at EN pin
with 200ms delay
Figure 10. System Attempts Four Retries with 200-ms Retry Delay Before Latch-off
3.3
Devices with Digital Interface
Some of the hot-swap controllers such as the LM25066 and LM5066 have digital (I2C™/SMBus™)
interface and provides flexibility to digitally configure the number of retries to the following:
• 0
• 1
• 2
• 4
• 8
• 16
• Infinite
The retry counts can be selected by setting the appropriate bits in the MFR_SPECIFIC_09:
DEVICE_SETUP (D9h) register. However, the delay between retries depends on the fault timer period
which is a function of the system design margin, thus providing controlled freedom on the selection of the
retry delay. For more information, refer to the LM25066 System Power Management and Protection IC
with PMBus Data Sheet.
3.4
eFuses with Configurable Fault Response
The TPS25982 allows the user to configure the system for latch-off (no auto-retry), limited number of autoretries before latch-off, or auto-retry indefinitely, depending on the system needs. The TPS25982 also
provides adjustable retry delay which allows a sufficient cooling period (which is system dependent), and
prevents the board from overheating in high ambient temperature installations. Both the auto-retry delay
and auto-retry count can be user-programmed using just two low voltage capacitors as shown in
Figure 11.
To configure the TPS25982 for a finite number of auto-retries with a finite auto-retry delay, choose the
capacitor value on the RETRY_DLY pin, then choose the capacitor value on the NRETRY pin as
discussed below.
8
Reducing Power Loss and Overheating During Faults with eFuses
Copyright © 2019, Texas Instruments Incorporated
SLVAED9 – June 2019
Submit Documentation Feedback
Techniques to Configure Fault Response
www.ti.com
TPS25982
RETRY
Retry Logic
NRETRY
RETRY_DLY
CRETRY_DLY
CNRETRY
Figure 11. User Adjustable Auto-Retry Scheme Using Two Capacitors
3.4.1
Setting the Auto-Retry Delay
The time delay between retries can be programmed by selecting capacitor CRETRY_DLY on the RETRY_DLY
pin. The value of CRETRY_DLY can be calculated using Equation 1.
CRETRY_DLY (pF) =
3.4.2
tRETRY_DLY (Ps)
46.83
4 pF
(1)
Setting the Number of Retries
The number of auto-retry attempts can be set by a capacitor CNRETRY on the NRETRY pin using
Equation 2.
NRETRY =
4 u CNRETRY (pF)
CRETRY_DLY (pF) + 4 pF
(2)
The number of auto-retries for TPS25982 is quantized to certain discrete levels as shown in Table 1.
Choose CNRETRY so that NRETRY falls within the range. Use Equation 3 to calculate CNRETRY.
CNRETRY (pF) <
NRETRY u CRETRY_DLY (pF) + 4 pF
4
(3)
A TPS25982xx Design Calculator Tool is also available for simplified calculations.
Table 1. NRETRY Quantization Levels for TPS25982
NRETRY CALCULATED FROM Equation 2
SLVAED9 – June 2019
Submit Documentation Feedback
NRETRY ACTUAL
0<N<4
4
4 < N < 16
16
16 < N < 64
64
64 < N < 256
256
256 < N < 1024
1024
Reducing Power Loss and Overheating During Faults with eFuses
Copyright © 2019, Texas Instruments Incorporated
9
Techniques to Configure Fault Response
3.4.3
www.ti.com
Test Results
Figure 12, Figure 13, and Figure 14 show the flexibility of configuring fault response with the TPS25982
device. In this test case, the TPS25982 is configured for four auto-retries with 100-ms retry delay.
Figure 12 shows successfully starting an unknown capacitive load in three attempts. If output short exists
before enabling the device or output gets shorted in running condition, the device retries only for four
times and then latches-off as shown in Figure 13 and Figure 14, respectively. The configurable fault
response of TPS25982 provides flexibility to recover from transient faults while avoiding unnecessary
power loss and higher operating temperatures under sustained fault conditions.
Systems installed in remote or inaccessible locations would require an indefinite retry setting to allow selfrecovery from faults. Designing such systems with longer retry delay (for example, 1-sec or higher) would
help to maintain lower operating temperatures under a persistent fault. Figure 15 shows the test waveform
where TPS25982 retries for every 1-sec until the short-circuit fault at the output is cleared.
Figure 16 to Figure 19 demonstrate the benefits of using longer retry delay. The longer retry delay allows
a sufficient cooling period to maintain lower case and board temperatures. Under a sustained short-circuit
fault, increasing the retry delay from 100-ms to 1-sec reduces the TPS25982 case temperature from
105.6°C to 61.5°C, thereby enhancing the system life.
10
Reducing Power Loss and Overheating During Faults with eFuses
Copyright © 2019, Texas Instruments Incorporated
SLVAED9 – June 2019
Submit Documentation Feedback
Techniques to Configure Fault Response
www.ti.com
Output builds up for
every retry event
Starts in current limit due to
large unknown capacitive load
Device auto-retries for
every ~100ms
RILIM = 332 Ω
COUT = 2.2 mF
CRETRY_DLY = 2.2 nF
CNRETRY = 2.2 nF
Figure 12. TPS25982 Starts an Unknown Load Successfully (In This Case, Three Attempts)
Short circuit exists at
output before power up
Auto-retries for 4 times with
100ms delay before latch-off
RILIM = 332 Ω
CITIMER = 4.7 nF
CRETRY_DLY = 2.2 nF
CNRETRY = 2.2 nF
Figure 13. When Enabled into Short, the TPS25982 Attempts Only Four Retries Before Latch-off
SLVAED9 – June 2019
Submit Documentation Feedback
Reducing Power Loss and Overheating During Faults with eFuses
Copyright © 2019, Texas Instruments Incorporated
11
Techniques to Configure Fault Response
www.ti.com
Short circuit event
at the output
Auto-retries for 4 times with
100ms delay before latch-off
RILIM = 332 Ω
CITIMER = 4.7 nF
CRETRY_DLY = 2.2 nF
CNRETRY = 2.2 nF
Figure 14. Output Hard Short-Circuit: TPS25982 Attempts Only Four Retries Before Latch-off
Short circuit event
at the output
Output recovers after short
circuit fault is cleared
Device auto-retries for every
1-sec till the fault is cleared
RILIM = 100 Ω
CITIMER = 4.7 nF
1-sec
CRETRY_DLY = 22 nF
CNRETRY = Short
Figure 15. TPS25982 Continues to Power Cycle with 1-sec Retry Delay till the Fault is Cleared
12
Reducing Power Loss and Overheating During Faults with eFuses
Copyright © 2019, Texas Instruments Incorporated
SLVAED9 – June 2019
Submit Documentation Feedback
Techniques to Configure Fault Response
www.ti.com
Short circuit event
at the output
Device continues to auto
retry for every 100ms
100ms
RILIM = 100 Ω, CITIMER = 4.7 nF, CRETRY_DLY = 2.2 nF,
CNRETRY = Short
Figure 16. TPS25982 eFuse Configured for Indefinite
Retires with 100 ms Retry Delay
Figure 17. Case Temperature of the TPS25982 Device
Under Sustained Fault with 100 ms Retry Delay
Short circuit event
at the output
Device continues to auto
retry for every 1-sec
1-sec
RILIM = 100 Ω, CITIMER = 4.7 nF, CRETRY_DLY = 22 nF,
CNRETRY = Short
Figure 18. TPS25982 eFuse Configured for Indefinite
Retires with 1-sec Retry Delay
SLVAED9 – June 2019
Submit Documentation Feedback
Figure 19. Case Temperature of the TPS25982 Device
Under Sustained Fault with 1-sec Retry Delay
Reducing Power Loss and Overheating During Faults with eFuses
Copyright © 2019, Texas Instruments Incorporated
13
Conclusion
4
Conclusion
•
•
•
•
5
www.ti.com
Auto-retry function is generally preferred in many applications to avoid permanent system shutdown for
any temporary faults and to allow systems to try and recover from faults without need for technician
intervention.
However, indefinite auto-retries can lead to higher device case temperature and board overheating in
high ambient temperature installations, thus raising concern on long term reliability under a persistent
fault.
Workaround solutions need additional components or sophisticated devices to configure the fault
response.
The TPS25982 smart eFuse provides flexibility to configure the system for latch-off (no auto-retry), limit
the number of auto-retries before latch-off, or auto-retry indefinitely as per the system needs, by using
just two low voltage capacitors.
References
1. Texas Instruments, Basics of eFuses Application Report (SLVA862)
2. Texas Instruments, TPS2595xx, 2.7 V to 18 V, 4-A, 34-mΩ eFuse With Fast Overvoltage Protection
Data Sheet (SLVSE57)
3. Texas Instruments, TPS25982 2.7 V to 24 V, 15-A, 2.7-mΩ Smart eFuse With 1.5% Accurate Load
Current Monitoring and Adjustable Transient Fault Management Data Sheet (SLVSEI3)
4. Texas Instruments, TPS25982xx Design Calculator Tool
14
Reducing Power Loss and Overheating During Faults with eFuses
Copyright © 2019, Texas Instruments Incorporated
SLVAED9 – June 2019
Submit Documentation Feedback
IMPORTANT NOTICE AND DISCLAIMER
TI PROVIDES TECHNICAL AND RELIABILITY DATA (INCLUDING DATASHEETS), DESIGN RESOURCES (INCLUDING REFERENCE
DESIGNS), APPLICATION OR OTHER DESIGN ADVICE, WEB TOOLS, SAFETY INFORMATION, AND OTHER RESOURCES “AS IS”
AND WITH ALL FAULTS, AND DISCLAIMS ALL WARRANTIES, EXPRESS AND IMPLIED, INCLUDING WITHOUT LIMITATION ANY
IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT OF THIRD
PARTY INTELLECTUAL PROPERTY RIGHTS.
These resources are intended for skilled developers designing with TI products. You are solely responsible for (1) selecting the appropriate
TI products for your application, (2) designing, validating and testing your application, and (3) ensuring your application meets applicable
standards, and any other safety, security, or other requirements. These resources are subject to change without notice. TI grants you
permission to use these resources only for development of an application that uses the TI products described in the resource. Other
reproduction and display of these resources is prohibited. No license is granted to any other TI intellectual property right or to any third
party intellectual property right. TI disclaims responsibility for, and you will fully indemnify TI and its representatives against, any claims,
damages, costs, losses, and liabilities arising out of your use of these resources.
TI’s products are provided subject to TI’s Terms of Sale (www.ti.com/legal/termsofsale.html) or other applicable terms available either on
ti.com or provided in conjunction with such TI products. TI’s provision of these resources does not expand or otherwise alter TI’s applicable
warranties or warranty disclaimers for TI products.
Mailing Address: Texas Instruments, Post Office Box 655303, Dallas, Texas 75265
Copyright © 2019, Texas Instruments Incorporated
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Related manuals

Download PDF

advertising