"DoN. Nichols" wrote in message
...
On 2013-09-17, Larry Jaques
wrote:
...
Huh? How can an open harm a circuit? (Other than arcing, and its
resultant damage and chaos?)

The short which should be there is holding the voltage low. If
the open occurs, the voltage shoots up to whatever the supply is
set to,
and this may be too much for something else connected to the same
wires
-- or not too much, but the wrong polarity.

OK, I'll almost buy that. But what happened to those components
-after- the incident but -before- the short dropped the voltage,
hmm?

Who knows -- but give them the best chance, at least. :-)

There are reverse-biased parasitic or intentional diode structures
between an IC's signal pins and its power and ground. These diodes may
not be able to handle much current.
http://www.analog.com/static/importe...09090AN109.pdf
Look at 8-24, "DIGITAL INPUT AND OUTPUT RELATED SCR STRUCTURES IN CMOS
DACS"
"Forward biasing any one of the four diodes shown in Figure 4 can lead
to [destructive] SCR action if the current through them is high
enough."

If you disconnect a car's battery and then ohm out signal wires, the
ohmmeter's voltage is higher than the car electronics' positive power
supply which is at zero volts. Current from the meter may pass through
the input protection diode into the + supply and try to power up the
device, which might then draw enough current to burn out the input. A
few milliamps from a DVM isn't so bad but the 1 Amp I suggested to
measure milliOhms is very capable of causing damage unless the power
supply voltage is set below the forward voltage of a diode.

This PS-305DM here can source 2A with the open-circuit voltage set to
100mV, which should be safe(?). The Analog Devices article warns next
to Figure 6:
The most common cause [of SCR latch-up] is when the input voltage
exceeds Vdd by about 0.3 to 0.7 volts."

jsw