Recently, I built the “Noise Inspector” with the objective to help me measuring inductance and transformers. After working with it for a little while I realised that the proposed testing circuit wasn’t correct. The bias current of the inductor under test won’t allow the use of the noise inspector unless the loop back to the variable HT supply is done before the sensing resistor:
Yet, there is a limitation with the above circuit. Unless we add an amplifier between the generator and the test circuit, the DUT will only see a couple of volts (rms) and measured inductance will be lower than the expected. To see this effect, let’s do a simple measurement of the inductance at different voltages. If we simply use a VARIAC we could find the inductance as a function of the current and voltage measured with accurate meters:
The series resistance cannot be measured accurately with a multimeter, so I used my bench supply and a CCS to derive the resistance by measuring the voltage across the inductor under DC conditions.
I used the Lundahl’ s LL1673/20H choke for my tests. I found that the series resistance is 58Ω. The following diagram shows the inductance as a function of the applied voltage:
We can see that when inductor is driven at very low voltages (e.g. less than 2Vrms) then the inductance drop from the expected 20H down to 16-17H.
Now, if we want to obtain the 6-model component values that can represent this inductor with a reasonable accuracy (i.e. 1-5%) then we use the proposed first test circuit to obtain the following results:
We can see from the previous plot that the model adjusts very well with exception of two clear points. At HF (above 200kHz) the self inductance of the test leads and resistance add a zero, which there is no point to measure beyond 200kHz or to factor this in the model. At very low frequency discrepancy is evident due to the adjustment of the model to a lower inductance in the resonance region as well covered by Morgan Jones in his latest edition of “Building Valve Amplifiers”.
I drove the inductor under test with about 1-1.5Vrms so expected inductance should be about 16-17H. From the model above we can see that total inductance is about 16.7H. To produce an accurate result I should have added an amplifier between the generator and the test circuit to excite the inductor properly.
What is interesting to see though is the shunt capacitance and resistance of this inductor which help us improve the accuracy of our simulations during the design process.
Very interesting indeed. What frequency did you do these tests at?