Hello everyone.
Does someone know where I can find a bidirectional TVS to clamp a voltage superior or inferior to +-1.5V? I am using an IC that states that the voltage on certain pins can not be superior than the values presented.
Thank you very much.

 

Does someone know where I can find a bidirectional TVS to clamp a voltage superior or inferior to +-1.5V? I am using an IC that states that the voltage on certain pins can not be superior than the values presented.

A TVS isn't typically used in this application. They're used to clamp voltages from high current sources and are sometimes used in conjunction with fuses to prevent fires.

What voltages are you trying to protect the device from? What is their sink/source capability? It's unusual for devices to allow voltages more than a diode below ground or above VCC. What device are you using and at what supply voltage?

 

4-generic Diodes in series will probably do what You want,
but You gave very little specific information.
.
.
.

 

Hello everyone.
Does someone know where I can find a bidirectional TVS to clamp a voltage superior or inferior to +-1.5V? I am using an IC that states that the voltage on certain pins can not be superior than the values presented.
Thank you very much.

Why it specifically needs to be TVS type ? Are you prone to have ESD strike on those pins, or why high speed diodes are required ???
There are multiple ways to limit voltage but there are also several reasons why you choose that particular method, that can be valuated here. Draw a circuit what you have.

 

Hello everyone. Thank you very much for your replies.
The IC is BL0937, an energy metering device. For IP, IN and VP pins, it is stated that:
"Because of the internal ESD protection circuit, If the voltage exceeds ± 1.5V the IC will not be damaged badly. "
I am using a current transformer for current (AC1020, 1000:1) and a current-type voltage transformer for voltage (ZMPT107-1, 1000:1000), in order to get an isolated circuit.
Imagine the following scenario: an huge current is presented at the current sense pins, which will result in a high voltage exceeding the ± 1.5V. I want to make sure the circuit is capable of withstanding this situation.
Do you suggest another type of component other than an TVS? A Zener diode may not be fast enough, and as for MOV, I did not find some with such a low voltage.
The circuit would be something like the photo attached.
As for the suggestion of LowQCab, the problem with diodes in series, is that when they are correctly polarized, although the current would not be high, it could be significant to influence the measure voltage.

 

"Because of the internal ESD protection circuit, If the voltage exceeds ± 1.5V the IC will not be damaged badly. "

Chip is already protected for ESD or other surge current, you sure you need more protection ?

 

For each input that needs protection I suggest a pair of silicon diode of opposite polarity connected between the input and the common. Those diodes will start to conduct around 600 millivoltsand be fully conducting by 800 millivolts. Since the maximum useful input is 500 millivolts that provides an adequate margin. Select diodes with a fairly fast response time and rated to carry at least half an amp. And use a 100 ohm resistor in series with the voltage source for the input, unless there is already a series resistor.

 

There are several circuit examples online, none of them have any external protection, except series resistors and caps.

 

There are several circuit examples online, none of them have any external protection, except series resistors and caps.

View attachment 248472

NONE of the circuits in post #8 is a clamp circuit. two are attenuators, a rather poor choice, and the last one is a power supply, complete with a regulator.
THE TS is looking to prevent an input from exceeding 1.5 volts either positive or negative. But since the useful input is much less, a lower clamp voltage provides a greater margin of protection.

 

NONE of the circuits in post #8 is a clamp circuit. two are attenuators, a rather poor choice, and the last one is a power supply, complete with a regulator.

oh really.

 

Chip is already protected for ESD or other surge current, you sure you need more protection ?

It says that if the voltage exceeds +-1.5V it will not be damaged severely. I want to make sure short-currents for current and surges for voltages will not damage the IC. This will be used in an industrial environment.

For each input that needs protection I suggest a pair of silicon diode of opposite polarity connected between the input and the common. Those diodes will start to conduct around 600 millivoltsand be fully conducting by 800 millivolts. Since the maximum useful input is 500 millivolts that provides an adequate margin. Select diodes with a fairly fast response time and rated to carry at least half an amp. And use a 100 ohm resistor in series with the voltage source for the input, unless there is already a series resistor.

I thought of it too, but my problem with putting two diodes of opposite polarity is that even if they do start conducting at 600 mV, in reality below that voltage they are already conducting although small currents (uA to mA), so they will be influencing the measured signal. Am I right?

NONE of the circuits in post #8 is a clamp circuit. two are attenuators, a rather poor choice, and the last one is a power supply, complete with a regulator.
THE TS is looking to prevent an input from exceeding 1.5 volts either positive or negative. But since the useful input is much less, a lower clamp voltage provides a greater margin of protection.

I did not understood:
"THE TS is looking" Who/which is the TS?
"lower clamp voltage provides a greater margin of protection" For example, the diodes you mentioned?

 

A clamp circuit is by definition NONLINEAR. The fundamental concept of a clamp arrangement is to not affect the signal below some voltage level, and then to prevent the amplitude of the output from exceeding that voltage level.
And now we see two conflicting statements, one telling us that damage will happen if the voltage exceeds 1.5 volts, and the other telling us that "the internal protection diodes will prevent any damage.".
Quite likely there is a limit to the power that those internal protection diodes can handle, and so an external clamp arrangement makes a lot of sense. AND, given that the specification claimes a usable input voltage of up to 0.50 volts, using a regular silicon diode that goes into full conduction around 0.7 volts is a reasonable protection scheme if a mens to limit the current is added. That would be a lower-value series resistor, possibly 1000 ohms.
The reason for not using a simple voltage divider is that to provide the greatest accuracy the intended range of inputs should match the full-scale range of the IC reasonably closely.

"TS" is the term used to refer to the Thread Starter. It is a commonly used term of reference on this site. Some folks use "OP", short for Original Poster. Both refer to the individual who asked the question.

 

Thank you @William Ketel . I will experiment with the diodes and resistor and I will see what happens.

 

oh really.

Yes, really.

 

The datasheet says that the absolute maximum voltage on the analogue pins is ±4V.
Use back to back 3V zeners.

 

The datasheet says that the absolute maximum voltage on the analogue pins is ±4V.
Use back to back 3V zeners.

Thank you @Ian0

 

The datasheet says that the absolute maximum voltage on the analogue pins is ±4V.
Use back to back 3V zeners.

No matter what sort of shunt clamping protection is used, a series resistorto limit the current will be required. And if the anticipated operating range is much lower, the higher clamp limits will not provide much added benefit. Also, zener diodes may have a lower current handling ability, and certainly they will cost more than silicon power diodes.

 

The recommended circuit has 1k resistors and 33nF capacitors for filtering. It looks as though it was lifted from the Atmel ATM90e26 datasheet.

 

The effect of a clamp arrangement on the voltage at levels below the clamp voltage does depend in the source impedance of the voltage source. A current transformer is a fairly low impedance voltage source, and so the few microamps drawn should not have a large effect. And the diode forward current at lower voltages varies quite a bit between different types of power diodes, so some testing will tell a lot about which type number will work.