Hi Eric:
There are three sets of insulations in a switching-mode
power supply:
1. Primary-to-ground. (Basic insulation)
2. Primary-to-secondary. (Double or reinforced insulation)
3. Primary, pole-to-pole and
component-to-component. (Operational or functional
insulation)
I presume your question is in regard of the last one.
By definition, these are operational/functional
insulations, not safety insulations. (The first two
are safety insulations.)
(One wonders why the various standards committees
specify spacings requirements for operational/functional
insulations? If these spacings are so important, why
is primary circuit SOLID operational/functional insulation
ignored? Also, why is operational/functional insulation
electric strength ignored?)
Nevertheless, we are required to apply spacings to the
operational/functional insulation of a switching-mode
power supply. To do so, we must divide the SMPS circuits
into manageable units. We have:
EMC filter
Rectifier and energy storage
Transformer
Power switching
PWM control
PWM LV power supply
Here's a simplified block diagram:
+------+ +-----+
| | | | Transformer
--->| |--->| |-------+-------------------+
| EMC | |Rect-| | | +---...
|filter| |ifier| +------+ )||(
--->| |--->| & |-+ |LV dc | Pri )||( Sec
| | | cap | | |supply|--+ )||(
+------+ +-----+ | +------+ | )||(
| | | | +---...
| | +-------+ |
| | | PWM | +------+
| | |control|--->|switch|
| | +-------+ +------+
| | | |
| | | |
+-----+------+------------+
EMC filter:
No problems. Pole-to-pole spacings.
Retifier/capacitor:
No problems. Pole-to-pole spacings. Except, of
course, we have two diodes (switches) across the line.
(Since these constitute alternating switches, I wonder
why spacings are required?) For the purposes of the
requirements, the spacings only apply to the PWB.
Transformer:
This has the full dc voltage across it when the
switch is closed. It has zero volts when the switch
is open. Therefore, the spacings across the transformer
must be based on the dc voltage.
However, the switching action causes transients to
appear across the transformer. The spacings must take
into account the magnitude of the switching transient
voltage.
Switch:
Since this is a switch that alternates opening and
closing, no spacings are required.
LV dc power supply:
Since this is a non-isolating supply, it is essentially
a voltage divider. The spacings across each part of
the divider are proportional to the voltage.
PWM control:
Usually, this is a low-voltage circuit. So, no spacings
are required within the circuit. But, spacings are
required between the low voltages and the higher voltages
in the supply.
(Usually, the PWM control and the switch are separate
devices. Some low-power SMPS use a single chip for both
the PWM and the SMPS, in which case all spacings are for
the high voltage.)
(High-power SMPS use multiple switches, but the principles
described here apply.)
Measuring voltages within a SMPS can only be done with an
oscilloscope. Furthermore, to measure accurately, the
scope must be an isolating scope. Voltages cannot be
accurately measured with conventional oscilloscopes, even in
the differential mode (insufficient common-mode rejection).
Because the spacing requirements are based on transient
overvoltages, it is also necessary to measure the voltage
between primary and ground, and between primary and secondary.
These spacings, too, must be dimensioned according to the
measured voltage (if you want strict adherence to the standard,
although many certification houses ignore these voltages).
I make a paper copy of the circuit board (both sides) and then
highlight the various voltage differences circuits. Then,
measure the voltages, determine the spacings, and see if the
board complies.
The spacings specified in various standards are based on
"standardized" overvoltages on the mains. Such overvoltages
simply don't appear on the dc side of the rectifier due to
the action of the EMC filter, the rectifier, and the capacitor.
(Put a scope on the dc when you do a transient test!) Yet,
our standards presume the transients do appear on the dc side
and furthermore coincide and stack up on the SMPS switching
transients! Imagine the coincidence of a single 2 microsecond
pulse occuring at the same time as a 100 kHz fractional
microsecond SMPS transient!
Be sure to check with your certifier. They will have some
guidelines for SMPS operational/functional spacings. Some
rules-of-thumb as to where to apply the spacings and what
values to use. They'll probably be relaxed from those I've
described here. They're beginning to recognize that these
are functional/operational spacings, and that they are not
critical to safety.
And, you can always fault the spacings in lieu of meeting
the dimensional requirements. Rather destructive, requires
many samples, and can be time-consuming.
Best regards,
Rich
---------
This message is coming from the emc-pstc discussion list.
To cancel your subscription, send mail to [email protected]
with the single line: "unsubscribe emc-pstc" (without the
quotes). For help, send mail to [email protected],
[email protected], [email protected], or
[email protected] (the list administrators).
