Inductor Tester II

Michael Monkenbusch

Abstract
This is a second part of the description of a versatile simple inductor tester to be used in
combination with a digital oscilloscope (https://www.elektormagazine.com/labs/simple-tool-to-
assess-inductors-and-transformers-with-100nh-l-100h). This part pertains the quick assessment
of non-linearities and saturation effects of small inductors. Giving hints on the maximum usage
currents.

Introduc on
Inductors that contain magnetic cores are subject to non-linear behaviour due to the
hysteresis properties of the core material. The detailed properties of a speci ic inductor
depend on the material properties and its geometry and the number (and position) of
winding turns. Typical effects are the loss of inductance with increasing current, sensitivity
of the inductance on external magnetic ields, inconsistent results of inductance
measurements by different measurements (size of excitation). The main cause is the
maximum possible magnetisation of the core material leading to saturation beyond a
strong enough excitation (=current × windings).
Here a tiny circuit is presented that in combination with an oscilloscope allows a quick
assessment of the saturation approach. The circuit suddenly applies a voltage to the
inductor under test and “measures” the current via a (small) voltage across a shunt resistor.
Due to the inductance the current starts a zero and increases more or less linearly with
time. As soon as this current i.e the shunt voltage voltage exceeds a preset limit the applied
voltage will be cut off. After a waiting time of some 100 ms the cycle is repeated.
Monitoring the current rise in the inductor with the oscilloscope allows easy assessment of
the non-linearities and saturation effects. Examples are shown below.
Further quantitative analysis (also shown below) may yield inductance values as function
of current.
ti
f
f
However, combined with the pulse excitation ringing frequency test to get the inductance
value and the current-increase test described here combined in a tester (shown in igure 1)
allows to get a comprehensive quick impression on the kind of inductor.

Figure 1: The tester

The circuit, schema cs

The schematics of the here discussed part of the tester is shown in igure 2.

The schematics
ti
f
f
Figure 2: Schematics

The inductor to be tested Lx is indicated in the blue box central to the schematics. The low
on-resistance p-mos transistor Q1 starts conducting by low state of the output (Q) of the
NE555 timer, which is con igured ass mono- lop. This results in an increasing current
through Lx and the shunt resistor Rshunt1. The shunt voltage is compared to a preset
voltage (adjusted by the potentiometer RV1). When the voltage reached the preset
threshold the output of the LM311 comparator triggers the mono- lop and the Q output
goes high and thereby closes the mosfet-switch Q1. This ends the current rise through Lx .
The current through the shunt stops immediately. The inductor current slowly decays on its
path through D2. The suppressor diode D1 and the Zener D3 just provide some extra
(exaggerated?) protection of the attached equipment from inductive spikes. The mono- lop
time of some 100 ms (set by R1 and C1) allows for the energy draining from Lx and some
recovery. For assessment and analysis the signals Uref and Signal are supplied to an
oscilloscope. The actual realisation of the circuit is shown in igure 3.
f
f
f
f
f
 p-MOS
LM 311
Lx
NE555
shunt
Figure 3: The realisation

Principle of opera on

If an inductor is connected to a voltage source it starts with 0 current and then, depending
on the applied voltage U and the inductance L , the current rises linearly (in the ideal
inductor case) with time. Or expressed differently, the rate of current change d I /dt through
the inductor is:
dI
= U/L (1)
dt
In case of a material core the inductance L will a some stage become current dependent
L = L (I ). If saturation sets in the inductance will decrease and thus the current increase
ti
 switch

L threshold
-> open sw.
I(t)
-> L(I) sets in..

shunt
dI U
=
dt L (I )

Figure 4: The basic principle of operation: at t=0 the switch is close and for an ideal
inductor the current would rise linearly with time, when the inductor core saturates the
current rises faster. Current is measured as shunt voltage; when a threshold is reached
the switch is open again. Not shown: diode to limit the yback voltage.

will become steeper. As shown below this may be immediately visible as a kind of kink in
the current-time curves.
For a more quantitative analysis one has to observe that the Uref provided by the circuit
only approximates the true drive voltage U. The latter rather is
U(t) = Uref − I(t)(Rshunt + RLx ) (2)

where RLx is the ohmic coil resistance. Initially I(t) is small and the correction may be
ignored. At later stages it becomes relevant and also Uref may drop due to loading the
internal voltage source resistance (and/or drain of the 300μF buffer capacitor).

Examples
Air coil, no core
Let us start the assessment of the circuit by applying the tester to coil without magnetic
core. For this purpose a spool of jump-wire (100m of 1 x 0.2mm copper wire) as bought
was used as test object. Previous measurements (with other methods) gave and inductivity
of L ≃ 11mH and a resistance of Rcoil ≃ 9.7Ω (see igure [ ig:6]b). Since this is an air-coil
there should be absolutely no saturation (non-linearity) effect. However, the slope of
current rise is expected to slow down due to the effect of the coil resistance.
f
fl
f
 Figure 5 displays the observed current rise following the application of the drive voltage in
terms ot the yellow (CH1) oscilloscope trace; the cyan (CH2) trace shows the applied
voltage Uref. At the end of the proper measurement cycle (here at about 450μs) the current
threshold is reached and the MOS-FET switch is closed. The stored energy sustains the
current for a while through the reverse diode D2 which gives rise to the recovery period
until about 2ms with Uref ≃ − 0.7V . Finally the voltage sustained by the inductor drops
below the conduction threshold of the diode and the coil is more or less decoupled and
start ringing as determined by the parasitic parallel capacitance and thus dissipates the last
bits of magnetic energy. When this is inished, the system would be ready for a new cycle.
The actual timing with respect to repeated cycling is shown in igure 6. The repetition rate
can be modi ied by changing R1C1.

Figure 5: Signal obtained from a 100m (1 × 0.2 mm2) jumper wire coil (inner diameter
40mm length 50mm, L = 11mH, R=9.7W. There is no core involved that could cause non-
linearities/saturation e ects. The slight down bend of the shunt voltage (i.e. current)
curve (CH1, yellow) is caused by the reduction of the e ective drive voltage due to the
ohmic resistance.
f
ff
f
ff
f
 Figure 6: On a larger time scale the repetition sequence of measurements cycling is
shown.

For a more quantitative analysis of the results –done here rather for assessment of the
setup and method than needed for practical use– the data from the oscilloscope were
extracted and converted to L (I ) values with equations 1 and 2. For comparison the same
coil was also analysed using the damped oscillation method (right part of igure and tester)
which was described in the previous project description. The traces and the procedures are
illustrated in igures 7 and 8.
f
f
 theory+ bkgr
for ch 0.000000 1
Coil_H5- Ldt- 5 level
slope
= 1.51929E
= 5.63710E
theory+ bkgr
for ch 1.500000 2
level = 5.85259E
slope = 0.00000E
- TRUwireC 90001 sc
x=x/xscale y=y
ch = 1.000000E
average = 1.000000E
V_ave = 1.163388E
0.4 V_ac_rms = 3.984351E
V_min
V_max
= - 3.780000E
= 2.036266E
rshunt = 4.660000E
wpart = 2.000000E
L1 = 1.038226E
nrange = 5.000000E
- fitTRUwi 89999 sca
x=x/xscale y=y
ch = 1.000000E
average = 1.000000E
V_ave = 1.163388E
V_ac_rms = 3.984351E
V_min = - 3.780000E
V_max = 2.036266E
rshunt = 4.660000E
wpart = 2.000000E
0.3 -
L1
nrange
= 1.038226E
= 5.000000E
TRUwireC 90002 sc
x=x/xscale y=y/100
ch = 2.000000E
average = 1.000000E
V_ave = 5.255531E
V_ac_rms = 2.223736E
V_min = - 2.918400E
V_max = 7.374399E
rshunt = 4.660000E
wpart = 2.000000E
numor1 = 2.000000E
numor2 = 1.000000E
L1 = 1.038226E
nrange = 5.000000E
-
0.2 fitTRUwi 89998 sca
x=x/xscale y=y/100
ch = 2.000000E
average = 1.000000E
V_ave = 5.255531E
V_ac_rms = 2.223736E
V_min = - 2.918400E
V_max = 7.374399E
rshunt = 4.660000E
wpart = 2.000000E
numor1 = 2.000000E
numor2 = 1.000000E
L1 = 1.038226E
nrange = 5.000000E
- fitTRUwi 89999 sca
x=x/xscale y=y
ch = 1.000000E
0.1 average = 1.000000E
V_ave = 1.163388E
V_ac_rms = 3.984351E
V_min = - 3.780000E
V_max = 2.036266E
rshunt = 4.660000E
wpart = 2.000000E
ssq = 2.727994E
parwght = 0.000000E
nrange = 5.000000E
- fitTRUwi 89998 sca
x=x/xscale y=y/100
ch = 2.000000E
average = 1.000000E
V_ave = 5.255531E
V_ac_rms = 2.223736E
0 V_min
V_max
= - 2.918400E
= 7.374399E
0 250 500 750 rshunt
wpart
numor1
= 4.660000E
= 2.000000E
= 2.000000E
numor2 = 1.000000E
ssq = 2.727994E
parwght = 0.000000E

Figure: 7 Same signal as shown in gure 5 but restricted to the current rise section,
created by the data read from the oscilloscope and evaluated by tting slopes and
08- 02- 2024 15:39:03 #0

values of the current rise (black) and the e ective drive voltage (red) here using the
sections where the lines are thick.
fi
ff
fi
 theory+ damped
a = 2.40764E
truwire- range- 1 Datreat Plot dt
f
= 5.29209E
= 2.18139E
damping = 1.01192E
theory+ bkgr
level =- 2.29993
slope = 0.00000E
TRUwireC 90001 s
x=x*1000 y=y
ch = 1.000000E
average = 1.000000E
V_ave = - 5.891734
V_ac_rms = 2.707045E
V_min = - 2.356133
2 V_max
xstart
= 2.366533E
= 2.300000E
xend = 2.300000E
maxcur = 1.551053E
- fitTRUwi 89999 sca
x=x*1000 y=y
ch = 1.000000E
average = 1.000000E
V_ave = - 5.891734
V_ac_rms = 2.707045E
V_min = - 2.356133
V_max = 2.366533E
L_4n7 = 1.132535E
R_eff = 2.238386E
xstart = 2.300000E
xend = 2.300000E
maxcur = 1.551053E
- fitTRUwi 89999 sca
x=x*1000 y=y
ch = 1.000000E
average = 1.000000E
V_ave = - 5.891734
V_ac_rms = 2.707045E
V_min = - 2.356133
V_max = 2.366533E
L_4n7 = 1.132535E
R_eff = 2.238386E
ssq = 9.081324E
0 parwght = 0.000000E
xstart
xend
= 2.300000E
= 2.300000E
maxcur = 1.551053E

-2

0 1 2 3

Figure 8: Data obtained by using the other part ot the inductivity tester (see the my previous
project description). The damped ringing from the inductivity parallel to a 4.7 nF capacitor
07- 02- 2024 16:49:14 #0

(circles) is tted using the green coloured zone, the rest is computed from the obtained
parameters: f=21.8139 kHz, t=1.012 ms yielding L = 11.33 mH.
fi
 theory+ bkgr
level = 1.04093E+01+- 3.33E- 01 1.0E- 03
simple Coil slope
TRUH5
= 0.00000E+00+- 0.00E+00 0.0E+00
90003 scale 1.000000E+00
x=x y=y*1000
- fitTRUH5 - 90003 scale 1.000000E+00
x =x y=y*1000
- TRUH5 90003 scale 1.000000E+00
x=x y=y*1000
- fitTRUH5 - 90003 scale 1.000000E+00
x =x y=y*1000

10.0

7.5

5.0

2.5

0
0 0.2 0.4

Figure
08- 02- 2024 9: L-values
15:48:07 #0 obtained by evaluating the current rise data (see gure
7 at increasing current values along the curve using combining slope and
e ective drive voltage yield a constant value of L =0.41 mH indicated by
the red horizontal line. The deviation a larger current probably stem from
uncertainty of the drive voltage correction. The current slope method has
less accuracy due to this uncertainty and the combined errors of shunt
and o set amplitudes. However, as discussed below its main purpose and
virtue is the assessment of the useful current range, i.e. the onset of
saturation e ects.

The extraction of the inductance L (I ) using the evaluation indicated in the in igure 7 is
shown here: initially a current independent inductance of 10.4 mH is extracted. At larger
ff
ff
ff
fi
f
 currents the correction for internal coil resistance seem to deviate (this may be further
investigated ??).

Ring core choke

Now a irst non-trivial example, where the saturation effects become obvious is presented.
Figure 10 shows the small (blue) ring core (iron powder??) with 10 turns of red Cu-wire. In
igure 11 –as in the example above– the current I(t) trace, CH1 together with the Uref trace,
Ch2 are shown. Here the current rises super linear indicating the onset of core saturation
effects.

Figure 10: Ring choke with strong saturation e ect the photo shows the
inductor attached to the tester.
f
f
ff
 Figure 11: Oscilloscope traces: yellow the shunt voltage (Rshunt = 0.466 Ohm ), blue: Uref.
The current (shunt voltage) initially rises linearly but around 100 mV (200 mA)
saturation e ects drastically accelerate the current rise. After stopping the current
increase the stored energy is dissipated in D2 during the period from about 25-200
micro-s followed by a damped oscillation when the D2 stops conduction at voltages
below about 0.7 V.

As in the previous example also here the extracted oscilloscope data were further
scrutinised as depicted in igure 12. The let side I(t) clearly shows that beyond a current of
about 0.2 A non-linearity is prominent, the inductance is more and more reduce with
further current increase. This is immediately clear by visual inspection without further
analysis (i.e. the intended use of this tester). But more details are also visible in the analysis
of the damped oscillation (other side of tester). In igure 13 it is shown that the initial
oscillation does not it until the end of range. This is due to the interlay of the current
dependent L and the amplitude of the oscillations even if the maximum current during
the damped oscillation is only 7.2 mA!).
ff
f
f
f
 theory+ bkgr
for ch 0.000000 1.500000
blue_ring_choke_10turns- Ldt- 1 level
slope
= 5.54089E- 03+- 3.38E- 06 1.0E
= 1.43290E+04+- 4.14E+00 1.0E
theory+ bkgr
for ch 1.500000 2.500000
level = 7.40855E+00+- 5.56E- 07 1.0E
slope = 0.00000E+00+- 0.00E+00 0.0E
- LdtBluer 90001 scale 1.000000E+00
x=x/xscale y=y
ch = 1.000000E+00
average = 1.000000E+01
V_ave = 4.939531E- 03
V_ac_rms = 2.783344E- 02
V_min = - 2.335733E- 01
1.25 V_max
rshunt
= 6.962133E- 01
= 4.660000E- 01
wpart = 2.000000E+01
L1 = 5.170311E- 04
nrange = 1.000000E+00
- fitLdtBl 89999 scale 1.000000E+00
x=x/xscale y=y
ch = 1.000000E+00
average = 1.000000E+01
V_ave = 4.939531E- 03
V_ac_rms = 2.783344E- 02
1.00 V_min
V_max
= - 2.335733E- 01
= 6.962133E- 01
rshunt = 4.660000E- 01
wpart = 2.000000E+01
L1 = 5.170311E- 04
nrange = 1.000000E+00
- LdtBluer 90002 scale 1.000000E+00
x=x/xscale y=y/100
ch = 2.000000E+00
average = 1.000000E+01
V_ave = 1.343361E- 02
0.75 V_ac_rms = 1.307808E+00
V_min
V_max
= - 8.738132E+00
= 8.137599E+00
rshunt = 4.660000E- 01
wpart = 2.000000E+01
numor1 = 2.000000E+00
numor2 = 1.000000E+00
L1 = 5.170311E- 04
nrange = 1.000000E+00
- fitLdtBl 89998 scale 1.000000E+00
x=x/xscale y=y/100
0.50 ch = 2.000000E+00
average = 1.000000E+01
V_ave = 1.343361E- 02
V_ac_rms = 1.307808E+00
V_min = - 8.738132E+00
V_max = 8.137599E+00
rshunt = 4.660000E- 01
wpart = 2.000000E+01
numor1 = 2.000000E+00
numor2 = 1.000000E+00
L1 = 5.170311E- 04
0.25 -
nrange
fitLdtBl
= 1.000000E+00
89999 scale 1.000000E+00
x=x/xscale y=y
ch = 1.000000E+00
average = 1.000000E+01
V_ave = 4.939531E- 03
V_ac_rms = 2.783344E- 02
V_min = - 2.335733E- 01
V_max = 6.962133E- 01
rshunt = 4.660000E- 01
wpart = 2.000000E+01
ssq = 2.422812E+07
0 parwght = 0.000000E+00
nrange = 1.000000E+00
- fitLdtBl 89998 scale 1.000000E+00
x=x/xscale y=y/100
ch = 2.000000E+00
average = 1.000000E+01
V_ave = 1.343361E- 02
V_ac_rms = 1.307808E+00
V_min = - 8.738132E+00
V_max = 8.137599E+00
0 5 10 15 20 rshunt
wpart
numor1
= 4.660000E- 01
= 2.000000E+01
= 2.000000E+00
numor2 = 1.000000E+00
ssq = 2.422812E+07
parwght = 0.000000E+00

Figure
06- 02- 2024 12: Oscilloscope
11:08:30 #0 data as in gure 11 , note the onset of saturation
e ects beyond about 0.2 A.
ff
fi
 theory+ damped
a = 2.33804E+00+- 1.72E- 07 1.0E
blue_ring_choke_10turns- range- 1 dt
f
= 2.83293E- 08+- 1.15E- 13 4.0E
= 1.03038E+05+- 4.57E- 04 2.5E
damping = 8.23343E- 05+- 1.90E- 11 4.0E
theory+ bkgr
level =- 3.06168E- 03+- 4.39E- 08 1.0E
slope = 0.00000E+00+- 0.00E+00 0.0E
LdtBluer 90001 scale 1.000000E+00
x=x*1000 y=y
ch = 1.000000E+00
average = 1.000000E+01
V_ave = - 2.299417E- 03
V_ac_rms = 3.836882E- 01
V_min = - 2.223066E+00
2 V_max
xstart
= 2.378000E+00
= 4.879999E- 06
xend = 4.879999E- 05
maxcur = 7.116680E- 03
- fitLdtBl 89999 scale 1.000000E+00
x=x*1000 y=y
ch = 1.000000E+00
average = 1.000000E+01
V_ave = - 2.299417E- 03
V_ac_rms = 3.836882E- 01
V_min = - 2.223066E+00
V_max = 2.378000E+00
L_4n7 = 5.074561E- 04
R_eff = 1.232673E+01
xstart = 4.879999E- 06
xend = 4.879999E- 05
maxcur = 7.116680E- 03
- fitLdtBl 89999 scale 1.000000E+00
x=x*1000 y=y
ch = 1.000000E+00
average = 1.000000E+01
V_ave = - 2.299417E- 03
V_ac_rms = 3.836882E- 01
V_min = - 2.223066E+00
V_max = 2.378000E+00
L_4n7 = 5.074561E- 04
R_eff = 1.232673E+01
ssq = 1.276442E+07
0 parwght = 0.000000E+00
xstart
xend
= 4.879999E- 06
= 4.879999E- 05
maxcur = 7.116680E- 03

-2

0 0.1 0.2 0.3 0.4

Figure 13: Oscilloscope data and t of test of the coil with the damped
06- 02- 2024 10:03:32 #0
oscillation method (right).

The results of the quantitative analysis (of igure 13 yields L (I ) and Lef f and τdamping from
the damped oscillation. The latter is done by itting various zone of the damped oscillation,
n indicated the number of zero crossings after which the range started. In this case the
inductance shows a drastic drop for currents beyond 0.2 A and even the low current
oscillation reveals a slight decrease of damping (=increase of τ ) when the amplitude gets
lower.
fi
f
f
 bluering 90001 scale 1.000000E+00 bluering 90010 scale 1.000000E+00
x=(X) y=y*1000 x=x y=y*1000
blue ringcore choke with 10 truns L(I) - bluering 90001 scale 1.000000E+00
x=(X) y=y*1000 blue ring choke 10turns L from ringing - bluering
x=x y=y*1000
90020 scale 1.000000E+00

0.6 0.6

0.4 0.4

0.2 0.2

0 0
0 0.25 0.50 0.75 1.00 0 25 50 75 100

Figure 14: Oscilloscope data as in gures 11-13 and test of the coil with the damped
08- 02- 2024 13:59:01 #0 08- 02- 2024 14:07:47 #0

oscillation method (right). In the latter the t is performed over region between the n-th
to the (n-10)-th zero crossing of the oscillation

Ring core choke, core magne zed

As a side remark and demonstration of the effects of saturating the magnetic lux in the
core here we look at the same ring choke as in the previous section, but with a NdB-
magnet attached to it (see igure 15). The effect of this strong external magnetic
disturbance is drastic, as the lower part of the igure illustrates the inductance of the coil is
100 fold reduced (i.e. the core has virtually lost its effect due to saturation) as is illustrated
in igure 16.
f
f
fi
ti
fi
f
f
 NdB-magnet

Figure 15: Ring core choke with attached NbB magnet

disc.

theory+ bkgr nrange = 3.000000E+00 theory+ damped

for ch 0.000000 1.500000 a = 2.33124E+00+- 1.75E- 07 1.0E+00
blue_ring_choke_10turnsMAG- Ldt- 3 level
slope blue_ring_choke_10turnsMAG- range- 1
= 4.43801E- 02+- 1.05E- 05 1.0E- 03
= 1.86402E+06+- 2.81E+01 1.0E+02
dt
f
= 4.74678E- 08+- 1.22E- 14 7.0E- 07
= 1.02225E+06+- 4.59E- 03 1.4E+04
theory+ bkgr damping = 1.17255E- 05+- 3.89E- 12 7.0E- 05
for ch 1.500000 2.500000 theory+ bkgr
level = 6.67760E+00+- 5.24E- 07 1.0E+00 level = 2.26485E- 03+- 4.88E- 08 1.0E- 02
slope = 0.00000E+00+- 0.00E+00 0.0E+00 slope = 0.00000E+00+- 0.00E+00 0.0E+00
- LdtBluer 90001 scale 1.000000E+00 LdtBluer 90001 scale 1.000000E+00
x=x/xscale y=y x=x*1000 y=y
ch = 1.000000E+00 ch = 1.000000E+00
average = 1.000000E+01 average = 1.000000E+01
V_ave = 2.797135E- 02 V_ave = 2.947680E- 03
V_ac_rms = 1.967675E- 01 V_ac_rms = 3.968490E- 01
V_min = - 4.139999E- 01 V_min = - 2.184533E+00
V_max
rshunt
= 1.849267E+00
= 4.660000E- 01
2 V_max
xstart
= 2.149733E+00
= 4.900000E- 07
wpart = 2.000000E+01 xend = 4.900000E- 06
L1 = 3.582372E- 06 maxcur = 7.038792E- 02
nrange = 3.000000E+00 - fitLdtBl 89999 scale 1.000000E+00
- fitLdtBl 89999 scale 1.000000E+00 x=x*1000 y=y
x=x/xscale y=y ch = 1.000000E+00
3 ch = 1.000000E+00
average = 1.000000E+01
V_ave = 2.797135E- 02
average = 1.000000E+01
V_ave = 2.947680E- 03
V_ac_rms = 3.968490E- 01
V_ac_rms = 1.967675E- 01 V_min = - 2.184533E+00
V_min = - 4.139999E- 01 V_max = 2.149733E+00
V_max = 1.849267E+00 L_4n7 = 5.156465E- 06
rshunt = 4.660000E- 01 R_eff = 8.795266E- 01
wpart = 2.000000E+01 xstart = 4.900000E- 07
L1 = 3.582372E- 06 xend = 4.900000E- 06
nrange = 3.000000E+00 maxcur = 7.038792E- 02
- LdtBluer 90002 scale 1.000000E+00 - fitLdtBl 89999 scale 1.000000E+00
x=x/xscale y=y/100 x=x*1000 y=y
ch = 2.000000E+00 ch = 1.000000E+00
average = 1.000000E+01 average = 1.000000E+01
V_ave = 6.549421E- 02 V_ave = 2.947680E- 03
V_ac_rms = 1.265391E+00
2 V_min
V_max
= - 8.738132E+00
= 8.574399E+00
V_ac_rms = 3.968490E- 01
V_min
V_max
= - 2.184533E+00
= 2.149733E+00
rshunt = 4.660000E- 01 L_4n7 = 5.156465E- 06
wpart = 2.000000E+01 R_eff = 8.795266E- 01
numor1 = 2.000000E+00 ssq = 2.250952E+07
numor2 = 1.000000E+00
L1
nrange
= 3.582372E- 06
= 3.000000E+00
0 parwght = 0.000000E+00
xstart
xend
= 4.900000E- 07
= 4.900000E- 06
- fitLdtBl 89998 scale 1.000000E+00 maxcur = 7.038792E- 02
x=x/xscale y=y/100
ch = 2.000000E+00
average = 1.000000E+01
V_ave = 6.549421E- 02
V_ac_rms = 1.265391E+00
V_min = - 8.738132E+00
1 V_max
rshunt
= 8.574399E+00
= 4.660000E- 01
wpart = 2.000000E+01
numor1 = 2.000000E+00
numor2 = 1.000000E+00
L1 = 3.582372E- 06
nrange = 3.000000E+00
- fitLdtBl 89999 scale 1.000000E+00
x=x/xscale y=y
ch = 1.000000E+00
average = 1.000000E+01
V_ave = 2.797135E- 02
V_ac_rms = 1.967675E- 01
V_min = - 4.139999E- 01
V_max = 1.849267E+00
0 rshunt
wpart
ssq
= 4.660000E- 01
= 2.000000E+01
= 1.860133E+08
parwght = 0.000000E+00
nrange = 3.000000E+00
- fitLdtBl -2
89998 scale 1.000000E+00
x=x/xscale y=y/100
ch = 2.000000E+00
average = 1.000000E+01
V_ave = 6.549421E- 02
V_ac_rms = 1.265391E+00
V_min = - 8.738132E+00
V_max = 8.574399E+00
0 0.5 1.0 1.5 2.0 2.5 rshunt
wpart
numor1
= 4.660000E- 01
= 2.000000E+01
= 2.000000E+00 0 0.02 0.04 0.06 0.08
numor2 = 1.000000E+00
ssq = 1.860133E+08
parwght = 0.000000E+00

06- 02- 2024 11:44:27 #0

06- 02- 2024 11:37:08 #0

.Figure 16: Oscilloscope data as in (left) and test of the coil with the damped
oscillation method (right) from the ring choke with attached magnet. Note the
changed time scale! The resulting e ective inductance here comes out to be L
with magnet as about 5 µH instead of L plain about 500 µH
ff
(No more) current compensated choke
Another very clear demonstration was achieved by (mis)using a current compensated
choke, which for this demo was rewired such that the two windings are no longer parallel
but in series such that their magnetisation effects add-up. Each winding had about 10 mH,
thus in series its now 22 = 4 times = 40 mH (see igure 17). The saturation effect is very
prominent and sets in at currents as low as about 10 mA and with a drastic kink in the I(t)
trace at about 50μs.

Figure 17: Small 2x 10mH current compensated choke, here the

two windings are not parallel but in series yielding L=40 mH.
f
 Figure 18: The saturation e ect is very prominent and sets in at currents as low
as about 10 mA and with a drastic kink in the I(t) trace at about 50 µs.

In igure 19 again the extracted I(t) and the damped oscillation probe data are shown it
some itted curves.
f
f
ff
f
 theory+ bkgr n
for ch 0.000000 1.500000
blue_rect_chope10- ser- 10mH- Ldt- 3 level
slope
= 3.51607E- 04+- 9.54E- 06 1.0E- 03
= 1.77284E+02+- 8.99E- 01 1.0E+02
theory+ bkgr
for ch 1.500000 2.500000
level = 7.36554E+00+- 9.10E- 07 1.0E+00
slope = 0.00000E+00+- 0.00E+00 0.0E+00
- Lsat_blu 90001 scale 1.000000E+00
x=x/xscale y=y
ch = 1.000000E+00
average = 1.000000E+01
V_ave = 8.893330E- 04
V_ac_rms = 7.424350E- 03
V_min = - 2.501333E- 02
0.25 V_max
rshunt
wpart
= 1.323360E- 01
= 4.660000E- 01
= 2.000000E+01
L1 = 4.154665E- 02
nrange = 3.000000E+00
- fitLsat_ 89999 scale 1.000000E+00
x=x/xscale y=y
ch = 1.000000E+00
average = 1.000000E+01
V_ave = 8.893330E- 04
V_ac_rms = 7.424350E- 03
V_min = - 2.501333E- 02
V_max = 1.323360E- 01
0.20 rshunt
wpart
= 4.660000E- 01
= 2.000000E+01
L1 = 4.154665E- 02
nrange = 3.000000E+00
- Lsat_blu 90002 scale 1.000000E+00
x=x/xscale y=y/100
ch = 2.000000E+00
average = 1.000000E+01
V_ave = - 8.356457E- 02
V_ac_rms = 1.539732E+00
V_min = - 6.744266E+00
V_max = 7.375999E+00
0.15 rshunt
wpart
numor1
= 4.660000E- 01
= 2.000000E+01
= 2.000000E+00
numor2 = 1.000000E+00
L1 = 4.154665E- 02
nrange = 3.000000E+00
- fitLsat_ 89998 scale 1.000000E+00
x=x/xscale y=y/100
ch = 2.000000E+00
average = 1.000000E+01
V_ave = - 8.356457E- 02
V_ac_rms = 1.539732E+00
V_min = - 6.744266E+00
0.10 V_max
rshunt
= 7.375999E+00
= 4.660000E- 01
wpart = 2.000000E+01
numor1 = 2.000000E+00
numor2 = 1.000000E+00
L1 = 4.154665E- 02
nrange = 3.000000E+00
- fitLsat_ 89999 scale 1.000000E+00
x=x/xscale y=y
ch = 1.000000E+00
average = 1.000000E+01
V_ave = 8.893330E- 04
0.05 V_ac_rms = 7.424350E- 03
V_min
V_max
= - 2.501333E- 02
= 1.323360E- 01
rshunt = 4.660000E- 01
wpart = 2.000000E+01
ssq = 1.175500E+05
parwght = 0.000000E+00
nrange = 3.000000E+00
- fitLsat_ 89998 scale 1.000000E+00
x=x/xscale y=y/100
ch = 2.000000E+00
average = 1.000000E+01
V_ave = - 8.356457E- 02
0 V_ac_rms = 1.539732E+00
V_min = - 6.744266E+00
V_max = 7.375999E+00
0 20 40 60 rshunt
wpart
numor1
= 4.660000E- 01
= 2.000000E+01
= 2.000000E+00
numor2 = 1.000000E+00
ssq = 1.175500E+05
parwght = 0.000000E+00

Figure 19: (Not-)current compensated choke responses. Note the

deviations in the oscillating decay from the initial part t (see gure 21
07- 02- 2024 14:21:52 #0

for resulting parameters).

fi
fi
 theory+ damped
a = 2.44421E+00+- 2.03E- 07 1.0E+00
blue_rect_chope10- ser- 10mH- range- 1 dt
f
= 5.50227E- 07+- 1.28E- 12 6.0E- 05
= 1.05090E+04+- 5.38E- 05 1.7E+02
damping = 6.49540E- 04+- 1.37E- 10 6.0E- 03
theory+ bkgr
level = 4.80534E- 02+- 4.80E- 08 1.0E- 02
slope = 0.00000E+00+- 0.00E+00 0.0E+00
Lsat_blu 90001 scale 1.000000E+00
x=x*1000 y=y
ch = 1.000000E+00
average = 1.000000E+01
V_ave = 4.280821E- 02
V_ac_rms = 3.267216E- 01
V_min = - 2.266400E+00
2 V_max
xstart
= 2.489466E+00
= 4.899999E- 05
xend = 4.899999E- 04
maxcur = 7.589486E- 04
- fitLsat_ 89999 scale 1.000000E+00
x=x*1000 y=y
ch = 1.000000E+00
average = 1.000000E+01
V_ave = 4.280821E- 02
V_ac_rms = 3.267216E- 01
V_min = - 2.266400E+00
V_max = 2.489466E+00
L_4n7 = 4.877358E- 02
R_eff = 1.501789E+02
xstart = 4.899999E- 05
xend = 4.899999E- 04
maxcur = 7.589486E- 04
- fitLsat_ 89999 scale 1.000000E+00
x=x*1000 y=y
ch = 1.000000E+00
average = 1.000000E+01
V_ave = 4.280821E- 02
V_ac_rms = 3.267216E- 01
V_min = - 2.266400E+00
V_max = 2.489466E+00
L_4n7 = 4.877358E- 02
R_eff = 1.501789E+02
ssq = 5.179999E+08
0 parwght = 0.000000E+00
xstart
xend
= 4.899999E- 05
= 4.899999E- 04
maxcur = 7.589486E- 04

-2

0 2 4 6

Figure 20: (Not-)current compensated choke responses. Note the

07- 02- 2024 14:39:38 #0

deviations in the oscillating decay from the initial part t (see gure 21 for
resulting parameters).
fi
fi
 buleRD 1 scale 1.000000E+00 bluering 90010 scale 1.000000E+00
data: Lsat_blueRect0 makro: blrecdB x=x y=y*10
small blue chocke windigs in series - buleRD 1 scale 1.000000E+00
data: Lsat_blueRect0 makro: blrecdB small rectangular choke (series) - bluering 20 scale 1.000000E+00
Lsat_blueRectRing0 makro: blrecdA

0.06

0.04

0.04

0.02
0.02

0 0
0 0.02 0.04 0 25 50 75 100

07- 02- 2024 17:38:49 #0

Figure 21: Result for the analysis of the data form gure 18. On the left the drastic
08- 02- 2024 14:28:38 #0

drop of inductance from about 40 mH to only a few mH for currents above 20mA can
bee seen. The damping curves probe the maximum current at start of the damped
oscillation only to a max of 0.76 mA. Nevertheless some e ects of the oscillation
amplitude on L and tau are visible. Most clearly the damping goes down with
decreasing amplitude (tau increases). Moreover the inductance seems to be a bit
larger at slightly larger current amplitude (0.7mA) indicating that probably the highest
points in the left part of the gure may be real an not a numerical artefact...

Limita ons
The automatic cycling as well as the maximum achievable current are limited by the ohmic
coil resistance Rc . If Rc is so large that its limits the current to values that only yield shunt
voltages below the lowest adjustable threshold (RV1) the measurement cycling will not be
sustained. For small inductors currents up to about 3⋯4 A may be probed. Large inductors
may drain the buffer capacitor leading to an excessive drop of Uref.
ti
fi
fi
ff
Finally
This is the description of the nonlinear induction part of a tiny induction tester (requiring a
digital scope). The irst part using the damped oscillation frequency to estimated the
inductance is presented in my previous project description. Here the properties and use of
the second part: assessment of non-linearity/saturation effects of inductors is described.
The intended use case is the quick assessment of the maximum useful current (besides the
limitation to resistive heating) of a given inductor. For that purpose a quick look on the
oscilloscope traces should be suf icient.
f
f