One zener diode vs two standard diodes for overvoltage protection of low-level voltage (analog and data) lines...

Hi guys,

I am working on a design where I have a couple of analog (ADC) lines measuring a range between 0V and 3V3, and a couple of digital data lines transmitting/receiving 0V and 3V3 voltage levels, and I want to protect them from overvoltage.

Hence, I was wondering if you have any reference to read about different configurations for overvoltage protection of low-level- voltage analog and data lines. I am specially interested in advantages/ disadvantages of a zener diode connected between a data line and GND to clamp voltages to 3V3 compared to two standard diodes: one connected between data and GND and the other connected between data and 3V3 (both inversely polarized under normal conditions). Speed? Price?

Best Regards,

Reply to
rfmg
Loading thread data ...

This is what you need: TPD4E001

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

formatting link

Reply to
Vladimir Vassilevsky

o

Hi Vladimir, thanks for the reference.

However, I am thinking of a continuous overvoltage protection, neither transient spikes or ESD. Thinks of the +3V3 data line being driven by a +5V signal. I would like the digital transmission to still be able to work under those condition.

In the same spirit, I would like to protect the analog line (clamp to

+3V3) even if the user attempts to measure voltage over 3V3.

Thanks,

Reply to
rfmg

On a sunny day (Thu, 17 Dec 2009 07:48:38 -0800 (PST)) it happened rfmg wrote in :

The main disadvantage of zeners is the high capacitance. The disadvantage of diodes is that, given enough current, you can lift the supply line above a safe voltage. So if you use the diodes, or a TDO4e001 like Vladimir suggested, then maybe you also need a circuit that prevents the 3.3V from rising too much. That could be a high power zener, as there more capacitance is not important.

Sometimes it is also needed to have some resistors in series before the diodes or zener to limit any current. Depends on what the data lines can make contact with, mains, 24 V DC, something else? For very low speed data lines zeners are the simplest solutions.. or transorbs.

Reply to
Jan Panteltje

A suitable pair of diodes will have much lower capacitance than a zener diode and will therefore have less effect on the wanted signal. Don't forget to include some series resistance to limit the fault current to a safe value for the expected overvoltage. BAV99 double diodes are good enough for most purposes and are cheap.

Clamping to the 3V3 supply needs caution, as the overvoltage signal can easily cause the supply to rise in voltage and damage components by that route. If you are just protecting against electrostatic discharge, the power supply decoupling capacitors will absorb the extra charge, but accidental connection to a 12V power supply, for example, could cause problems.

To avoid this issue, connect the upper diode to a 3V zener diode the other end of which is grounded. This prevents the fault current from getting into any of the active circuits. It is a good idea to prebias the zener with a resistor to the 3V3 supply. This ensures that the wanted signal does not have to charge up the zener at startup which might otherwise cause momentary distortion.

John

Reply to
John Walliker

e supply line above a safe voltage.

much.

ant.

Ok.

iodes

sure,

ething else?

sorbs.

Thanks for the info.

Reply to
rfmg

Thanks for the info, it is just what wanted to hear.

Reply to
rfmg

supply line above a safe voltage.

much.

something else?

transorbs.

A pair of depletion-mode fets, like LND150s, in series with an input will limit the current to about 1 mA up to hundreds of volts of overload. After that, most chip ESD diodes can handle it.

John

Reply to
John Larkin

A zener diode is not very precise and a 3V3 is also quite soft. I usually go for a BAV99 against VCC and GND or if leakage is a concern a BAV199. Then a small resistor between that and the chip pin so that the chip has no chance to get hit with the lion's share of the spike current through its substrate diode. A resistor on the outside is also best but you may not be able to do that for ports that have to drive long data lines.

With diodes you must also watch the rail. If your crcuitry consumes little power and you experience a lot of ESD hits these may drive up your rail. This is because nearly all regulators, whether switch-mode or linear, can only source current but not sink any. A TL431 is often enough to remedy this.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
Reply to
Joerg

o

es.

Thanks for the diodes' suggestion.

Reply to
rfmg

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.