Waveforms
from a 187 KHz transmitter circuit. The lower trance
is 200 milliamps per division,
measured with this current probe.
To use this probe, a wire is connected in series with the load that you
want to see the current through, and that wire is passed through the
hole in the probe. An oscilloscope is used to view and measure the
current waveform, which is calibrated in terms of volts output per amp
of input current. The sensitivty of the circuit can be multipled by
using multiple turns in the sensing winding. For example, to get 5
times the sensitivity, pass the input wire through the hole in the
toroid five times. Of cousre, it would be much better to design
the probe to have the sensitivity required for your application, but
this trick is ok if you don't have to do it too often.
This probe only measures AC current.
The Circuit
Current from the single loop of
wire passing though the hole in
the toroid induces current in the
secondary, which creates a voltage drop.
The current probe is a current transformer. A single loop of wire
passes through the center of the toroid core, inducing current in the
secondary. A 2.7 Ohm resistor is the termination resistor and a 1K pot
is used to calibrate the output voltage as a function of input current.
The resistance of the wire in the
secondary
affects the probe's low frequency
response.
The current induced in the secondary is inversely proportional to the
number of turns in the secondary. My probe has a 42 turn secondary, so
the current in the secondary is 1/42 that of the primary, which means
there are 24 milliamps of secondary current for every amp of input
current. The illustration above is a low frequency analytical model of
the probe's circuit. The secondary current causes a voltage drop across
the 2.7 Ohm sense resistor, and this voltage is further divided by the
1K potentiometer.
The sensitivity of the probe in volts out per amps input
is:
,
which in this case 1/42 x 2.7 = 64 millivolts per amp. The 1K pot is
used to adjust this sensitivity downward to 50 millivolts per amp.
After scale factor, the low frequency corner frequency is the next most
important consideration. At this frequency, sensitivity is -3 db (-30%)
from what it was at significantly higher frequencies, and below this
frequency the sensitivity is cut in half every time the frequency is
halved, which in other words, is -6 db per octave.
The low frequency corner is a function of the inductance of the
secondary and the total resistance across the secondary.
The total resistance is the sum of the resistance of the
secondary winding (Rwire below) and the termination resistance. We will
ignore the effects of the 1K pot and the connection to the
oscilloscope.
The secondary of one probe was wound with #30 wire magnet wire and the
secondary of the other was wound with #30 Kynar insulated wire wrapping
wire. A larger
wire diameter would mean better low frequency response, but at the
would
have also induced more high frequency losses because of eddy currents
resulting from the flux gradient from the core across the conductor.
The resistance of #30 copper wire is 0.104 Ohms per foot, which is 340
microohms per millimeter. One turn of wire is 34 millimeters long.
Therefore, the resistance of the secondary is:
.
To this, you have to add the value of the 2.7 Ohm termination
resistance, so total resistance is:
,
which is equal to 3.185 Ohms.
The other value needed to compute the low frequency corner is
inductance.
Where L is inductance, AL is the inductance index of the core, and N is
the number of turns in the secondary. I used a Ferroxcube 846T500
toroid
cores for these probes. A test winding of 7 turns
gave an inductance of 319 microhenries., so I calculated the AL of
these cores to be about 6.51
microhenries per turn squared.
This coil has 42 turns on it, so the inductance is:
With a total resistance of 2.7 Ohms for the termination + 0.485 Ohms
for the wire = 3.185 Ohms, and an inductance of 11.5 millihenries, the
single pole high pass RL filter value can be calculated:
.
Calibration
After assembly of the toroid inductor and the circuit board it was time
for calibration.
The calibration fixture was a FET switch, a 10 Ohm power resistor, a
large decoupling capacitor pulse generator and a bench power supply.
Using an oscilloscope, I adjusted the gate drive and bench supply
voltage so that 10 volts was switched across the 10 Ohm resistor. This
provided a 1 amp peak-to-peak square wave through the wire going
through the current probe. I adjusted the 1K pot to get 50 millivolts
peak-to-peak on the scope. Calibration was complete.
Construction
I cut a piece of pre punched fiberglass
circuit board to an appropriate size and drilled a hole in the spot
over which I mounted the toroid. I made matching holes in the plastic
box and the plastic cover.
The toroid is held in place by a couple of plastic wire ties (Panduit
brand tie-wraps to be precise) and inserted a length of vinyl
insulation I stripped off of a piece of multiconductor cable though the
hole in the toroid and the hole in the circuit board. The termination
resistor and the calibration pot are mounted on the circuit board,
their leads bent over and soldered to each other and the leads from the
transformer. The other probe used the barrel from a BIC ballpoint pen
as the insulted sleeve, but this was rigid and not very forgiving of
mechanical misalignment of the holes, thus the use of vinyl in this one.
The shielded cable with the BNC connector on it was fastened to the
circuit board, the vinyl sleeve was installed, and the board was
fastened into the box with a few dabs of hot melt glue. After
calibration, the cover was fastened into place with the screws and the
probe was ready to go.
Plastic boxes are very nice for this
sort of use. They are easy to cut holes in and there is no worry about
the box shorting to the circuit under test, as might happen with a
metal box.