Testing surge protector by giving it high voltage when 380VAC is present

[samuelr]

Hi

I am trying to test my surge protector. I have a HV pulse generator which will give me a peak voltage of 15kV connected parallel to a 380VAC and inbetween is the TDK surge protector. There is 3kohm and 3.3mH in series.

When the 380VAC is absent (off) the pulse generator works nicely.
However its not going all the way to 10kV when the 380VAC is present (on)

I believe this should work and the firing of the pulse is at 0VAC of the 380VAC ( using zero-cross detect)

The best I am getting is 3kV which nearly not enough to do any damage on the surge protector or see how effective it is.

Any advise?

 

[chuckey]

You are only getting 3 KV because the surge suppressor is working?
Frank

 

[westom]

The best I am getting is 3kV which nearly not enough to do any damage on the surge protector or see how effective it is.

Any advise?

Voltage is not the independent value. In surge protection, the independent value is current. Voltage (a dependent value) should rise as much as possible with current limited to single digit milliamps. Obviously width of that single test pulse is important since the setup could output somthing less than 3 amps.

Also appreciate that you are only measuring its normal mode response. Critical is its longitudinal mode response. IOW a protector is only effective when part of a system that includes another 'system' part: the protection.

 

[samuelr]

My concern is that now there is 2 different voltage sources connected in parallel.
I know that if I strike the surge protector when the 380VAC is switch off it would show as high as 10kV, (surge protector I used is of higher value) but when introducing the 380VAC it seems to "eat" some of this high voltage and makes it 3kV and the pulse width of the HV is reduced from 1ms to 1us. It should work if the strike is done at 0V of the 380VAC input. right?

 

[westom]

No. Two voltage sources are not described. Described is a voltage source (that is problematic) and a current source (that does the testing). Applying 380 volts does not accomplish anything.

Previous post describes what is important. A current pulse (not a voltage pulse) that is milliamps.