ESD sensitivity of low frequency discrete devices

[engr_joni_ee]

Hi, I have seen mostly on integrated circuits and RF devices that they are ESD sensitive. What about low frequency discrete devices ? Are they less ESD sensitive ?

 

[dick_freebird]

You'll often see sub-500V HBM ESD ratings
on smaller discretes. I've seen RF parts with
10V ESD failure threshold -results-. Power
MOSFETs sometimes have built in gate
clamps and sometimes not, ones without
may have 100V gate ESD while protected
ones might deliver >1000V.

You won't go wrong by assuming the worst
and preparing accordingly.

 

[KlausST]

Hi,

What are "low frequency discrete devices"?
I assume they come with a datasheet... there I'd look for this information.

Klaus

 

[engr_joni_ee]

Which type of devices are more sensitive to ESD ? What are the units to measure ESD

 

[BigBoss]

No matter RF or LF Device, ESD is always an issue for ALL Semiconductor Devices.

 

[dick_freebird]

Which type of devices are more sensitive to ESD ? What are the units to measure ESD

You do not "measure ESD" directly. You test
and retest a part subjected to increasing
"voltage" levels, as part of a standardized
threat-source model (human body, charged
device, machine model, back in the day, but
now supplemented by additional specs that
apply to particular application segments
(e.g. "dumb consumer on carpet" applied
to products that are attached to accessible
connectors like HDMI or USB chips).

The "number" is the last passing voltage,
where the part still meets both spec table
and any assigned parameter drift limits.
It is full pin-pin, for "handling" ESD (HBM,
CDM, MM) but special cases might be
applied to parts that need to meet powered,
in-system ESD / EOS threats (you would not
care about, say, the inward-facing enable
or data pins beyond sensible "handling ESD"
expectations, but demand more from the
connector-side pins against each other
and any exposed power / ground pins.

When you see a "500V HBM" part that means
it was OK after a one-time 500V "zap set"
(all pin-pin combos) and failed subsequently
(or, the testing was ceased after reaching an
"acceptable" (per the person doing it, or who
set the bar) outcome. Properly, you would
test-to-failure some population of parts and
report the highest all-passing level.