To describe this effect, emission coatings in the tube attract the
contaminants to the electrodes. As the impurities decrease in the
gas atmosphere, the DCBD voltage also decreases. This can result
in variations seen in the measurements during repeated tests on
the same device.
Calculating Expected Peak Voltage
Lightning strikes usually range from 30 to 250 kA with a
very short duration, but the energy seen by a primary protector
at the end of a twisted pair is more like a 500-A, 10/250-µs
surge. There are also ground-potential rises (GPRs), with
surges entering the ground plane and damaging equipment
through the backplane.
Furthermore, there is the belief that a faster rising voltage
will provide a higher impulse voltage. While this belief
is correct, some components can see a decline of the surge
and ramp impulse voltage as the rate of rise of surge voltages
increases (Fig. 3).
The ramp or surge impulse voltage is calculated using VSURGE = (1.2 × VDCBD) + 280 to find the expected peak voltage with a
known DCBD voltage option. This is fairly simple when the only
impact on the surge-impulse voltage rating comes from the shape
of the surge rising-edge voltage waveform.
A Bourns GDT requires a surge voltage of about 15% above the
dc spark-over rating for a 10/1000-µs surge. This increases to 50%
under a 2/10-µs impulse to ensure the GDT operates. This identifies
the stress applied to the secondary protection when the GDT
is about to switch into arc mode by just knowing the impulsevoltage
value of the GDT primary protector.
A common practice is to utilize the 1-kV/µs rating, which is
often 150% above the DCBD rating in high-lightning conditions.
Using the 1-kV/µs figure will provide a worst-case scenario or
maximum tolerance, though at the expense of higher secondary
protection requirements.
As per Figure 4, worst-stress conditions
occur in secondary protection or the equipment
when subjecting the GDT to a 2-kV, 2/10-µs or
8/20-µs surge. Increasing the generator voltage
past this point will reduce the surge impulse voltage, and the
higher current will only be testing the ground return current path.
Determining the surge-impulse-voltage rating of the GDT with a
2-kV, 2-O, 8/20-µs surge will help to ensure coordination with the
secondary protection components.
This test can also be useful for equipment in high-lightning,
remote-access applications. The GR-1089-CORE, issue 4, section
4.7 specification covers lightning protection tests for equipment
in high-exposure premises or online service provider (OSP)
facilities and uses a 10/250-µs generator delivering up to 4 kV at
500 A. Further investigation of this waveform is necessary to see if
4 kV provides maximum stress on the equipment.
Primary protectors have to meet Telcordia GR-974-CORE
and GR-1361-CORE general requirements for telecom-line
protection in the U.S. and ITU-T K.12 for most other countries.
In the U.S., secondary protection needs to meet GR-1089-CORE
or TIA-968-A general requirements and ITU-T K.20/K.21 for
other countries.
Previously, the design of primary and secondary protectors in
telecom systems was totally exclusive to each. Now, standards such
as Telcordia GR-1089-CORE, issue 4 present the first opportunity
for considering the benefits of a GDT during secondaryprotection
design work.
Selecting the Correct GDT
Though considered slow operating overvoltage protectors,
GDTs can handle impulse currents many times higher than faster
technologies like solid-state devices. The components have to
develop enough voltage across their conductors to create an arc
while the gas between the contacts controls the electrical parameters
and makes them repeatable under a variety of conditions.
Additionally, evolving GDT technology now enables fasterswitching
devices that reduce the impulse sparkover rating, but
often at the expense of other key parameters such as tube life and
high glow-to-arc transition currents or high arc voltages. Addressing
budget limits, a component may also employ lower-quality
materials such as the electrode material or electrode coating material,
though at the expense of the GDT life in the field.
Essentially, to select the proper GDT for the job, establishing
the tradeoffs of dynamic parameters, longevity, and price is key to
designing the correct product for the application. For example, it
may not be necessary to design a GDT with a 20-year lifespan for
an end product with a five-year service life.
Tim Ardley is the senior telecom field applications engineer with
the Bourns Inc. circuit protection division. He has more than 20
years of experience in the semiconductor industry and holds a BSc
(Hons) from Luton University in England.