Inductance measurement problem

[CJ]

ok we checked the resistor method against the v/i method at a set of low freqs, here is what we found:


resistor method - v/i method
10 hz - 91.0 H----88.6 H
15 hz - 88.2 H---86.6 H
20 hz - 83.5 H---84.7 zh
30 hz - 81.0 H---81.4 H
40 hz - 78.0 H---77.9 H

matches up pretty darn good, any error is probably non exact resistor-voltage nulling as it can be a tedious process,  so probably no need to use smaller resistors and compensate for Z angle,

experiment over-back to the 217 xfmr 

 

[MicDaddy]

CJ was right about 20Hz saturation, at lower and lower frequencies it takes smaller levels to mangle the waveform. The max voltage I could put across a 100 turns winding was  ~100mV.  Anything higher and the waveform starts to mangle. As frequency increases, less mangling.

1000 turns took the whole 1.22V without distorting at all. (viewing the waveform on the scope)

For added practice I'm reading DaveP's write up and will try the PRR method and see how things line up.

I'll stop gathering data above 2kHz    still learning.

Thanks everyone for the discussion!  ;D

 

[PRR]

> As frequency increases, less mangling.

As a rough approximation: voltage proportional to frequency.

10Hz = 0.05V
20Hz = 0.1V
40Hz = 0.2V
80Hz = 0.4V
160Hz = 0.8V?(*)
320hz = 1.6V??

(*) Here we must stop extrapolating and go back to the lab, actually beat the iron and watch the damage. The eddy effect will be throwing flux out of the lams significantly by 200Hz-400Hz (depending on lam thickness, alloy, and bulk resistance).

 

[CJ]

there are a lot of constants in the flux formula,

they are turns, lam size, stacking factor , Faraday's constant of 100,000,000 on top,

so from B=100,000,000 V-rms / 4.44 K1 A N f,

so for a particular transformer, we can simplify the formula to show what is happening,

Flux = V-rms / frequency

so now we can say

as voltage goes up, flux goes up, in a linear fashion,

as frequency goes up, flux goes down, in a hyperbolic way (1/x)