1
\$\begingroup\$

Last summer I built a current source to power a custom-made pressure sensor. This source uses a REF200 IC and is powered by a wall wart that outputs 5V. The whole circuit is encased in a plastic box that has holes drilled for the energy input (+5V) and energy output (100uA) cables (see circuit).

The circuit worked well for over 7 months and then mysteriously died one day when a friend was using it. He claims he felt a slight shock when he touched a surface that was connected to the pressure sensor. Yeah whatever.

Except that we built another (equivalent) current source and, after a few days of good performance, I felt a slight shock (when I touched a rubber surface connected to the system) and the current source stopped working. Could this shock (which I'm pretty sure is ESD) have damaged the REF200 IC that was inside its own case?

I suppose the charge could have traveled from my hand to the rubber, jumped to the power output (100 uA) cables, then fried the REF200. But I thought that ICs were mostly vulnerable to ESD when not mounted on a board.

So... did ESD fry my defunct current source? And if so, what is a simple way to prevent this? Will grounding the system help? (This is what my friend suggested)

REF200 current source circuit

\$\endgroup\$
1
\$\begingroup\$

ESD may fry a component in a board, but it's less common. What happens is that the potential damage is inversely proportional to the area of the terminal you touch. When mounted, the copper tracks and vias increase this; also you have paths to ground through passive components that reduce the effective voltage produced by the ESD discharge on it.

But I don't think it could so easily happen unless you touch the terminal or a metal conductor that leads to it; air has about 1000V of maximum isolation per mm, and materials like rubber are even higher than that. You'd need a huge charge in your body.

Anyway, the usual way to protect against ESD is to place an inverted TVS diode between any vulnerable terminal and ground (by 'vulnerable' I mean those that have connections to the outside of your case).

I'd need to understand better what your friend meant when he mentioned 'grounding the system'. If it is replacing the case to a metallic one connected to ground, yes, that'd help a lot. It'd be much more probable that when you touch the device the discharge would go directly through the case to ground than through a vulnerable contact from the circuit.

\$\endgroup\$
  • \$\begingroup\$ What is strange is that he talks about a plastic enclosing... Maybe the sensor broke in the first place... \$\endgroup\$ – Vladimir Cravero Apr 22 '14 at 19:31
  • \$\begingroup\$ Agree, seems the most probable cause. Maybe the ESD damaged the sensor and it became shorted. \$\endgroup\$ – fceconel Apr 22 '14 at 21:20
  • \$\begingroup\$ Unfortunately, the sensor is not the problem. After these 'shocks', the current source does not put out the correct current. Even as it only connected to the wall-wart and an ammeter (the ammeter shows ~12 uA now). Something damaged the electronics in the circuit, but the circuit was not exposed to the outside except through its input/output connections. \$\endgroup\$ – Michael Santiago Apr 22 '14 at 21:28
  • \$\begingroup\$ If the sensor is brokent is in a subtle way. A short circuit is good for a current source, and it does not fry it instantly... \$\endgroup\$ – Vladimir Cravero Apr 22 '14 at 21:28
  • 1
    \$\begingroup\$ It has to be the same ground reference the component to be protected is using. The idea is that when an ESD occurs, it'll flow through the diode and not the protected component. The grounding wire can be used only if it's connected somewhere else (usually your '-' from the wall wart, unless you have a better ground). If you want to be extra safe, use two bidirectional TVS, one between the protected terminal and the '-' and other between the same terminal and the '+' from the wall wart. \$\endgroup\$ – fceconel Apr 23 '14 at 21:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.