Wilson Lee, Director of Product Marketing, Newark element14
Engineers may underestimate the importance of thermal management, but it is critical to utilize techniques at the
production level that reflect the real environment in which
their build or components may exist. In my experience, I have
found engineers make two common mistakes during tempera-
ture management. The first mistake is engineers rely strictly
on the product specification sheet, which tests components in
a lab setup, without a true frame of reference for how the part
operates or its functionality within a larger design environ-
ment. There are nuances to the design process that the spec
sheet won’t consider: forced air cooling for an out-of-bound
temperature range, power source reliability and quality/clean-
liness of the operating environment, to name a few. Instances
where a design engineer is designing components for use in
another part of the world can pose chal-
lenges too, as they are sometimes not able
to sufficiently simulate the in-use envi-
ronment. Disregarding the real operating
environment in which components are used
can result in thermal failure.
A second common thermal management mistake, partic- ularly in extended temperature designs, is that engineers
expect full performance characteristics across wide temperature ranges. This leads to over compensation or over driving
of the devices, which often leads to component failure. In
those situations, it is important to take into consideration
performance drop offs and to build backup and redundancy
measures. Also, it is important to integrate other mechanisms into the cooling process, rather than relying solely on
forced air cooling. There are several reliable options available: temperature sensors integrated into the board itself;
and thermal padding to isolate “hot spots” on the board.
As board space continues to shrink, engineers are further
challenged in how to incorporate design elements that
manage temperature. However, in any extreme environment
where thermal management is not well thought-out, new
designs are more likely to fail and experience catastrophic
(immediate) rather than soft failures.
John Gammel, Senior Staff Applications Engineer, Sensor
Products, Silicon Labs
When it comes to environmental monitor- ing, few designers consider the benefits
of including relative humidity (RH) sensing in
their systems. There are numerous cases where
sensing RH is useful:
•A telecommunications equipment cabinet requiring cli-
mate control for thermal management. Increasing RH due to
higher cooling coil temperature can indicate air condition-
ing problems ahead of rising temperature, giving an early
indication of potential system failure before electronics are
damaged or an outage occurs.
•Outdoor electronics that have a weatherproof enclosure
that vents to the outside. Generally, in this kind of cabinet,
RH levels will be low even if outside humidity is high due
to heat generated in the cabinet. However, if water accumulates in the cabinet, this will not be the case. High RH in this
type of cabinet is a good indicator of liquid water, which
could lead to corrosion.
•An electronic system used in a rugged application with
a watertight enclosure. A small amount of desiccant in the
enclosure should keep RH low. RH measurement provides a
measure of the integrity of the seal.
Sensing humidity is no longer a complex or board
space-consuming task. In fact, highly integrated CMOS-based, single-chip RH and temperature sensors are now
available in small surface mount packages that operate on a
single I2C interface. Many RH sensing devices are factory
calibrated, meaning no added system test cost. RH sensors
are also available with optional factory-installed covers to
protect the devices from dust and liquids and to ensure reliable RH and temperature measurement needed to safeguard
sensitive electronics in extreme environments.
Ramesh Khanna, Texas Instruments
In harsh environments where temperatures can be as high as 210°C, it is critical to have proper thermal management and thermal protection.
You cannot overestimate the importance of
sensing system temperature as it’s a critical ele-
ment that protects the system. There are various approaches
to protect the system including:
1) Resistance temperature detector (RTD): Semiconductor
IC sensors available in digital as well as analog output have
a linear increase in resistance as temperature rises, they are
commonly constructed of platinum.
2) Negative temperature coefficient thermistor (NTC) are
best suited for precision temperature measurements. The
resistance of the material is linearly proportional to the
3) Positive temperature coefficient thermistor (PTC)
are used in conjunction with an op amp. PTCs work as
temperature monitors. PTCs are best suited for switching
applications, where the resistance rises suddenly at a critical
4) Platinum resistance temperature detectors (PRTD) are
very stable temperature sensors, which are not affected by
corrosion or oxidation. PRTDs are a resistive device and
require an excitation current, or a constant current drive.
Voltage is then read across its terminals. Using the Kelvin
connection force leads provide constant current drive to
PRTD. Constant current introduces voltage drop in the
The first mistake is that
engineers rely strictly on the
product specification sheet.