Don, I think you were onto something.?
I made a few changes to the setup since the previous session.? I used to have just one of the low-voltage secondaries loaded.? Now, the 6.3 V secondary is loaded with 0.2 ohms and both 5V secondaries are loaded with 1 ohm.? This causes the currents in these windings to be about what they would be in normal operation.? I¡¯m still energizing the transformer with the 31 Vrms source.? I inverted the current waveform to make it easier to see the phase relationship with the voltage on part of the HV secondary.? I¡¯m still using the Omron G3MB-202P SSR to turn the power on and off.
It turns out that if the transformer ends a cycle with one polarity and starts the next cycle with the same polarity, the current on one side of the AC cycle starts out higher than usual in most trials.? The magnitude of the extra current varies.? But if it ends on one polarity and re-starts on the other, it operates normally.? Example images 14 (opposite) and 17, 19 (same) illustrate this.? Image 27 shows how the extra current decays over several cycles.
To investigate further, I tried to pre-magnetize the core by applying power through a diode just before a turn-on test.? Afterwards, turn-ons with one initial polarity gave extra current while turn-ons with the other polarity looked normal.? Image 63 shows one case where not only did current increase on one side of the cycle, but it also decreased on the other side of the cycle.
But at no time during these experiments did I see a voltage spike.? This only serves to show the magnitude and nature of excess current that can flow due to remanance in this transformer.?
The primary current during these tests was about 0.7 A rms.? Actual current when not transmitting is about 1.4A rms.? Thus, the transformer magnetization would be about double in actual use relative to the conditions of this test.? Thus, the surge current would be at least the fraction shown here and possibly a little more.? It¡¯s not likely to be much more than double because the transformer is designed to transform twice again as much power when transmitting at full tilt and should not saturate at that time.
Using an SSR at turn-on does not enable one to select whether the polarity at turn-on will be the same polarity as it was at the prior turn-off.? Use of an SSR or TRIAC at turn-off might even maximize the remanence in the transformer by ensuring that it has a full half-cycle of current at one polarity just before power is removed.? But it might be more gentle on the transformer by avoiding multiple on-off cycles during turn-on using a switch with mechanical bounce.
Next post: a few findings about voltage spikes which are more likely to be relevant to the initial purpose of this enquiry.
Cheers
Halden VE7UTS
?
