Emery's Projects
Inductance Meter
Building your own instruments to measure
electrical quantities and parameters of various components can
be a satisfying and worthwhile passtime. An inductance meter,
for example.
There are a great variety of circuits posted online, some are
wonderfully simple but perhaps less than ideal. They may not,
for example, account for coil resistance or self capacitance.
Each approach will have its pros and cons. An experimenter will
experiment...This is what I invite you to do:
How it works:
-The unknown inductor is inserted into the oscillator circuit
which then self adjusts to produce a precise stabilized
amplitude sinusoid. To achieve this, the oscillator amplitude is
fed back to control a negative resistance within the tank
circuit.
-The output is integrated twice in succession.
-The resulting waveform is demodulated and averaged out to give
a reading proportional to the inductance being measured.
Further details:
L1 is the inductance to be measured.
R13 sets the oscillator output amplitude, which should be at 400
mv peak.
The circuit uses a "home made" Led-Cadmium sulfid optocoupler
consisting of a single Led illuminating two separate photocells.
One of the cells is used to linearize the response, the other
for controlling the oscillator.
The low frequency noise components tend to be greatly amplified
in the process of integration, therefore a high pass filter is
an absolute must before demodulation.
High pass filters 1 and 2 are identical except for an offset
adjustment in filter 1. The purpose of filter 1 is to eliminate
noise while filter 2 assures that the phase conditions for
demodulation are correct.
Potentiometer R33 of the demodulator is meant to offset the U6
output, it should be adjusted to obtain a clean half wave
rectified signal. U6 output is meant to be as large an amplitude
as possible, without driving the LM318 into saturation.
Saturating it will result in unwanted phase shifts.
Finally, a digital voltmeter is connected to the averaged out
demodulated signal.
The circuit below will measure inductances up to about 20
milihenries. Meaningful readings can be taken even at the
milivolt level. All op amps used here were TL072-s except for op
amp U6 which is an LM318. The integrators can be scaled to any
custom range desired. The circuits here are capable of accurate
readings down to 10 microhenries.
I also include a clip showing the actual demodulated output and
control voltage levels. The measurement consecutively displays
outputs for 10, 20 and 100 uh inductors. The last measurement is
then repeated with a 200 ohm resistor put in series with the
inductor. Note that the output stays the same while the control
voltage settles at a much higher value.
Capacitance
Meter
The meter is based on the idea of charge
transfer featuring a continuous operating mode that avoids
dielectric absorption and AC hum problems.
The device was breadboarded and tested:
An opamp square wave generator was used, of approximately 1 khz
frequency. The period or the symmetry need not be exact, only
the amplitude. The complementary power mosfet driver is there to
fix the output
amplitude precisely to the well regulated supply rail of +-10
volts. This precise amplitude square signal is then applied to
the measured capacitance C3. The periodically transferred charge
is integrated by opamp U1A and the standard capacitance of C2.
The resulting square waveform is then scaled by opamp U5A and
sampled differentially with two transistor driven Fet circuits.
The final result is displayed on a high impedance digital volt
meter.(10 MOhm input, buffers would be needed for lower
impedance meters)
-Sampling is done in the center (see simulation diagram),
allowing sufficient time for full charge transfer as well as
having the added advantage of compensating for the upper and
lower edge slope altering effect of the 20M resistance across
C2. Sampling is differential as well, which eliminates AC fields
interference.
C2=10nf and R6=49k will yield a measurement range up to 100 pf-s,
the display being 10 volts at that value. 1/100 th pf will show
up as 1 mv. Output values above 12 volts are not supported
because of the limits of sampling amplitudes.
C2=100nf and R6=4k will yield a range up to 100nf. (one can try
any number of combinations here)
Observations:
I have tried this circuit up to 250 nf. For capacitance values
larger than 250nf, one must increase the oscillator period or
the the current capacity of opamp U1A. Low capacitance
measurements were stable and reproducable, easily accurate to
1/100 pf. Linearity excellent as well.
A Fast
Precision Rectifier
Opamp rectifiers are not particularly fast.
-Take for example a 15mhz bandwidth TL072CP with the basic diode rectifier
configuration as in the circuit displayed on the bottom. At an input of 100mv
100khz signal the output will show no rectification at all.
-The more standard circuitry (in the middle) employs an extra diode and a
configuration that will not let the opamp saturate in the half cycle. The
results are visibly better, but still quite weak at the low amplitude and high
frequency employed here.
-Take a look at the top circuitry. The extra components added will further
reduce the opamp output swings so that it can respond much faster during the
zero transitions. The resulting scope image speaks for itself. In this setup the
diodes will yield full temperature compensation as well.
-outputs 8,17 and 30 are signal averaging outputs. For any single type of
waveform, they will produce dc levels proportional to the input amplitude. The
resistors in these low pass RC elements must be kept at a high value so that
they will not interfere with rectified signal. This part of the circuit could
also be separated by an opamp buffer if necessary.
A Simple Regenerative Receiver
It is hard to resist the simplicity of these receivers as well
as their potential. Charles Kitchin, among others, has published fine working
models that are worth looking at.
After building some of these over the years, I settled on a list of what I
wanted in a regenerative receiver:
1-There should be no external coupling inductors, be it for antenna signal or
regeneration. A single small diameter coil only.
2-To be able to tune through the bands without plugging in a variety of
inductors.
3-Proximity of objects should not detune the receiver to a noticeable extent.
4-An antenna of a relatively short length, no more than 1.5 meters but still
sufficient to rival the sensitivity of commercial single transposition superhets.
In the entire frequency range of reception.
5-Regeneration should be smooth, without any hysteresis.
6-Bandwidth of reception should be narrow enough even at the higher frequencies.
7-Simple to build.
The design below aims to achieve all of the above.
-The single most important performance determinant feature of this circuit is
the separate demodulator. Demodulation is not done in the oscillator stage, as
is most common. The large source resistor shunt capacitor, used to obtain the
demodulated signal in the majority of regenerative receivers, will necessarily
cause non-linearities in the drain current. This limits the achievable Q.
Comparing the total harmonic distortion figures in the drain current (at a 100mv
amplitude sinusoidal gate signal) will yield .06% without and 3% with a 1nf
source resistor shunt. Leading to a 4 times increase in the minimum bandwidth
figures! Here lies the problem as well as the solution based on the addition of
a separate fet demodulator stage that will not degrade bandwidth. This
demodulator is biased to very low currents to enable great sensitivity.
-The grounded base antenna matcher is standard, its other function is to
minimize antenna proximity effects and power radiation. What is different is the
capacitive coupling.
-The throttle capacitor is connected in a way to current feed directly into the
tank inductor's grounding capacitance. Its done on the inductance side to invert
the phase. The parallel choke provides the dc path. (simply grounding the gate
with a high valued resistor instead of using the choke, is not a good idea,
leaves the gate vulnerable to AC fields)
-The Tank is slug tuned with a high-frequency ferrite rod of 10mms diameter and
7cms length. The coil's maximum inductance is about 7uh, its diameter is 12 mm,
11 turns, length at 30 mm for my particular rod. I was able to receive on
frequencies from 5.5 to 20 Mhz or so. Fine-tuning is with a 25pf variable
capacitor.
A little help for the build:
-The fequency compensator source resistor shunt capacitance C3 is there to
enable oscillation at the higher frequency bands.
-Tank inductance grounding capacitor C5 should be chosen so that the feedback is
in the proper range, to a value that would just start oscillation with the
throttle capacitance at middle setting and the tuning slug fully inserted.
-All 3 RF chokes are identical, 3mh on small 16 mm dia. toroidal cores.
-If any longer antenna is to be attached, an attenuator would be a good idea. I
found the 1.5 meter length more than sufficient to match the sensitivity of my
Realistic DX-370 digital receiver.
High Impedance Probe
I have used a shielded probe tip (pictured above), fashioned
from a 50 ohm coax and some tin can material, housing the components indicated
in the circuit diagram. The other components of the probe were bread-boarded,
not being critical for performance.
-measurements indicate a 0.02 pf input capacitance up to 100 khz, still managing
0.07pf at 1 mhz. The probe tip is of such high impedance that a mere proximity
to the point of measurement is sufficient to obtain most of the signal. (I think
that the 0.02 pf input capacitance could be lowered further by decreasing coax
diameter and decreasing tip extension length, simulations say that this is
possible) As far as amplitude transfer, accurate to about 1/10-th of one percent
below 100 khz, and less than 1% at 1 mhz. (for very low frequencies, the
blocking capacitor values become a factor)
High Voltage Generator using Mechanical
Energy
(disclaimer: High voltage kills!, anyone experimenting with
this device must be knowledgeable in handling such devices. This device, if
built, has sufficient energy to kill a human being, extreme caution is required)
This is a little "off the wall", but seems feasible and can generate
considerable power. So far it is only simulated...
Principle:
A motor driven variable capacitor generates high voltage DC at high power
levels. An electric motor drives a circular disc of a diameter approximately
10-15 cms. The motor spins at 50-60 cycles/sec. The rotating disc forms one pole
of a capacitor. It is divided into 10 symmetric sectors of conducting "pies",
evenly spaced, alternating the conducting and non coducting regions. The
stationary disk is exactly the same layout as the rotating one. These two disks
together form a variable high voltage capacitor of approximately 20pf value.
(depending on spacing and insulation )
Function:
An arrangement of 3 high voltage diode stacks and two high voltage capacitors
channel the charge into the load and the same time feedback the potential
necessary to operate the device. A battery provided startup potential is used to
initiate the buildup.
Specs:
This design will generate 20 kilovolts at 7 watts of power. The three high
voltage diode stacks and the two 500pf capacitors are rated at 25kv. The
rotating disk's total capacitance is 20 pf, rated at 40 kilovolts. (possibly
using glass insulation in between the disks)
Note: It is a "runaway" type of generator, needs regulator to avoid self
destruct.