Hello,

I have been developing a remote data acquisition system composed by one data acquisition and operation circuit which reads values from 4-20mA sensors and sensors with digital pulse outputs, one circuit to received solar energy and recharges a 12 V battery, and a circuit to do the 12v -> 24 v conversion (to supply the 4-20 mA sensors). A battery supplies the power to all the circuits and sensors.

Besides the EMC protection, what others kinds of protection should I implement in each circuit I have?
- ESD?
- Power input protection for overvoltages ? Using TVS?
- Power input protection for reverse polarity ? Using diodes in parallel ?
- Current-limiting circuit for each 4-20mA sensor present?
- What about an earth connection? Should I connect it to the battery GND?

Is there any book that treat these issues above?

Thank you,
rprodrigues

 

Sounds like a robust design in progress. I'd be wary of earthing 0V though - keep it floating from the planet if you can so as to avoid earth loops when connected to other equipment. Earth the metal chassis if need be, and take any input pi filters to this earth, but only earth 0V if safety demands it, for example a guitar amplifier.

Series input resistors cure many ills - they provide some current limiting and ESD protection, and work well going to inputs that already contain diode clamps to the supply rails.

TVS and similar devices are also good design-ins. 4-20mA circuits running off a 24 V rail normally use 36 V rated suppressors, as the 24 V rail can be up to 32 V in industrial applications.

I prefer parallel reverse-voltage protection diodes as they're more efficient in operation and don't raise the input impedance on the power line. The downside is they crowbar the power supply (and so need to be meatier than a series diode for the same circuit), but that's just punishment for plugging a circuit in backwards I reckon.

You won't find answers to all your issues in the same book, but here's a couple of handy ones:

"EMC for product designers" 2nd edition, by Tim Williams - the definitive reference.
"Troubleshooting analog circuits" by Robert A Pease - just because. A bit more random in its content, but full of juicy nuggets nonetheless.