Continuous-Current Pulse Induction

[Anonymous]

Please see my post of 29 December 01 for a description of this (presumably) new approach to pulse induction. Here are some further thoughts on it.
COMPUTATION OF VOLTAGE INDUCED
The voltage induced in a target, or into the receiver coil by induction imbalance, by loop imbalance, is proportional to the rate of change of current. During the pulse itself (Phases I and III), the rate of change is proportional to the applied voltage and inversely proportional to the inductance of the transmitter coil, as previously explained. During the flyback period (Phases II and IV), the rate of change is proportional to the current and to the transmitter circuit resistance (the product of which is voltage), and inversely proportional to the transmitter coil inductance. Ordinarily the resistance of the transmitter circuit would consist primarily of the electrical resistance of the coil itself.
POWER CONSUMPTION
Over a broad range of operating parameters, the power consumption is (to a first approximation) independent of the pulse rep rate; or, looking at it in the frequency domain, the fundamental frequency. This is contrary to both conventional PI practice and conventional MF practice.
If you cut the fundamental frequency (i.e., the rep rate) in half, you're cutting the transmitter on-time in half, which, if it were a conventional PI, would cut the power consumption in half. However, in a CCPI system, when the transmitter is on, part of that time the field is returning power to the power supply. The efficiency of this process is proportional to the Q at the fundamental frequency. So, when you double the period (cut the pulse rep rate in half), the transmitter duty cycle is cut in half, but the efficiency of the on-time is also cut in half, and power consumption comes out about the same.
(Note: I am not certain the above analysis is correct. There's another way of looking at it which suggests that power may decrease as the square root of the rep rate. Maybe someone will jump into this forum with a verified answer.)
SPECTRAL CHARACTERISTICS OF THE TRANSMITTER
If the transmitter voltage were a square wave, say for instance like the Fisher CZ's, the voltage of the harmonics would be inversely proportional to the harmonic number. This puts about 90% of the voltage energy at the fundamental. The current waveform is a triangle wave, the integral of the the voltage waveform.
In a CCPI system, the current waveform is approximately a square wave-- that is, the first derivative of a triangle wave. Over a broad range, the voltage of the harmonics is proportional to their frequency. This relationship no longer holds at frequencies whose period is less than about twice that of the transmit pulse duration (Phases I and III). There is a null in the spectrum for harmonics which have a period about that of the transmit pulse duration, and the higher frequencies roll off at a 1/f rate.
The square wave (as used by the Fisher CZ) is good for finding objects within a somewhat restricted size (i.e. equivalent conductivity) range, for instance US coinage and aluminum pulltabs. The transmitted spectrum of a CCPI machine is particularly well suited to finding objects of unknown size-- for instance gold prospecting, which requires high sensitivity to tiny nuggets while retaining good sensitivity to larger nuggets as well.
LOW-Q TRANSMITTER COILS
The use of a low-Q transmitter coil (say less than unity) forfeits most of the power efficiency advantage of CCPI. However, this type of operation may prove useful in some applications. By operating at low Q, the current can be allowed to diminish to a low value which induces little voltage, facilitating discrimination of iron and high-conductivity objects which have long decay times.
There is still a substantial power advantage compared to conventional pulse induction. In CCPI, the energy for the transmit pulse comes straight from the power supply. In conventional pulse induction, the pulse which energizes the target is the flyback pulse, the energy for which is whatever is left over from the lengthly transmit period during which the energy drawn from the power supply is dissipated primarily as heat. To put it another way, CCPI is like powering the target straight from the battery, whereas conventional PI is like powering the target indirectly through an extraordinarily inefficient switching power supply.
PREAMP ARCHITECTURES
One of the disadvantages of conventional PI's is that the preamp looks at the receiver coil through a resistor typically several hundred ohms or higher, which is there to allow shunt diodes to protect the preamp from overvoltage. (The Fisher system is an exception to this.)
CCPI by its very nature requires a separate receiver coil, which would ordinarily be in approximate induction balance relative to the transmitter. This allows the preamp to operate without protection and without gating, looking straight at the coil (through a series capacitor resonant at the fundamental frequency if desired). This allows the preamp to operate at a lower noise level than is possible in conventional PI practice.
Furthermore, the preamp does not have to be operated in saturation. It can be operated at fairly low gain to avoid saturation, but at gain high enough to offer a good noise figure to the following stages.
One of the difficulties in conventional multifrequency induction balance practice is in maintaining reactive alignment of the searchcoil. The CCPI system can be used to make this requirement less stringent. Either at the receiver coil itself or at the output of a low-gain preamp, the resistive signals can be gated into a high-gain amplifer during Phases II and IV when the reactive signals are largely absent. The reactive signals, which are "piled up" mostly into Phases I and III, can be demodulated separately at low gain. In most applications the reactive signals will be used for ground balancing and/or reactive target discrimination, which do not usually require as good a signal-to-noise ratio as the resistive signals do.
-------------
Your comments and critique are invited.
--Dave Johnson

 

[Anonymous]

Sorry, I done misspoked. In the section titled "Spectral Characteristics of the Transmitter", the voltage spectrum of a CCPI system is FLAT with frequency, not proportional to frequency.
It is well known that a unit impulse has a flat power spectrum. If the pulses are periodic, if one assigns the power of the fundamental and each harmonic uniformly to the slot of spectrum it occupies between its neighbors, then the spectrum is flat from the fundamental up to the point where pulse itself rolls off due to its finite width (since there are no perfect unit impulses in the real world).
--Dave J.

 

[Anonymous]

Those who deal primarily with VLF technology are used to thinking in the frequency-phase domain. Those who are familiar with PI think primarily in the time domain. Which way one thinks is a matter of convenience: mathematically the two ways of thinking are equivalent, in the same sense that polar and rectangular coordinate systems are equivalent.
The following discussion is not intended to be rigorously correct from a mathematical point of view. It is intended to provide insight. So please don't get hung up on those things which would get me flunked out of math class if this were a test. If you're used to thinking more rigorously, you may prefer to swap the words "sine" and "cosine" in the following blurb.
----------
Suppose we have a sinusoidal magnetic field. We find that the reactive (e.g. magnetite) response is at zero degrees, and the resistive response (e.g. salt water) is a sinusoid with a lag of 90 degrees. In other words, a cosine. And, in other words, the integral of the sinusoid.
A unit impulse is a superposition of an infinite number of sinusoids. If they are not integrated, the impulse represents the reactive component of the magnetic field.
Suppose we integrate all those sinusoids individually and sum them up. That's equivalent to integrating the unit pulse. And what is the integral of a unit pulse? a flat line. The resistive component of a unit impulse magnetic field is a flat line.
In a CCPI metal detector, the pulses are periodic and finite in width. The superposition of the reactive components of the field induces a voltage in the receiver which has nearly the same shape as the transmit voltage pulse. This is because the transmitter coil integrates the voltage to current, and the receiver coil differentiates the current back to voltage. If you integrate a function and then differentiate it, you get back to the original function.
Now, assuming the presence of a low-conductivity nonferrous target: The resistive components of the superimposed sinusoids are integrated-- that is, shifted by 90 degrees. The time domain representation of the superposition of the integrated sinusoids is a square wave-- that is, the integral of the pulses.
In actual practice, of course things aren't quite that perfect. At the fundamental you're limited by the resistive losses in the transmitter coil, and at the high frequencies you get time distortion resulting primarily from capacitances and damping in the receiver coil circuit.
Hope this helps make the relationship between time and frequency-phase in CCPI a little clearer. If any of you whizzes out there would like to provide a clearer explanation, please do so.
--Dave J.

 

[Anonymous]

In the post "a way of thinking about CCPI" I spoke of a hypothetical low-conductivity target integrating the pulse.
Oops. There aren't any physical low-conductivity targets which do that. Real low-conductivity targets have a resistive amplitude response which is proportional to frequency. When you scale the corresponding sinusoids according that rule and superimpose them, you get that short little "tail" that PI engineers are so familiar with.
In order to get that resistive square wave, you'd need a target which has a 90 degree (resistive) response at all frequencies in question, AND equal amplitude response at all frequencies. There aren't any such targets. About as close as you'd get, is some iron targets, but that's not very close at all. High-conductivity targets also have a response which is not as dependent on frequency as that of low-conductivity targets.
-- Dave J.