Hi Jeffrey:
In answer to your questions...
> 1. Advantages / Disadvantages of using a ramping function vs. zero crossing
> ?
Back in the old days, some hi-pot testers simply had
an off-on switch. No ramp. No zero crossing. And,
no problems with the equipment.
The ramp function is absolutely necessary for dc hi-
pot testing so as to limit the charging current and
not trip the tester.
The ramp function is nice if you want to record the
voltage at which a failure occurs. Of course, you
should have no failures so the ramp simply extends
the test time. If you do have a failure, you will
want to examine the unit with a manual voltage
control hi-pot tester in order to pinpoint the
failure.
A zero-crossing start is also nice, but not necessary.
After all, the idea of insulation withstanding the
high voltage is to withstand the normally-occuring
voltage transients on the power line. So, a non-
zero-crossing hi-pot tester better simulates what is
likely to occur in service.
> 2. For ITE requirements Ramp vs. Zero Crossing (Does 1950, 950 care? What
> about ramping rates)
No. Just zap the product for at least 2 seconds.
> 3. What if my product passes using one function but fails using the other?
Not likely.
The startup is not where breakdowns occur. They
occur at 500 volts and more.
If you do see a difference, then you had better find
the root cause and fix it because the product
(nee insulation) is not operating normally.
The solid insulations and clearances required by IEC
60950 and clones results in dielectric withstands of
4500 V rms or more. During your type testing, you
should test to breakdown to find the weakest
insulation in your product. Once the weakest
insulation is identified, you should examine it and
decide whether the design and the production process
controls that insulation so that it will always break
down at or near that voltage. Once you have done
this, then the hi-pot tester parameters are largely
of no matter.
> 4. Which function is acceptable for production line testing? Certification?
Both or neither. Your choice, based on your
particular situation.
If you meet the standard, and have contol over
the weakest insulation (about 4500 V rms), then the
startup is largely inconsequential to the test
result.
> 5. Which settings will allow the most stringent criteria when performing a
> HiPot?
The choice of ac or dc voltage, and the choice of
trip current setting are the most critical.
A colleague here in San Diego has a neat hi-pot test
process. He uses a programmable hi-pot tester.
After the ground-continuity test, he programs the
hi-pot tester for 250 V rms, and measures leakage
current. This tells him if the Y caps are in place
and about the right value because he checks for both
low-limit and high-limit currents. Then he ramps a
dc voltage to the hi-pot test value. He believes dc
is less severe to components, and improves component
reliability over ac. With dc, after the ramp, the
trip current can be set very low.
Voltage breakdown is always preceded by corona
(partial discharge). Dc has much less corona than
ac. So, to minimize false trips, dc is somewhat
better than ac. But, ac is more likely to find
marginal construction.
If the design is not marginal (i.e., minimum
breakdown voltage is 4500 V rms or more), then, in
my opinion, the choice of ac or dc, ramp, trip
current, etc., are largely inconsequential.
Personally, I use ac, and set the trip current to
twice or more the normal current at the test voltage
(because, when a failure occurs, it truly draws more
current than almost any current trip setting).
> My concern is not passing this test but ensuring I've design the most robust
> PCB that's feasible. This particular PCB will have stringent reliability
> requirements as it will be considered a NON-Field replaceable unit.
As previously mentioned, the way to determine the
robustness of your design is to test to failure,
and then decide if that degree of robustness (i.e.,
margin between required withstand voltage and true
breakdown voltage) is good enough for you. If you
have good margin, and if you have control over the
design and manufacturing parameters that control
that margin, then the parameters of your production-
line hi-pot test are inconsequential.
Best regards,
Rich
-------------------------------------------
This message is from the IEEE EMC Society Product Safety
Technical Committee emc-pstc discussion list.
To cancel your subscription, send mail to:
[email protected]
with the single line:
unsubscribe emc-pstc
For help, send mail to the list administrators:
Jim Bacher: [email protected]
Michael Garretson: [email protected]
For policy questions, send mail to:
Richard Nute: [email protected]
