Power-supply decoupler protects your UUT
Raju Baddi, Tata Institute of Fundamental Research, Pune, India - May 10, 2012
You can install the bench-top
accessory in Figure 1 between a
variable-voltage bench power supply
and a breadboard or UUT (unit under
test) to protect against accidental overvoltage
and reverse polarity. It draws its
power from the supply and pulses a 5V,
two-coil latching relay to interrupt
power to the load under abnormal conditions.
The latching relay uses permanent
magnets to hold the DPDT (double-pole/double-throw) contacts in the
position, which the most recent pulse
to the associated coil sets. It is nominally
rated at 5V for the coils but operates
at voltages as low as 3.5V.
In the normal state, the relay passes power from the attached supply at the input voltage to the load at the output voltage through the inductor and holds Q3’s emitter at 0.6V less through BR1. C2 charges to 1.2V less than the input voltage. C1 cannot charge through reverse-biased D1 or D2.
Q1 forms a variable-zener function, which you adjust by using the coarse and fine potentiometers to set the overvoltage-threshold base to the emitter voltage. If the voltage should exceed the threshold—for example, if a user accidentally bumps the power-supply voltage knob—the base-to-emitter voltage increases to the 0.6V necessary to turn on Q1. This action then turns on Q2, which, in turn, turns on Q3 through D1. Q3 draws current from C2 through the appropriate relay coil to open the contact to VOUT and BR1 and close the contact to the collector of Q4 and an ac input terminal of BR2, which lights the error LED. The charge on C2 allows the relay to fully complete its latching action, even though it has disconnected its own voltage supply.
You should calibrate a dial for the
potentiometers or adjust them with the
power supply on but with the UUT load
unconnected. This approach helps to
quickly and easily set the decoupler’s
overvoltage threshold.
Voltage follower Q4 charges its emitter capacitor to no more than 4.5V, which the 5.1V zener diode at its base sets. After the overvoltage condition is cleared, momentarily pressing the RST (reset) pushbutton switch discharges Q4’s emitter capacitor into the other relay coil, latching it back into its normal position.
If you accidentally apply reverse polarity to the VIN and ground terminals, Q3 biases on through D2. BR1 applies the normal polarity to C2 and the relay coil to allow Q3 to operate the relay for the overvoltage condition. BR2 lights the LED despite the reversed polarity.

The data sheet usually specifies the relay’s operating time (Reference 1). Alternatively, you can measure it with an oscilloscope or a dual-event timing circuit (Reference 2). The circuit uses only one of two sets of contacts. If your design requires a second protection circuit for a negative-supply voltage—with NPN swapped for PNP and the diodes reversed—you can cross-couple the extra contacts in series with the opposite supply so that a fault of either supply would disconnect both.
This Design Idea describes a general approach. You should verify that the various parameters meet the requirements of the power-recipient circuit and make appropriate modifications.
In the normal state, the relay passes power from the attached supply at the input voltage to the load at the output voltage through the inductor and holds Q3’s emitter at 0.6V less through BR1. C2 charges to 1.2V less than the input voltage. C1 cannot charge through reverse-biased D1 or D2.
Q1 forms a variable-zener function, which you adjust by using the coarse and fine potentiometers to set the overvoltage-threshold base to the emitter voltage. If the voltage should exceed the threshold—for example, if a user accidentally bumps the power-supply voltage knob—the base-to-emitter voltage increases to the 0.6V necessary to turn on Q1. This action then turns on Q2, which, in turn, turns on Q3 through D1. Q3 draws current from C2 through the appropriate relay coil to open the contact to VOUT and BR1 and close the contact to the collector of Q4 and an ac input terminal of BR2, which lights the error LED. The charge on C2 allows the relay to fully complete its latching action, even though it has disconnected its own voltage supply.
You should calibrate a dial for the
potentiometers or adjust them with the
power supply on but with the UUT load
unconnected. This approach helps to
quickly and easily set the decoupler’s
overvoltage threshold.Voltage follower Q4 charges its emitter capacitor to no more than 4.5V, which the 5.1V zener diode at its base sets. After the overvoltage condition is cleared, momentarily pressing the RST (reset) pushbutton switch discharges Q4’s emitter capacitor into the other relay coil, latching it back into its normal position.
If you accidentally apply reverse polarity to the VIN and ground terminals, Q3 biases on through D2. BR1 applies the normal polarity to C2 and the relay coil to allow Q3 to operate the relay for the overvoltage condition. BR2 lights the LED despite the reversed polarity.

The design includes a 47- to 500-mH inductor and a 1000- to 4700-μF
capacitor at the output stage to delay
the rise of output voltage and current.
This step avoids damage to the recipient
circuit during the operating-time delay
of the relay. Choose the inductor for
sufficient current rating and minimum
dc resistance for the load current you
anticipate.
The following equation calculates
the rise of current, I, through an inductor,
L, as a function of time, T, when a
voltage, V, exists across it: I=(V/L)T.
The next equation calculates the rise
in voltage of a capacitor, C, when a
charge, Q, is deposited into it: V=Q/C.
You can use these equations to calculate
the required values of L and C if the
relay were to operate in time T. You
can obtain the charge by integrating the
first equation with time limits zero to T.
Because the voltage across the capacitor
at the output must be within safe limits
for the recipient circuit, the charge
on the capacitor must be small enough
that its voltage hardly changes, meaning
that the voltage and the current
are more or less constant. In this case,
the charge is (I×T)/2 and I=(V/L)T. Click here to see the circuit layout.
The data sheet usually specifies the relay’s operating time (Reference 1). Alternatively, you can measure it with an oscilloscope or a dual-event timing circuit (Reference 2). The circuit uses only one of two sets of contacts. If your design requires a second protection circuit for a negative-supply voltage—with NPN swapped for PNP and the diodes reversed—you can cross-couple the extra contacts in series with the opposite supply so that a fault of either supply would disconnect both.
This Design Idea describes a general approach. You should verify that the various parameters meet the requirements of the power-recipient circuit and make appropriate modifications.
| References |
|
Optoisolation for intelligent power modules
Why bypass caps make a difference - Part 4: Op amp macromodels: A cautionary tale
Managing inrush and system protection for electrical systems
Teardown: Cell-phone charger: nice idea done right
Cray -1 super computer: The power supply
Some thoughts on DC/DC converters, part five
Has Thomas Edison ultimately won the DC vs AC power transmission controversy against Tesla?
Datasheets.com Parts Search
185 million searchable parts
(please enter a part number or hit search to begin)
KNOWLEDGE CENTER

