U
user2934303
Recently I've disassembled, and rewound a 12+12 V 2A transformer as an autotransformer for a test in impedance matching, but it ended not being enough for what I wanted (although it did work).
Then I tried using it as an inductor (using the two outer taps) in another application, but it has a very strange behavior: It only seems to let VERY low frequencies, like < 5Hz pass properly, and surely I don't expect it to has such a huge inductance, it only had 150 turns of a thick gauge wire, half the amount of turns it had in its original secondary.
More strange even, frequencies a bit above that (like 20 Hz+) seem to be passed all with equal amplitude, instead of having a decay up to a few kHz at least, before the iron core presents lower permeability.
Its DC resistance is 0.5 ohms and there doesn't seem to be any poor contact or fracture in the wires.
Then, I went to try and measure its impedance VS frequency curve, using an audio amplifier and a resistance in series with it, but even before I found something even more strange: It's behaving non-linearly in the SAME frequency (60 Hz) in very low currents (< 100 mA), much before core saturation or heating should have any effect.
To test if there was no non-linearity in my measurement equipment, I first tested the setup with the reference resistance (R1) plus a load resistor (R2) in series. As expected, the voltage drop in my reference resistance (R1) was proportional to the voltage drop in the load resistor (R2), and crosses the zero as it should. The slope was exactly was it should be too, from its resistance (20 ohm).
Then, I made another measurement, in the same frequency, around the same range of voltages etc, and measured the voltage drop across the reference resistor and the inductor, and it was very nonlinear:
Sure, an inductor is a reactive load, but this should not make it nonlinear. The voltage drop across its legs should still be proportional to the current, which is proportional to the voltage drop across R1, disregard of any phase difference introduced by the inductance (which only affects the difference between the sum of these RMS voltages and the total RMS voltage).
The transformer is behaving as it had a larger impedance at higher voltages, after the voltage drop across it becomes larger than 0.1V or so.
The currents here are so low that there's negligible heating, nothing gets even warm at all, and the core of the transformer should be very far from saturation as well.
Perhaps some insulation problem? Although that I really doubt there is any at these low voltages.
Plus, my amplifier is really really far from distortion, it's maximum voltage swing is more than 5x the maximum voltage swing here, and R1 alone is larger than its minimum load impedance (R1 is 5.4 ohm, it works even with ).
What could be happening?
Perhaps I should take a normal transformer and repeat the test using its secondary as an inductor, see if this repeats.
Then I tried using it as an inductor (using the two outer taps) in another application, but it has a very strange behavior: It only seems to let VERY low frequencies, like < 5Hz pass properly, and surely I don't expect it to has such a huge inductance, it only had 150 turns of a thick gauge wire, half the amount of turns it had in its original secondary.
More strange even, frequencies a bit above that (like 20 Hz+) seem to be passed all with equal amplitude, instead of having a decay up to a few kHz at least, before the iron core presents lower permeability.
Its DC resistance is 0.5 ohms and there doesn't seem to be any poor contact or fracture in the wires.
Then, I went to try and measure its impedance VS frequency curve, using an audio amplifier and a resistance in series with it, but even before I found something even more strange: It's behaving non-linearly in the SAME frequency (60 Hz) in very low currents (< 100 mA), much before core saturation or heating should have any effect.
To test if there was no non-linearity in my measurement equipment, I first tested the setup with the reference resistance (R1) plus a load resistor (R2) in series. As expected, the voltage drop in my reference resistance (R1) was proportional to the voltage drop in the load resistor (R2), and crosses the zero as it should. The slope was exactly was it should be too, from its resistance (20 ohm).
Then, I made another measurement, in the same frequency, around the same range of voltages etc, and measured the voltage drop across the reference resistor and the inductor, and it was very nonlinear:
Sure, an inductor is a reactive load, but this should not make it nonlinear. The voltage drop across its legs should still be proportional to the current, which is proportional to the voltage drop across R1, disregard of any phase difference introduced by the inductance (which only affects the difference between the sum of these RMS voltages and the total RMS voltage).
The transformer is behaving as it had a larger impedance at higher voltages, after the voltage drop across it becomes larger than 0.1V or so.
The currents here are so low that there's negligible heating, nothing gets even warm at all, and the core of the transformer should be very far from saturation as well.
Perhaps some insulation problem? Although that I really doubt there is any at these low voltages.
Plus, my amplifier is really really far from distortion, it's maximum voltage swing is more than 5x the maximum voltage swing here, and R1 alone is larger than its minimum load impedance (R1 is 5.4 ohm, it works even with ).
What could be happening?
Perhaps I should take a normal transformer and repeat the test using its secondary as an inductor, see if this repeats.
