Static - Re: [PLUG-TALK] "regular" compressed air

[Keith Lofstrom]

On Fri, Jan 20, 2006 at 02:59:51PM -0800, alan wrote:
> 
> My understanding is that if you use a regular air compressor, the static 
> charge will kill most electronics.  (As well as being a bit too powerful.)

Others have talked about regulating the flow/pressure down, I'll 
address the static and the moisture.

When air is compressed, it heats up - part of the energy of compression
is turned into heat.  When air is allowed to expand, it cools down;  
part of the thermal energy stored in the air is turned into mechanical
force, so it cools down.   If the air is close to the dew point (like
the air in your unheated garage), then cooling it will cause some of the
moisture to turn into droplets.  If you can keep air from making dew,
then the moisture actually helps make it conductive, and reduces the
potential static buildup.  That's why a proper pressure regulator is
good - it helps the expanding gas turn into the appropriate air velocity
(good) rather than cold gas (bad). 

Most chips are potentially sensitive to static - it doesn't take much
energy at all to wipe them out.  To combat this, well-designed chips
have special circuitry on the pins to route ElectroStatic Discharge
currents safely from pin to pin.  However, not all chips have properly
designed "ESD protection", and the innards of the chips do not have any
significant protection at all.  Because many chip tests do not verify
properly functioning ESD protection on each pin, it is possible for a
circuit to perform its electronic function and pass the test, while
the ESD protection on a particular pin may be broken.  Multiplied by
thousands of pins on hundreds of chips in a computer,  there is a
significant chance that one of those pins somewhere in that case is
ready to be zapped by smaller-than-normal amounts of ESD.

In semiconductor labs, where the chips are not packaged yet and the 
innards are exposed, we are extremely sensitive to even tiny amounts
of charge stored in the gas from a pressure jet.  Adding insult to
injury, we are usually using pure nitrogen (from boiled-off liquid
nitrogen) because it is cleaner than compressed air.  However, that
nitrogen has zero moisture and is quite non-conductive; to get 
around that, the nozzles of the air guns often have a radioactive(!)
plug in them.  Typically the plug is an alpha emitter like Americium
that ionizes the air travelling through it without emitting particles
that can travel significant distances and hurt people (just don't
inhale or eat the Americium!).  Ionized air is not "charged" air per
se, but the ions are good at conducting electricity and preventing
little packets of air from developing enough charge to produce tiny,
tiny little zaps capable of roasting a femtogram-sized transistor.

And no, you are probably not allowed to buy one of these little
Americium plugs or the air guns that use them, for consumer use.
But that is what the manufacturers are using.

So, if you use a traditional air compressor and "automotive shop"
style air gun, which doesn't regulate the pressure, you can generate
charged air that has some small chance of damaging an otherwise
functional chip with an occasional failed ESD protection circuit.  
Or you can generate dew that corrodes a close-to-failure connection. 
Most of the time, the circuitry does not suffer from these defects,
so most of the time you are safe.  If you want to be safe more often,
find a pressure regulator that will deliver appropriately warm and
below-dew-point-moist air.  Since I don't run an electronic repair
shop,  I cannot point to specific solutions, but I'm sure they are
out there.

Keith

-- 
Keith Lofstrom          keithl at keithl.com         Voice (503)-520-1993
KLIC --- Keith Lofstrom Integrated Circuits --- "Your Ideas in Silicon"
Design Contracting in Bipolar and CMOS - Analog, Digital, and Scan ICs