If you can technically understand it this is a link to a paper by Bruce Candy:
http://www.minelab.com/__files/f/11043/METAL%20DETECTOR%20BASICS%20AND%20THEORY.pdf
The most important concept to understanding how the "frequencies" are derived is "multi-period rectangular waves". Now an obvious question is why do I understand it? And the most relevant answer is that I've designed circuitry that utilizes "multi-period rectangular waves" to synthesize multiple frequencies, but not for anything metal detecting related and long before the release of the above paper.
What makes it very difficult to explain the above is that it is a departure from what was long the technological base line for metal detector detector design. That being single frequency technology, akin to playing a single key on a piano. Next along came the dual frequency technology with a Fundamental frequency and a Harmonic frequency akin to playing two keys on a piano towards building a chord. And now another manufacture's use of the Fundamental plus two Harmonics to build a chord. And the other big issue which I won't even try to explain is that it's being done in the Time Domain and not the Frequency Domain.
First in order to understand why the above technology is being used is to understand the mission statement of why the FBS detectors exist. Their primary mission is to hunt for goodies in trashy parks and this means having superior discrimination. The FBS method provides the best information for discrimination with a very in depth explanation in the above paper.
So let's see if we can bring this more into layman's terms with things that are more familiar.
The FBS detectors listen in on what Minelab calls three Channels. The purpose of the three Channels is to have optimum frequencies for low conductive targets, medium conductive targets, and high conductive targets being processed all at once.
Oddly enough some of the above general concept is related to the X-terra platform. On the X-terra you have a choice of 3KHz, 7.5KHz, plus 18.75KHz coils and those frequencies are NOT harmonically related. Now let's design a new X-terra model called the Variable X-terra. It will transmit on all three frequencies at once, but it will have the ability to automatically vary it's three core frequencies widely within a certain Band(or as Minelab calls it on FBS a Channel). Let's predefine our band(Channel)limits; our 3KHz can automatically vary from 2KHz to 5Khz, our 7.5KHz from 6KHz to 12Khz, and our 18.75KHz from 15KHz to 23KHz. So we are standing in a field and turn on the Variable X-terra.....the internal computer takes a look at the conditions and says wait a minute we have some problems here. The ground is showing a strong response at 18.75KHz so off the computer goes and chooses 17KHz as our Low Conductivity Channel. Likewise 3KHz is showing EMI problems so the computer searches around and finds 4KHz as quiet and that is our High Conductivity Channel. It turns out 7.5KHz is just ducky so the computer leaves that alone. None of the chosen frequencies are harmonically related because of the use of "multi-period rectangular waves".
Why is this superior to a Fundamental plus two harmonics? Because in order to shift the Harmonics I "have to" shift the Fundamental. And if I end up in a situation where I "have to" shift the Fundamental then the two harmonics are going to shift also. Now consider where I have an EMI issue on the Fundamental frequency but when I shift it I run into an EMI issue on one of my Harmonics, you're stuck between a rock and a hard place. Using the FBS method it can arbitrarily shift the frequencies to optimize the detector. Therefore because of the "multi-period rectangular waves" Minelab can transmit 28 frequencies and listen in on the best frequency within each Channel(Band) for the three different conductivity levels.
Therefore, and I hope this is clear, the three Frequencies being listened to today may not be the same as being listened to the next day depending on local conditions. The computer is designed to pick the best three and can arrive at that decision in seconds with what could take you half an hour of twiddling to do.
Is FBS perfect? Uh no! It's not very adept at finding small gold jewelry, but on the flip side you won't be driven mad by small bits of foil while searching for gold rings either.
BarnacleBill
I think Andy's post about what is under the hood is right on the money.
