Experiment sw V inductance

[gar]

161108-1448 EDT

Following are three results of a simple experiment switching an inductance. The purpose of this experiment was to observe the arc voltage drop following the opening of a mechanical switch at atmospheric pressure in air. This was a quick experiment consisting of a DC source in current limiting to set the initial current in inductor, a reverse biased diode across current source, an ordinary Ohimite power resistor, 1 ohm, for current measurement (means it was very slightly inductive), 1 ufd capacitor in parallel with said 1 ohms resistor, a single Potter and Brumfield KUP relay contact as the switch (10 A rating), and the transformer primary wired in parallel of a A41-175-16 Standard Transformer transformer.

Initially upon opening of the switch contact the switch voltage drop appears to be a constant of about 15 to 20 V. As current drops the voltage gradually increases to around 50 V.

If the arc extinguishes before near zero current, then a rather large voltage spike appears. The stored magnetic energy has to go somewhere.

A more controlled experiment is needed with higher initial current, and I believe we will see the initial voltage drop stay around 15 V.

Many trials were required to get the following three plots.

Blue is switch voltage, and red is current.

.

.

.

.

.

 

[gar]

Slight change in the citrcuit

Slight change in the citrcuit

161109-1415 EST

I have slightly changed the experimental circuit to start with greater initial current.

Initial voltage and current source is a 12 V car battery. Current measured with a 3 A 50 mV shunt with a 1 k 1 ufd filter. Same transformer primary is the inductor. Its DC resistance is about 2 ohms and in combination with the battery sets the initial current measured at about 6.6 A.

To avoid overheating the transformer primary the P & B KUP 24 V DC relay is only pulsed on for several seconds to current charge the inductor and allow time to init the scope single shot mode. This is done by charging a 2800 ufd capacitor to 25 V and then discharging the capacitor into the relay coil. RC time constant is about 1.3 seconds. Relay drop out is about 4 V. 21/25 = about 84%,

Separately I measured the relay contact transit time at about 2 mS, and some times with bounce at transit end.

The voltage and current are more logical in this plot, Voltage scaling is obvious. Current scaling is 0.1 V = 6 A.

Initially the currrent is 6.6 A thru the contacts and coil, and the contacts are closed with contact voltage drop = 0.

At the instant the contacts open the arc voltage jumps to 15 V, then gradually rises to about 25 V over the 2 mS period the contacts are opening. This increase is a result of increasing path length.

After contact transit time (2 mS) the distance for arc length remains constant and as arc current drops there is a gradual increase in arc voltage drop as the currrent decreases, notice the inverse relationship (what does that mean?), to about 50 V where the arc extinguishes. In this sample the arc restarted for a short time at a higher voltage.

When the arc turns off there is still energy stored and some oscillation occurs.

.

.

.

 

[gar]

161109-2141 EST

All the oscillation on the current waveform of the previous post bothered me. Thus, I made a slight change in the test circuit. A battery is not necessarily a good low impedance at high frequencies. Added a 2800 ufd 25 V electrolytic capacitor in parallel with the battery, but close to the switch (relay contacts), and inductor. This capacitor is a different one than the one used to pulse the relay.

Attached is a resulting plot. Much cleaner.
.

.

.

 

[gar]

161110-2047 EST

The above posts were using a P&B KUP relay as a switch.

The initial air gap for any switch upon opening is basically 0. Thus, the voltage drop at this time is essentially the sum of the cathode fall of potential and the anode fall of potential, and essentially 0 drop along the plasma. Anode is typically a few volts, and cathode is greater. The sum is generally less than 25 V. See "Fundamentals of Engineering Electronics", by Dow.

Free electrons are fed into the arc from the cathode, and collected at the anode. Ionization of the gas exists within the plasma. A sheath encloses the plasma with little electron migration thru it.

In all my measurements the sum of cathode and anode fall of potentials were about 15 V. Also all plots showed a negative resistance characteristic. Very large currents are required to shift away from negative resistance.

The KUP contact gap when opem is only about 0.015".

To see how things changed with a larger opened air gap I change the switch to an ordinary cheap hardware store switch. I have no information on its gap. Following are two plots with this switch.

The first plot circuit is no different than the previous post. Because of the expected much longer arc path there appears much more arc noise. The arc extinguishes at a higher current level than for a short plasma. Because of the longer path with more current in the inductor at currernt turn off (larger stored energy is in the inductor at this time), and thus, a lot of oscillation.

In the second plot I added a clamping diode to about +95 V. This removed most of the inductive energy and eliminated the massive oscillation. Here you can see the greater current point where the arc is extinguished. One could do much the same with an MOV or Xener diode.

.

.

.

I can not delete the following plot that does not belong here.
.

.

 

[Ingenieur]

Doing something similar for my transients class
deenergize loaded 109 MVA shunt reactor (has Cu R, core R and side C losses)
Study TRV and freq of fault V
calc chopping current for use in model and hand calcs
PSCAD software
345 kv at 1.025 pu
system avail fault i 35 kA , needed to calc system Xl

Build model and run sim
draw single line and compute by hand
compare sim vs hand calcs
I got a TRV of 2.17 pu By sim and 2.16 by hand
freq by sim 2000 (as best as I can determine) 2012 by hand calc